Open NAND Flash Interface Specification Revision 2.3a 19-October-2011 Hynix Semiconductor Intel Corporation Micron Technology, Inc. Phison Electronics Corp. SanDisk Corporation Sony Corporation Spansion This 2.3a revision of the Open NAND Flash Interface specification ("Final Specification") is available for download at www.onfi.org. SPECIFICATION DISCLAIMER THIS SPECIFICATION IS PROVIDED TO YOU “AS IS” WITH NO WARRANTIES WHATSOEVER, INCLUDING ANY WARRANTY OF MERCHANTABILITY, NONINFRINGEMENT, OR FITNESS FOR ANY PARTICULAR PURPOSE. THE AUTHORS OF THIS SPECIFICATION DISCLAIM ALL LIABILITY, INCLUDING LIABILITY FOR INFRINGEMENT OF ANY PROPRIETARY RIGHTS, RELATING TO USE OR IMPLEMENTATION OF INFORMATION IN THIS SPECIFICATION. THE AUTHORS DO NOT WARRANT OR REPRESENT THAT SUCH USE WILL NOT INFRINGE SUCH RIGHTS. THE PROVISION OF THIS SPECIFICATION TO YOU DOES NOT PROVIDE YOU WITH ANY LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS. Copyright 2005-2011, Hynix Semiconductor, Intel Corporation, Micron Technology, Inc., Phison Electronics Corp., SanDisk Corporation, Sony Corporation, Spansion. All rights reserved. For more information about ONFI, refer to the ONFI Workgroup website at www.onfi.org. All product names are trademarks, registered trademarks, or servicemarks of their respective owners. ONFI Workgroup Technical Editor: Amber Huffman Intel Corporation 2111 NE 25th Ave M/S JF2-53 Hillsboro, OR 97124 USA Tel: (503) 264-7929 Email: [email protected] ii Table of Contents 1. Introduction ............................................................................................................................... 1 1.1. Goals and Objectives ........................................................................................................ 1 1.2. EZ NAND Overview .......................................................................................................... 1 1.3. References ........................................................................................................................ 1 1.4. Definitions, abbreviations, and conventions ...................................................................... 2 1.4.1. Definitions and Abbreviations .................................................................................... 2 1.4.2. Conventions ............................................................................................................... 4 2. Physical Interface ..................................................................................................................... 7 2.1. TSOP-48 and WSOP-48 Pin Assignments ....................................................................... 7 2.2. LGA-52 Pad Assignments ............................................................................................... 10 2.3. BGA-63 Ball Assignments ............................................................................................... 12 2.4. BGA-100 Ball Assignments ............................................................................................. 16 2.5. Signal Descriptions ......................................................................................................... 20 2.6. CE_n Signal Requirements ............................................................................................. 26 2.6.1. Source Synchronous Data Interface Requirements ................................................ 26 2.7. Absolute Maximum DC Ratings ...................................................................................... 26 2.8. Recommended DC Operating Conditions ....................................................................... 27 2.8.1. I/O Power (VccQ) and I/O Ground (VssQ)............................................................... 27 2.9. AC Overshoot/Undershoot Requirements ...................................................................... 28 2.10. DC and Operating Characteristics ............................................................................... 28 2.11. Calculating Pin Capacitance ....................................................................................... 31 2.12. Staggered Power-up.................................................................................................... 32 2.13. Power Cycle Requirements ......................................................................................... 32 2.14. Independent Data Buses ............................................................................................. 32 2.15. Bus Width Requirements ............................................................................................. 32 2.16. Ready/Busy (R/B_n) Requirements ............................................................................ 33 2.16.1. Power-On Requirements ...................................................................................... 33 2.16.2. R/B_n and SR[6] Relationship ............................................................................. 34 2.17. Write Protect ................................................................................................................ 34 3. Memory Organization ............................................................................................................. 35 3.1. Addressing ...................................................................................................................... 36 3.1.1. Multi-plane Addressing ............................................................................................ 37 3.1.2. Logical Unit Selection .............................................................................................. 38 3.1.3. Multiple LUN Operation Restrictions ........................................................................ 38 3.2. Factory Defect Mapping .................................................................................................. 39 3.2.1. Device Requirements............................................................................................... 39 3.2.2. Host Requirements .................................................................................................. 40 3.3. Extended ECC Information Reporting ............................................................................. 40 3.3.1. Byte 0: Number of bits ECC correctability ............................................................... 41 3.3.2. Byte 1: Codeword size ............................................................................................. 41 3.3.3. Byte 2-3: Bad blocks maximum per LUN ................................................................. 41 3.3.4. Byte 4-5: Block endurance ....................................................................................... 41 3.4. Discovery and Initialization.............................................................................................. 42 3.4.1. CE_n Discovery ....................................................................................................... 42 3.4.2. Target Initialization ................................................................................................... 42 4. Data Interface and Timing ...................................................................................................... 44 4.1. Data Interface Types ....................................................................................................... 44 4.1.1. Signal Function Reassignment ................................................................................ 44 4.1.2. Bus State ................................................................................................................. 45 4.1.3. Source Synchronous and Repeat Bytes .................................................................. 46 4.1.4. Data Interface / Timing Mode Transitions ................................................................ 47 4.2. Timing Parameters .......................................................................................................... 48 4.2.1. General Timings ....................................................................................................... 48 4.2.2. Asynchronous .......................................................................................................... 49 iii 4.2.3. Source Synchronous................................................................................................ 56 Number of LUNs per x8 data bus ........................................................................................... 64 4.3. Timing Diagrams ............................................................................................................. 69 4.3.1. Asynchronous .......................................................................................................... 69 4.3.2. Source Synchronous................................................................................................ 76 4.4. Command Examples ....................................................................................................... 88 4.4.1. Asynchronous .......................................................................................................... 88 4.4.2. Source Synchronous................................................................................................ 91 5. Command Definition ............................................................................................................... 96 5.1. Command Set ................................................................................................................. 96 5.2. Command Descriptions ................................................................................................... 99 5.3. Reset Definition ............................................................................................................. 102 5.4. Synchronous Reset Definition ....................................................................................... 102 5.5. Reset LUN Definition ..................................................................................................... 103 5.6. Read ID Definition ......................................................................................................... 104 5.7. Read Parameter Page Definition .................................................................................. 106 5.7.1. Parameter Page Data Structure Definition ............................................................ 108 5.7.2. Extended Parameter Page Data Structure Definition ............................................ 122 5.8. Read Unique ID Definition............................................................................................. 125 5.9. Block Erase Definition ................................................................................................... 127 5.10. Read Status Definition ............................................................................................... 127 5.11. Read Status Enhanced Definition ............................................................................. 130 5.12. Read Status and Read Status Enhanced required usage ........................................ 131 5.13. Status Field Definition................................................................................................ 131 5.14. Read Definition .......................................................................................................... 132 5.15. Read Cache Definition............................................................................................... 135 5.16. Page Program Definition ........................................................................................... 139 5.17. Page Cache Program Definition ................................................................................ 141 5.18. Copyback Definition................................................................................................... 144 5.19. Small Data Move ....................................................................................................... 149 5.20. Change Read Column Definition ............................................................................... 152 5.21. Change Read Column Enhanced Definition ............................................................. 152 5.22. Change Write Column Definition ............................................................................... 154 5.23. Change Row Address Definition ............................................................................... 154 5.24. Set Features Definition .............................................................................................. 156 5.25. Get Features Definition.............................................................................................. 158 5.26. Feature Parameter Definitions .................................................................................. 158 5.26.1. Timing Mode ....................................................................................................... 159 5.26.2. I/O Drive Strength ............................................................................................... 160 5.26.3. EZ NAND control ................................................................................................ 160 6. Multi-plane Operations ......................................................................................................... 161 6.1. Requirements ................................................................................................................ 161 6.2. Status Register Behavior .............................................................................................. 162 6.3. Multi-plane Page Program ............................................................................................ 162 6.4. Multi-plane Copyback Read and Program .................................................................... 165 6.5. Multi-plane Block Erase ................................................................................................ 168 6.6. Multi-plane Read ........................................................................................................... 169 7. Behavioral Flows .................................................................................................................. 174 7.1. Target behavioral flows ................................................................................................. 174 7.1.1. Variables ................................................................................................................ 174 7.1.2. Idle states ............................................................................................................... 174 7.1.3. Idle Read states ..................................................................................................... 176 7.1.4. Reset command states .......................................................................................... 178 7.1.5. Read ID command states ...................................................................................... 180 7.1.6. Read Parameter Page command states................................................................ 181 7.1.7. Read Unique ID command states .......................................................................... 182 iv 7.1.8. Page Program and Page Cache Program command states ................................. 183 7.1.9. Block Erase command states ................................................................................ 186 7.1.10. Read command states ....................................................................................... 188 7.1.11. Set Features command states ........................................................................... 190 7.1.12. Get Features command states ........................................................................... 191 7.1.13. Read Status command states ............................................................................ 191 7.1.14. Read Status Enhanced command states ........................................................... 192 7.2. LUN behavioral flows .................................................................................................... 193 7.2.1. Variables ................................................................................................................ 193 7.2.2. Idle command states .............................................................................................. 193 7.2.3. Idle Read states ..................................................................................................... 195 7.2.4. Status states .......................................................................................................... 196 7.2.5. Reset states ........................................................................................................... 197 7.2.6. Block Erase command states ................................................................................ 197 7.2.7. Read command states ........................................................................................... 199 7.2.8. Page Program and Page Cache Program command states ................................. 201 A. Sample Code for CRC-16 (Informative) ........................................................................... 205 B. Spare Size Recommendations (Informative) .................................................................... 207 C. Device Self-Initialization with PSL (Informative) ............................................................... 208 D. ICC Measurement Methodology ....................................................................................... 209 E. Measuring Timing Parameters to/From Tri-State ............................................................. 217 F. EZ NAND: End to End Data Path Protection (INFORMATIVE) ........................................... 218 v 1. Introduction 1.1. Goals and Objectives This specification defines a standardized NAND Flash device interface that provides the means for a system to be designed that supports a range of NAND Flash devices without direct design pre-association. The solution also provides the means for a system to seamlessly make use of new NAND devices that may not have existed at the time that the system was designed. Some of the goals and requirements for the specification include: Support range of device capabilities and new unforeseen innovation Consistent with existing NAND Flash designs providing orderly transition to ONFI Capabilities and features are self-described in a parameter page such that hard-coded chip ID tables in the host are not necessary Flash devices are interoperable and do not require host changes to support a new Flash device Define a higher speed NAND interface that is compatible with existing NAND Flash interface Allow for separate core (Vcc) and I/O (VccQ) power rails Support for offloading NAND lithography specific functionality to a controller stacked in the NAND package (EZ NAND) 1.2. EZ NAND Overview EZ NAND includes the control logic packaged together with NAND to perform the NAND management functionality that is lithography specific (e.g. ECC), while retaining the NAND protocol infrastructure. EZ NAND delivers an ECC offloaded solution with minimal command and/or protocol changes. The device parameter page will specify if EZ NAND is supported. 1.3. References This specification is developed in part based on existing common NAND Flash device behaviors, including the behaviors defined in the following datasheets: Micron MT29F8G08ABC data sheet available at http://www.micron.com/products/partdetail?part=MT29F8G08ABCBBH1-12 Numonyx NAND04GB2D data sheet available at http://www.numonyx.com/Documents/Datasheets/NAND04G-B2D_NAND08G-BxC.pdf The specification also makes reference to the following specifications and standards: ONFI Block Abstracted NAND revision 1.1. Specification is available at http://www.onfi.org. 1 1.4. Definitions, abbreviations, and conventions 1.4.1. Definitions and Abbreviations The terminology used in this specification is intended to be self-sufficient and does not rely on overloaded meanings defined in other specifications. Terms with specific meaning not directly clear from the context are clarified in the following sections. 1.4.1.1. address The address is comprised of a row address and a column address. The row address identifies the page, block, and LUN to be accessed. The column address identifies the byte or word within a page to access. The least significant bit of the column address shall always be zero in the source synchronous data interface. 1.4.1.2. asynchronous Asynchronous is when data is latched with the WE_n signal for writes and RE_n signal for reads. 1.4.1.3. block Consists of multiple pages and is the smallest addressable unit for erase operations. 1.4.1.4. column The byte (x8 devices) or word (x16 devices) location within the page register. 1.4.1.5. defect area The defect area is where factory defects are marked by the manufacturer. Refer to section 3.2. 1.4.1.6. device The packaged NAND unit. A device consists of one or more targets. 1.4.1.7. DDR Acronym for double data rate. 1.4.1.8. Dword A Dword is thirty-two (32) bits of data. A Dword may be represented as 32 bits, as two adjacent words, or as four adjacent bytes. When shown as bits the least significant bit is bit 0 and most significant bit is bit 31. The most significant bit is shown on the left. When shown as words the least significant word (lower) is word 0 and the most significant (upper) word is word 1. When shown as bytes the least significant byte is byte 0 and the most significant byte is byte 3. See Figure 1 for a description of the relationship between bytes, words, and Dwords. 1.4.1.9. latching edge The latching edge describes the edge of the CLK or the DQS signal that the contents of the data bus are latched on for the source synchronous data interface. For data cycles the latching edge is both the rising and falling edges of the DQS signal. For command and address cycles the latching edge is the rising edge of the CLK signal. 2 1.4.1.10. LUN (logical unit number) The minimum unit that can independently execute commands and report status. There are one or more LUNs per target. 1.4.1.11. na na stands for “not applicable”. Fields marked as “na” are not used. 1.4.1.12. O/M O/M stands for Optional/Mandatory requirement. When the entry is set to “M”, the item is mandatory. When the entry is set to “O”, the item is optional. 1.4.1.13. page The smallest addressable unit for read and program operations. 1.4.1.14. page register Register used to read data from that was transferred from the Flash array. For program operations, the data is placed in this register prior to transferring the data to the Flash array. If EZ NAND is supported a buffer exists in the EZ NAND controller that may be used to facilitate Copyback operations. Refer to section 5.18 for information on EZ NAND Copyback operations. 1.4.1.15. partial page A portion of the page, referred to as a partial page, may be programmed if the target supports more than one program per page as indicated in the parameter page. The host may choose to read only a portion of the data from the page register in a read operation; this portion may also be referred to as a partial page. 1.4.1.16. read request A read request is a data output cycle request from the host that results in a data transfer from the device to the host. Refer to section 4.1.2 for information on data output cycles. 1.4.1.17. row Refers to the block and page to be accessed. 1.4.1.18. source synchronous Source synchronous is when the strobe (DQS) is forwarded with the data to indicate when the data should be latched. The strobe signal, DQS, can be thought of as an additional data bus bit. 1.4.1.19. SR[ ] SR refers to the status register contained within a particular LUN. SR[x] refers to bit x in the status register for the associated LUN. Refer to section 5.13 for the definition of bit meanings within the status register. 1.4.1.20. target An independent Flash component with its own CE_n signal. 3 1.4.1.21. Uncorrectable Bit Error Rate, or ratio (UBER) A metric for the rate of occurrence of data errors, equal to the number of data errors per bits read. Mathematically, it may be represented as: UBER = cumulative number of data errors / cumulative number of bits read Note: The cumulative number of bits read is the sum of all bits of data read back from the device, with multiple reads of the same memory bit as multiple bits read. For example, if a 100GB device 12 is read ten times then there would be about 1TB (8x10 bits) read. The cumulative number of data errors is the count of the physical pages for which the device fails to return correct data. Note: This metric only applies to devices that support EZ NAND. EZ NAND delivers an ECC offloaded solution, and thus this metric applies. For raw NAND solutions where the host provides the ECC solution, the UBER is dependent on the host controller capability and UBER for that solution is not within the scope of this specification. 1.4.1.22. Vtt Termination voltage. 1.4.1.23. word A word is sixteen (16) bits of data. A word may be represented as 16 bits or as two adjacent bytes. When shown as bits the least significant bit is bit 0 and most significant bit is bit 15. The most significant bit is shown on the left. When shown as bytes the least significant byte (lower) is byte 0 and the most significant byte (upper) is byte 1. See Figure 1 for a description of the relationship between bytes, words and Dwords. 1.4.2. Conventions The names of abbreviations and acronyms used as signal names are in all uppercase (e.g., CE_n). Fields containing only one bit are usually referred to as the "name" bit instead of the "name" field. Numerical fields are unsigned unless otherwise indicated. 1.4.2.1. Precedence If there is a conflict between text, figures, state machines, timing diagrams, and tables, the precedence shall be state machines and timing diagrams, tables, figures, and then text. 1.4.2.2. Keywords Several keywords are used to differentiate between different levels of requirements. 1.4.2.2.1. mandatory A keyword indicating items to be implemented as defined by this specification. 1.4.2.2.2. may A keyword that indicates flexibility of choice with no implied preference. 1.4.2.2.3. optional A keyword that describes features that are not required by this specification. However, if any optional feature defined by the specification is implemented, the feature shall be implemented in the way defined by the specification. 4 1.4.2.2.4. reserved A keyword indicating reserved bits, bytes, words, fields, and opcode values that are set-aside for future standardization. Their use and interpretation may be specified by future extensions to this or other specifications. A reserved bit, byte, word, or field shall be cleared to zero, or in accordance with a future extension to this specification. The recipient shall not check reserved bits, bytes, words, or fields. 1.4.2.2.5. shall A keyword indicating a mandatory requirement. Designers are required to implement all such mandatory requirements to ensure interoperability with other products that conform to the specification. 1.4.2.2.6. should A keyword indicating flexibility of choice with a strongly preferred alternative. Equivalent to the phrase “it is recommended”. 1.4.2.3. Byte, word and Dword Relationships Figure 1 illustrates the relationship between bytes, words and Dwords. 7 6 5 4 3 2 1 0 Byte 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0 Word Byte 1 3 3 2 1 0 9 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 Byte 0 1 0 9 8 7 6 5 4 3 2 1 0 Dword Word 1 Byte 3 Word 0 Byte 2 Figure 1 Byte 1 Byte 0 Byte, word and Dword relationships 1.4.2.4. Behavioral Flow Diagrams For each function to be completed a state machine approach is used to describe the sequence and externally visible behavior requirements. Each function is composed of several states to accomplish a set goal. Each state of the set is described by an individual state table. Table 1 5 below shows the general layout for each of the state tables that comprise the set of states for the function. State name Action list Transition condition 0 Next state 0 Transition condition 1 Next state 1 Table 1 State Table Cell Description Each state is identified by a unique state name. The state name is a brief description of the primary action taken during the state. Actions to take while in the state are described in the action list. Each transition is identified by a transition label and a transition condition. The transition label consists of the state designator of the state from which the transition is being made followed by the state designator of the state to which the transition is being made. The transition condition is a brief description of the event or condition that causes the transition to occur and may include a transition action that is taken when the transition occurs. This action is described fully in the transition description text. Transition conditions are listed in priority order and are not required to be mutually exclusive. The first transition condition that evaluates to be true shall be taken. Upon entry to a state, all actions to be executed in that state are executed. If a state is re-entered from itself, all actions to be executed in the state are executed again. It is assumed that all actions are executed within a state and that transitions from state to state are instantaneous. 6 2. Physical Interface 2.1. TSOP-48 and WSOP-48 Pin Assignments Figure 2 defines the pin assignments for devices using 48-pin TSOP or 48-pin WSOP packaging for 8-bit data access. Figure 3 defines the pin assignments for devices using 48-pin TSOP or 48-pin WSOP packaging for 16-bit data access. The package with 16-bit data access does not support the source synchronous data interface. The physical dimensions of the TSOP package is defined in the JEDEC document MO-142 variation DD. The physical dimensions of the WSOP package is defined in the JEDEC document MO-259. 7 Ssync Async R R R R/B3_n R/B2_n R/B1_n R/B0_n W/R_n CE0_n CE1_n R Vcc Vss CE2_n CE3_n CLE ALE CLK WP_n VSP3 R R R R R R R R/B3_n R/B2_n R/B1_n R/B0_n RE_n CE0_n CE1_n R Vcc Vss CE2_n CE3_n CLE ALE WE_n WP_n VSP3 R R R R Async Ssync 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 48-pin TSOP and 48-pin WSOP Figure 2 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 VssQ R R R IO7 IO6 IO5 IO4 R VccQ VSP1 Vcc Vss VSP2* VccQ R IO3 IO2 IO1 IO0 R R R VssQ VssQ R R R DQ7 DQ6 DQ5 DQ4 R VccQ VSP1 Vcc Vss DQS VccQ R DQ3 DQ2 DQ1 DQ0 R R R VssQ 48-pin TSOP/WSOP pinout for 8-bit data access NOTE: For a source synchronous capable part, pin 35 is not used when configured in the asynchronous data interface. Specifically, VSP2 is present for asynchronous only parts. 8 R R R R/B3_n R/B2_n R/B1_n R/B0_n RE_n CE0_n CE1_n R Vcc Vss CE2_n CE3_n CLE ALE WE_n WP_n VSP3 R R R R 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 48-pin TSOP and 48-pin WSOP Figure 3 48-pin TSOP/WSOP pinout for 16-bit data access 9 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 VssQ IO15 IO14 IO13 IO7 IO6 IO5 IO4 IO12 VccQ VSP1 Vcc Vss VSP2 VccQ IO11 IO3 IO2 IO1 IO0 IO10 IO9 IO8 VssQ 2.2. LGA-52 Pad Assignments Figure 4 defines the pad assignments for devices using 52-pad LGA packaging with 8-bit data access. An option is specified for two independent 8-bit data buses. Figure 5 defines the pad assignments for devices using 52-pad LGA packaging with 16-bit data access. The physical dimensions of the package are 12mmx17mm or 14mmx18mm. Figure 6 defines the pad spacing requirements for the 52-pad LGA package for both package dimensions. These LGA packages do not support the source synchronous data interface. A 8 7 R/B1_ 1_n B C D E VDDi R/B1_ 0_n 6 H J K L R RE_0 _n M N R R/B0_ 1_n RE_1 _n Vcc 5 F G IO7 _1 IO6 _1 IO7 _0 Vss R VssQ IO5 _1 IO5 _0 VccQ Vcc CE0_ 0_n CE0_ 1_n R/B0_ 0_n WP_1 _n IO6 _0 IO4 _0 IO4 _1 CLE _0 CLE _1 WE_0 _n IO0 _0 IO2 _0 Vss IO3 _1 4 3 2 1 0 ALE _1 Vss CE1_ 0_n CE1_ 1_n OA ALE _0 R OB Figure 4 WP_0 _n WE_1 _n IO1 _0 IO0 _1 R OC IO3 _0 IO1 _1 VssQ IO2 _1 R R R OD OE OF LGA pinout for 8-bit data access 10 VccQ A 8 R/B3 _n B C D E VDDi 6 RE _n Vcc 5 H J K L R R/B2 _n 7 F G M N R R/B1 _n R IO15 Vss IO14 IO7 R VssQ IO13 IO5 VccQ Vcc CE0 _n CE1 _n R/B0 _n R IO6 IO4 IO12 CLE R WE _n IO0 IO2 Vss IO11 4 3 2 Vss 1 0 CE2 _n CE3 _n OA WP _n R ALE R OB Figure 5 R IO1 IO8 R OC IO3 IO9 VssQ IO10 VccQ R R R OD OE OF LGA pinout for 16-bit data access 11 14.00 +/- 0.15 12.00 +/- 0.10 10.00 2.00 1.00 1.00 1 mm pad diameter 1.30 OA A B G 1.00 H J 18.00 +/- 0.15 F 17.00 +/- 0.10 OC E 2.50 2.50 12.00 D 1.00 OB C OD K L OE M N OF 8 7 6 5 4 3 2 1 0 0.7 mm pad diameter Figure 6 LGA-52 pad spacing requirements (bottom view, dimensions in millimeters) 2.3. BGA-63 Ball Assignments Figure 7 defines the ball assignments for devices using 63-ball BGA packaging with 8-bit data access for the asynchronous data interface. Figure 8 defines the ball assignments for devices using 63-ball BGA packaging with 8-bit data access for the source synchronous data interface. Figure 9 defines the ball assignments for devices using 63-ball BGA packaging with 16-bit data access for the asynchronous data interface. The 63-ball BGA package with 16-bit data access does not support the source synchronous data interface. Figure 10 defines the ball spacing requirements for the 63-ball BGA package. The solder ball diameter is 0.45 mm post reflow. 12 1 2 A R R B R 3 4 5 6 7 8 C WP_n ALE VSS CE0_n WE_n R/B0_n D VCC RE_n CLE CE1_n CE2_n R/B1_n E R R R R CE3_n R/B2_n F R R R R VSS R/B3_n G VSP3 VCC VSP1 R R VSP2 H R IO0 R R R VCCQ J R IO1 R VCCQ IO5 IO7 K VSSQ IO2 IO3 IO4 IO6 VSSQ 9 10 R R R R L R R R R M R R R R Figure 7 BGA-63 ball assignments for 8-bit data access, asynchronous only data interface 13 Note that WE_n is located at ball H7 when a source synchronous capable part is used in asynchronous mode. 1 2 A R R B R 3 4 5 6 7 8 C WP_n ALE VSS CE0_n R R/B0_n D VCC W/R_n CLE CE1_n CE2_n R/B1_n E R R R R CE3_n R/B2_n F R R VREFQ R VSS R/B3_n G VSP3 VCC VSP1 R R VSP2 H R DQ0 DQS_c CLK_c CLK_t VCCQ J R DQ1 DQS_t VCCQ DQ5 DQ7 K VSSQ DQ2 DQ3 DQ4 DQ6 VSSQ 9 10 R R R R L R R R R M R R R R Figure 8 BGA-63 ball assignments for 8-bit data access, source synchronous data interface 14 1 2 A R R B R 3 4 5 6 7 8 C WP_n ALE VSS CE0_n WE_n R/B0_n D VCC RE_n CLE CE1_n CE2_n R/B1_n E R R R R CE3_n R/B2_n F R R R R VSS R/B3_n G VSP3 VCC VSP1 IO13 IO15 VSP2 H IO8 IO0 IO10 IO12 IO14 VCCQ J IO9 IO1 IO11 VCCQ IO5 IO7 K VSSQ IO2 IO3 IO4 IO6 VSSQ 9 10 R R R R L R R R R M R R R R Figure 9 BGA-63 ball assignments for 16-bit, asynchronous only data access 15 0.80 TYP A10 A1 0.80 TYP 0.45 mm ball diameter post reflow 8.80 4.40 3.60 7.20 Figure 10 BGA-63 ball spacing requirements (top view, dimensions in millimeters) 2.4. BGA-100 Ball Assignments Figure 11 defines the ball assignments for devices using 100-ball BGA packaging with dual 8-bit data access for the asynchronous data interface. Figure 12 defines the ball assignments for devices using 100-ball BGA packaging with dual 8-bit data access for the source synchronous data interface. Figure 13 defines the ball spacing requirements for the 100-ball BGA package. The solder ball diameter is 0.45 mm post reflow. The 100-ball BGA has two package sizes: 12mm x 18mm and 14mm x 18mm. 16 1 2 A R R B R 3 4 5 6 7 8 9 10 R R R C D R RFT VSP3_1 WP_1_n VSP2_1 VSP1_1 RFT R E R RFT VSP3_0 WP_0_n VSP2_0 VSP1_0 RFT VDDi F VCC VCC VCC VCC VCC VCC VCC VCC G VSS VSS VSS VSS VSS VSS VSS VSS H VSSQ VCCQ R R R/B0_1_n R/B1_1_n VCCQ VSSQ J IO0_1 IO2_1 ALE_1 CE1_1_n R/B0_0_n R/B1_0_n IO5_1 IO7_1 K IO0_0 IO2_0 ALE_0 CE1_0_n CE0_1_n CE0_0_n IO5_0 IO7_0 L VCCQ VSSQ VCCQ CLE_1 RE_1_n VCCQ VSSQ VCCQ M IO1_1 IO3_1 VSSQ CLE_0 RE_0_n VSSQ IO4_1 IO6_1 N IO1_0 IO3_0 NC NC NC WE_1_n IO4_0 IO6_0 P VSSQ VCCQ NC NC NC WE_0_n VCCQ VSSQ R T R U R Figure 11 R R R R BGA-100 ball assignments for dual 8-bit data access, asynchronous data interface 17 1 2 A R R B R 3 4 5 6 7 8 9 10 R R R C D R RFT VSP3_1 WP_1_n VSP2_1 VSP1_1 RFT R E R RFT VSP3_0 WP_0_n VSP2_0 VSP1_0 RFT VDDi F VCC VCC VCC VCC VCC VCC VCC VCC G VSS VSS VSS VSS VSS VSS VSS VSS H VSSQ VCCQ VREFQ_1 VREFQ_0 R/B0_1_n R/B1_1_n VCCQ VSSQ J DQ0_1 DQ2_1 ALE_1 CE1_1_n R/B0_0_n R/B1_0_n DQ5_1 DQ7_1 K DQ0_0 DQ2_0 ALE_0 CE1_0_n CE0_1_n CE0_0_n DQ5_0 DQ7_0 L VCCQ VSSQ VCCQ CLE_1 W/R_1_n VCCQ VSSQ VCCQ M DQ1_1 DQ3_1 VSSQ CLE_0 W/R_0_n VSSQ DQ4_1 DQ6_1 N DQ1_0 DQ3_0 DQS_1_c DQS_1_t CLK_1_c CLK_1_t DQ4_0 DQ6_0 P VSSQ VCCQ DQS_0_c DQS_0_t CLK_0_c CLK_0_t VCCQ VSSQ R T R U R Figure 12 R R R BGA-100 ball assignments for dual 8-bit data access, source synchronous data interface 18 R 0.45 mm ball diameter post reflow 14mm wide package 14 A1 A10 9 8 7 16 18 5 1 TYP 1 TYP 4.5 9 12 Figure 13 12mm wide package BGA-100 ball spacing requirements (top view, dimensions in millimeters) 19 2.5. Signal Descriptions Table 2 provides the signal descriptions. Where the pin function has an inverted logic sense, that is, the function is true or invoked for a low signal, the overbar, trailing slash (\), #, or _n symbol is appended to the symbol. The _t and _c symbols represent true and complementary logic states for differential pairs. Signal Name Input / Output R/B_x_n O RE_x_n I W/R_x_n I CE_x_n I Vcc I VccQ I Vss I VssQ I VREFQ_x I VDDi na CLE_x I ALE_x I Description Ready/Busy The Ready/Busy signal indicates the target status. When low, the signal indicates that one or more LUN operations are in progress. This signal is an open drain output and requires an external pull-up. See section 2.16 for requirements. Read Enable The Read Enable signal enables serial data output. This signal shares the same pin as W/R_x_n in the source synchronous data interface. Write/Read Direction The Write/Read Direction signal indicates the owner of the DQ bus and DQS signal in the source synchronous data interface. This signal shares the same pin as RE_x_n in the asynchronous data interface. Chip Enable The Chip Enable signal selects the target. When Chip Enable is high and the target is in the ready state, the target goes into a low-power standby state. When Chip Enable is low, the target is selected. See section 2.6 for additional requirements. Power The Vcc signal is the power supply to the device. I/O Power The VccQ signal is the power supply for input and/or output signals. Refer to section 2.8.1. Ground The Vss signal is the power supply ground. I/O Ground The VssQ signal is the ground for input and/or output signals. Refer to section 2.8.1. Voltage Reference This signal is reserved for future use. ASIC Voltage Control This signal is used to assist in stabilizing the internal power supply to a NAND controller ASIC (e.g. EZ NAND) by connecting to an external capacitor. Command Latch Enable The Command Latch Enable signal is one of the signals used by the host to indicate the type of bus cycle (command, address, data). Refer to section 4.1.2. Address Latch Enable The Address Latch Enable signal is one of the signals used by the host to indicate the type of bus cycle (command, address, data). Refer to section 4.1.2. 20 Signal Name Input / Output WE_x_n I CLK_x_t I CLK_x_c I WP_x_n I IO0_0 – IO7_0 (DQ0_0 – DQ7_0) DQS_x_t I/O DQS_x_c I/O IO8 – IO15 I/O IO0_1 – IO7_1 (DQ0_1 – DQ7_1) I/O VSP_x R RFT I/O Description Write Enable The Write Enable signal controls the latching of input data in the asynchronous data interface. Data, commands, and addresses are latched on the rising edge of WE_x_n. This signal shares the same pin as CLK_x in the source synchronous data interface. Clock The Clock signal is used as the clock in the source synchronous data interface. This signal shares the same pin as WE_x_n in the asynchronous data interface. Clock Complement This signal is reserved for future use. Write Protect The Write Protect signal disables Flash array program and erase operations. See section 2.17 for requirements. I/O Port 0, bits 0-7 The I/O port is an 8-bit wide bidirectional port for transferring address, command, and data to and from the device. Also known as DQ0_0 – DQ7_0 for the source synchronous data interface. Data Strobe The data strobe signal that indicates the data valid window for the source synchronous data interface. Data Strobe Complement This signal is reserved for future use. I/O Port 0, bits 8-15 These signals are used in a 16-bit wide target configuration. The signals are the upper 8 bits for the 16-bit wide bidirectional port used to transfer data to and from the device. I/O Port 1, bits 0-7 The I/O port is an 8-bit wide bidirectional port for transferring address, command, and data to and from the device. These pins may be used as an additional 8-bit wide bidirectional port for devices that support two independent data buses. Also known as DQ0_1 – DQ7_1 for the source synchronous data interface. Vendor Specific The function of these signals is defined and specified by the NAND vendor. Devices shall have an internal pull-up or pull-down resistor on these signals to yield ONFI compliant behavior when a signal is not connected by the host. Any VSP signal not used by the NAND vendor shall not be connected internal to the device. Reserved These pins shall not be connected by the host. Reserved for Test These pins shall not be connected by the host. Table 2 Signal descriptions Table 3 provides the signal mapping to pin/pad/ball for each package type listed within the ONFI specification. These signal mappings are required if the packages listed in this specification are implemented. The “Async Only” signal mappings apply to packages where the device is not source synchronous capable. When the device is source synchronous capable, the “Src Sync” signal mappings shall be used. If a signal is marked as “na” then the corresponding package 21 does not implement that signal. Any signal that does not have an associated number is implicitly numbered “0”. For example, WP_n is equivalent to WP0_n. Devices may be implemented with other package types and be ONFI compliant if all other ONFI requirements within this specification are satisfied. 22 TSOP / WSOP Async only x8 TSOP / WSOP Src Sync x8 TSOP / WSOP Async only x16 LGA Async only x8 LGA Async only x16 BGA-63 Async only x8 BGA-63 Src Sync x8 BGA-63 Async only x16 BGA-100 Async only x8 BGA-100 Src Sync x8 na na na na J6 H6 J7 H7 M6 L6 na na na na na na K7 K6 K5 J5 F2 F3 F4 F5 F6 F7 F8 F9 H3 H8 L2 L4 L7 L9 P3 na na na na J6 H6 J7 H7 M6 L6 M6 L6 na na na na K7 K6 K5 J5 F2 F3 F4 F5 F6 F7 F8 F9 H3 H8 L2 L4 L7 L9 P3 Signal Name M/O/R R/B0_n R/B1_n R/B2_n R/B3_n R/B0_0_n R/B0_1_n R/B1_0_n R/B1_1_n RE_0_n RE_1_n W/R_0_n W/R_1_n CE0_n CE1_n CE2_n CE3_n CE0_0_n CE0_1_n CE1_0_n CE1_1_n Vcc M O O O M O O O M O M O M O O O M O O O M 7 6 5 4 na na na na 8 na na na 9 10 14 15 na na na na 12 37 7 6 5 4 na na na na 8 na 8 na 9 10 14 15 na na na na 12 37 7 6 5 4 na na na na 8 na na na 9 10 14 15 na na na na 12 37 na na na na E5 E7 A7 OA8 C7 D6 na na na na na na A5 C5 A1 OA0 B6 M6 E5 E7 A7 OA8 na na na na C7 na na na A5 C5 A1 OA0 na na na na B6 M6 C8 D8 E8 F8 na na na na D4 na na na C6 D6 D7 E7 na na na na D3 G4 C8 D8 E8 F8 na na na na D4 na D4 na C6 D6 D7 E7 na na na na D3 G4 C8 D8 E8 F8 na na na na D4 na na na C6 D6 D7 E7 na na na na D3 G4 VccQ M 34 39 34 39 34 39 N1 N7 N1 N7 H8 J6 H8 J6 H8 J6 23 TSOP / WSOP Async only x8 TSOP / WSOP Src Sync x8 TSOP / WSOP Async only x16 LGA Async only x8 LGA Async only x16 BGA-63 Async only x8 BGA-63 Src Sync x8 BGA-63 Async only x16 13 36 B2 F6 L3 B2 F6 L3 C5 F7 C5 F7 C5 F7 25 48 25 48 M2 OE8 M2 OE8 K8 K3 K8 K3 K8 K3 na na na 16 na 17 na 18 na 18 na na na 19 na na na na 16 na 17 na 18 na na na na na 19 na na na OB8 A3 C3 C1 D2 E3 E1 na na na na F2 G5 na na OB8 A3 na C1 na E3 na na na na na F2 na na na na D5 na C4 na C7 na na na na na C3 na F5 na na D5 na C4 na H7 na H7 na H6 na C3 na na na na D5 na C4 na C7 na na na na na C3 na Signal Name M/O/R Vss M 13 36 13 36 VssQ M 25 48 VREFQ_0 VREFQ_1 VDDi CLE_0 CLE_1 ALE_0 ALE_1 WE_0_n WE_1_n CLK_0_t CLK_1_t CLK_0_c CLK_1_c WP_0_n WP_1_n R R O M O M O M O M O R R M O na na na 16 na 17 na 18 na na na na na 19 na 24 BGA-100 Async only x8 BGA-100 Src Sync x8 P8 G2 G3 G4 G5 G6 G7 G8 G9 H2 H9 L3 L8 M4 M7 P2 P9 na na E9 M5 L5 K4 J4 P7 N7 na na na na E5 D5 P8 G2 G3 G4 G5 G6 G7 G8 G9 H2 H9 L3 L8 M4 M7 P2 P9 H5 H4 E9 M5 L5 K4 J4 P7 N7 P7 N7 P6 N6 E5 D5 Signal Name M/O/R TSOP / WSOP Async only x8 IO0_0 / DQ0_0 IO1_0 / DQ1_0 IO2_0 / DQ2_0 IO3_0 / DQ3_0 IO4_0 / DQ4_0 IO5_0 / DQ5_0 IO6_0 / DQ6_0 IO7_0 / DQ7_0 IO8 IO9 IO10 IO11 IO12 IO13 IO14 IO15 IO0_1 / DQ0_1 IO1_1 / DQ1_1 IO2_1 / DQ2_1 IO3_1 / DQ3_1 IO4_1 / DQ4_1 IO5_1 / DQ5_1 IO6_1 / DQ6_1 IO7_1 / DQ7_1 DQS_0_t DQS_1_t DQS_0_c DQS_1_c VSP0_0 VSP1_0 VSP2_0 VSP0_1 VSP1_1 VSP2_1 M M M M M M M M M M M M M M M M O O O O O O O O M O R R O O O O O O 29 30 31 32 41 42 43 44 na na na na na na na na na na na na na na na na na na na na 38 35 20 na na na TSOP / WSOP Src Sync x8 TSOP / WSOP Async only x16 LGA Async only x8 LGA Async only x16 BGA-63 Async only x8 BGA-63 Src Sync x8 BGA-63 Async only x16 BGA-100 Async only x8 BGA-100 Src Sync x8 29 30 31 32 41 42 43 44 na na na na na na na na na na na na na na na na 35 na na na 38 na 20 na na na 29 30 31 32 41 42 43 44 26 27 28 33 40 45 46 47 na na na na na na na na na na na na 38 35 20 na na na G3 H2 J3 K2 L5 K6 J5 H6 na na na na na na na na G1 J1 L1 N3 N5 L7 J7 G7 na na na na na na na na na na G3 H2 J3 K2 L5 K6 J5 H6 G1 J1 L1 N3 N5 L7 J7 G7 na na na na na na na na na na na na na na na na na na H4 J4 K4 K5 K6 J7 K7 J8 na na na na na na na na na na na na na na na na na na na na G5 G8 G3 na na na H4 J4 K4 K5 K6 J7 K7 J8 na na na na na na na na na na na na na na na na J5 na H5 na G5 G8 G3 na na na H4 J4 K4 K5 K6 J7 K7 J8 H3 J3 H5 J5 H6 G6 H7 G7 na na na na na na na na na na na na G5 G8 G3 na na na K2 N2 K3 N3 N8 K8 N9 K9 na na na na na na na na J2 M2 J3 M3 M8 J8 M9 J9 na na na na E7 E6 E4 D7 D6 D4 K2 N2 K3 N3 N8 K8 N9 K9 na na na na na na na na J2 M2 J3 M3 M8 J8 M9 J9 P5 N5 P4 N4 E7 E6 E4 D7 D6 D4 Table 3 25 Signal mappings: TSOP, LGA, BGA packages 2.6. CE_n Signal Requirements If one or more LUNs are active and the host sets CE_n to one, then those operations continue executing until completion at which point the target enters standby. After the CE_n signal is transitioned to one, the host may drive a different CE_n signal to zero and begin operations on another target. Note that if using a dual x8 package (e.g. BGA-100), then operations may execute in parallel on two different CE_n’s if they are connected to different 8-bit data buses. When SR[6] for a particular LUN is cleared to zero and the CE_n signal for the corresponding target is cleared to zero, the host may only issue the Reset, Synchronous Reset, Read Status, or Read Status Enhanced commands to that LUN. 2.6.1. Source Synchronous Data Interface Requirements When using the source synchronous data interface, the following requirements shall be met if the device does not support CLK being stopped during data input: 1. CLK shall only stop or start when CE_n is high. When using the source synchronous data interface, the following requirements shall be met if the device supports CLK being stopped during data input: 1. CLK shall only stop or start when either: a. CE_n is high, or b. CE_n is low and the bus state is data input When using the source synchronous data interface, the following requirements shall always be met: 1. CLK shall only change frequency when CE_n is high. 2. When CE_n is low, CLK shall maintain the same frequency. 3. CE_n shall only transition from one to zero when the CLK is stable and has a valid period based on the timing mode selected. 4. The interface shall be in an idle state (see section 4.1.2) when CE_n changes value. CE_n shall only transition when the following are true: a. ALE and CLE are both cleared to zero, and b. There is no data transfer on the DQ/DQS signals during the current clock period. 2.7. Absolute Maximum DC Ratings Stresses greater than those listed in Table 4 may cause permanent damage to the device. This is a stress rating only. Operation beyond the recommended operating conditions specified in Table 5 and the DC and operating characteristics listed in Table 9 and Table 10 is not recommended. Except as defined in section 2.9, extended exposure beyond these conditions may affect device reliability. Table 4 defines the voltage on any pin relative to Vss and/or VssQ for devices based on their Vcc and VccQ typical voltages. 26 Parameter Symbol Rating Vcc = 3.3V and VccQ = 3.3V nominal Vcc Supply Voltage VCC -0.6 to +4.6 Voltage Input VIN -0.6 to +4.6 VccQ Supply Voltage VCCQ -0.6 to +4.6 Vcc = 3.3V and VccQ = 1.8V nominal Vcc Supply Voltage VCC -0.6 to +4.6 Voltage Input VIN -0.2 to +2.4 VccQ Supply Voltage VCCQ -0.2 to +2.4 Vcc = 1.8V and VccQ = 1.8V nominal Vcc Supply Voltage VCC -0.2 to +2.4 Voltage Input VIN -0.2 to +2.4 VccQ Supply Voltage VCCQ -0.2 to +2.4 Table 4 Units V V V Absolute maximum DC ratings 2.8. Recommended DC Operating Conditions Parameter Supply voltage for 3.3V devices Supply voltage for 1.8V devices Supply voltage for 3.3V I/O signaling Supply voltage for 1.8V I/O signaling Ground voltage supply Ground voltage supply for I/O signaling Table 5 Symbol Min Typ Max Units VCC 2.7 3.3 3.6 V VCC 1.7 1.8 1.95 V 2.7 3.3 3.6 V 1.7 1.8 1.95 V 0 0 0 V 0 0 0 V VCCQ (VCCQH) VCCQ (VCCQL) VSS VSSQ Recommended DC operating conditions 2.8.1. I/O Power (VccQ) and I/O Ground (VssQ) VccQ and Vcc may be distinct and unique voltages. VccQ shall be less than or equal to Vcc, including during power-on ramp. The device shall support one of the following VccQ/Vcc combinations: Vcc = 3.3V, VccQ = 3.3V Vcc = 3.3V, VccQ = 1.8V Vcc = 1.8V, VccQ = 1.8V All parameters, timing modes, and other characteristics are relative to the supported voltage combination. If a device has the same Vcc and VccQ voltage levels, then VccQ and VssQ are not required to be connected internal to the device. Specifically, the device may use Vcc and Vss exclusively as the I/O and core voltage supply. 27 2.9. AC Overshoot/Undershoot Requirements The device may have AC overshoot or undershoot from VccQ and VssQ levels. Table 6 defines the maximum values that the AC overshoot or undershoot may attain. These values apply for both 3.3V and 1.8V VccQ levels. Maximum Value Parameter Peak amplitude allowed for overshoot area Peak amplitude allowed for undershoot area Maximum Overshoot area above VccQ Maximum Undershoot area below VssQ Table 6 > 133 MT/s and <= 166 MT/s > 166 MT/s and <= 200 MT/s Unit <= 100 MT/s > 100 MT/s and <= 133 MT/s 1 1 1 1 V 1 1 1 1 V 3 2.25 1.8 1.5 V-ns 3 2.25 1.8 1.5 V-ns AC Overshoot/Undershoot Maximum Values Figure 14 displays pictorially the parameters described in Table 6. Maximum Amplitude Volts (V) Overshoot Area VccQ VssQ Maximum Amplitude Figure 14 2.10. Undershoot Area Overshoot/Undershoot Diagram DC and Operating Characteristics All operating current ratings in this section are specified per active logical unit (LUN). A LUN is active when there is a command outstanding to it. All other current ratings in this section are specified per LUN (regardless of whether it is active). For high performance applications it may be desirable to draw increased current for ICC1-ICC4. For these applications, the device may draw up to 100 mA per active LUN in both 3.3V and 1.8V devices. Increased current may be used to improve sustained write performance. 28 All ICC measurements are measured with each Vcc pin decoupled with a 0.1 µF capacitor. The ICC definition assumes outputs change between one and zero every other data cycle (once per CLK period, every other DQS transition) for data signals. The test conditions and measurement methodology for the ICC values is defined in Appendix D. Parameter Symbol Array read current Array program current Array erase current ICC1 ICC2 ICC3 I/O burst read current 4 ICC4R I/O burst write current Bus idle current Test Conditions Refer to Appendix D ICC4W ICC5 Standby current, CMOS ISB Staggered power-up current 1 IST CE_n=VccQ0.2V, WP_n=0V/VccQ CE_n=VccQ0.2V tRise = 1 ms cLine = 0.1 µF Min Typ Max Units - - 50 50 50 mA mA mA - - 50 mA - - 50 mA - - 10 mA - - 50 µA - - 10 mA NOTE: 1. Refer to Appendix C for an exception to the IST current requirement. 2. ICC1, ICC2, and ICC3 as listed in this table are active current values. For details on how to calculate the active current from the measured values, refer to Appendix D. 3. During cache operations, increased ICC current is allowed while data is being transferred on the bus and an array operation is ongoing. For a cached read this value is ICC1 + ICC4R; for a cached write this value is ICC2(active) + ICC4W. 4. For ICC4R the test conditions in Appendix D specify IOUT = 0 mA and requires static outputs with no output switching. When outputs are not static, additional VccQ current will be drawn that is highly dependent on system configuration. IccQ may be calculated for each output pin assuming 50% data switching as (IccQ = 0.5 * CL * VccQ * frequency), where CL is the capacitive load. Table 7 DC and Operating Conditions for raw NAND, measured on Vcc rail 29 Parameter Symbol Array read current Array program current Array erase current I/O burst read current I/O burst write current Bus idle current Standby current, CMOS Staggered power-up current Min Typ Max Per LUN Controller Per LUN Per LUN Controller Per LUN Per LUN Controller Per LUN - - 85 200 15 85 75 15 85 75 15 LUN - - 50 Controller - - 50 ICC4W LUN - - 50 ICCQ4W ICC5 ICCQ5 ICCQ5 ISB Controller - - 50 Per LUN Controller Per LUN Per LUN - - 45 10 15 50 Controller - - 1000 Per LUN - - 10 Controller - - 20 ICC1 ICCQ1 ICCQ1 ICC2 ICCQ2 ICCQ2 ICC3 ICCQ3 ICCQ3 4 ICC4R 4 ICCQ4R ISBQ 1 IST ISTQ Test Conditions Refer to Appendix D CE_n=VccQ0.2V, WP_n=0V/VccQ CE_n=VccQ0.2V tRise = 1 ms cLine = 0.1 µF Units mA mA mA mA mA mA µA mA NOTE: 1. Refer to Appendix C for an exception to the IST current requirement. 2. ICC1, ICC2, and ICC3 as listed in this table are active current values. For details on how to calculate the active current from the measured values, refer to Appendix D. 3. During cache operations, increased ICC current is allowed while data is being transferred on the bus and an array operation is ongoing. For a cached read this value is ICC1 + ICC4 R on Vcc and ICCQ1 on VccQ; for a cached write this value is ICC2(active) + ICC4W on Vcc, and ICCQ2 on VccQ. 4. For ICC4R the test conditions in Appendix D specify IOUT = 0 mA and requires static outputs with no output switching. When outputs are not static, additional VccQ current will be drawn that is highly dependent on system configuration. IccQ may be calculated for each output pin assuming 50% data switching as (IccQ = 0.5 * C L * VccQ * frequency), where CL is the capacitive load. Table 8 DC and Operating Conditions for EZ NAND, measured on Vcc or VccQ rail The maximum leakage current requirements (ILI and ILO) in Table 9 and Table 10 are tested across the entire allowed VccQ range, specified in Table 5. DC signal specifications apply to the following signals and only when using the source synchronous data interface: CLK, DQ[7:0], DQS, ALE, CLE, and W/R_n. For all signals in asynchronous and all other signals in source synchronous, the AC signal specification shall be met. For signals where DC signal specifications apply, the transition times are measured between VIL (DC) and VIH (AC) for rising input signals and between VIH (DC) and VIL (AC) for falling input signals. The receiver will effectively switch as a result of the signal crossing the AC input level and remain in that state as long as the signal does not ring back above (below) the DC input LOW (HIGH) level. The parameters in Table 9 and Table 10 apply to power-on default values in the device. If I/O drive strength settings or other device settings are changed, these values may be modified. The output characteristics for a device that supports driver strength settings (as indicated in the parameter page) are specified in the impedance tables (Table 28 and Table 29). 30 Parameter Symbol Test Conditions Min Typ Max Units Standby current, CE_n=VccQ-0.2V, ISBQ 25 µA CMOS WP_n=0V/VccQ Input leakage current ILI VIN=0V to VccQ +-10 µA Output leakage ILO VOUT=0V to VccQ +-10 µA current DC Input high VIH (DC) VccQ * 0.7 VccQ + 0.3 V voltage AC Input high VIH (AC) VccQ * 0.8 VccQ + 0.3 V voltage DC Input low voltage VIL (DC) -0.3 VccQ * 0.3 V AC Input low voltage VIL (AC) -0.3 VccQ * 0.2 V VccQ * 1 VOH IOH=-400 µA V Output high voltage 0.67 1 VOL IOL=2.1 mA 0.4 V Output low voltage Output low current IOL(R/B_ VOL=0.4 V 8 10 mA (R/B_n) n) NOTE: 1. VOH and VOL defined in this table shall only apply to devices in asynchronous mode that do not support driver strength settings. If driver strength settings are supported then Table 28 shall be used to derive the output driver impedance values. Table 9 Parameter DC and Operating Conditions for VccQ of 3.3V, measured on VccQ rail Symbol Test Conditions Min Typ Max Units Standby current, CE_n=VccQ-0.2V, ISBQ 25 µA CMOS WP_n=0V/VccQ Input leakage current ILI VIN=0V to VccQ +-10 µA Output leakage ILO VOUT=0V to VccQ +-10 µA current DC Input high VIH (DC) VccQ * 0.7 VccQ+0.3 V voltage AC Input high VIH (AC) VccQ * 0.8 VccQ+0.3 V voltage DC Input low voltage VIL (DC) -0.3 VccQ * 0.3 V AC Input low voltage VIL (AC) -0.3 VccQ * 0.2 V 1 VOH IOH=-100 µA VccQ – 0.1 V Output high voltage 1 VOL IOL=100 µA 0.1 V Output low voltage Output low current IOL(R/B_n) VOL=0.2 V 3 4 mA (R/B_n) NOTE: 1. VOH and VOL defined in this table shall only apply to devices in asynchronous mode that do not support driver strength settings. If driver strength settings are supported then Table 29 shall be used to derive the output driver impedance values. Table 10 2.11. DC and Operating Conditions for VccQ of 1.8V, measured on VccQ rail Calculating Pin Capacitance To calculate the pin capacitance for all loads on the I/O bus, the host should utilize the reported pin capacitance per target in Read Parameter Page (refer to section 5.7). The maximum 31 capacitance may be used, or the typical capacitance if provided by the device may be used. The algorithm to use is: PinCapacitance = 0; for (target = 0; target < TotalTargets; target++) PinCapacitance += GetCapacitanceFromRPP(target); This methodology will calculate an accurate maximum or typical pin capacitance, respectively, accounting for all targets present. 2.12. Staggered Power-up Subsystems that support multiple Flash devices may experience power system design issues related to the current load presented during the power-on condition. To limit the current load presented to the host at power-on, all devices shall support power-up in a low-power condition. Until a Reset (FFh) command is received by the target after power-on, the target shall not draw more than IST of current per LUN and ISTQ (if present for devices that support EZ NAND). For example, a target that contains 4 LUNs may draw up to 40 mA of current until a Reset (FFh) command is received after power-on. This value is measured with a nominal rise time (tRise) of 1 millisecond and a line capacitance (cLine) of 0.1 µF. The measurement shall be taken with 1 millisecond averaging intervals and shall begin after Vcc reaches Vcc_min and VccQ reaches VccQ_min. 2.13. Power Cycle Requirements As part of a power cycle, the host shall hold both the Vcc and VccQ voltage levels below 100 mV for a minimum time of 100 ns. If these requirements are not met as part of a power cycle operation, the device may enter an indeterminate state. 2.14. Independent Data Buses There may be two independent 8-bit data buses in some ONFI packages (i.e. the LGA and the 100-ball BGA package). For packages that support either two independent data buses or a single data bus (e.g. LGA-52) then CE0_n and CE2_n shall use the same pins as the first data bus CE_n pins (marked as CE0_0_n and CE1_0_n) and CE1_n and CE3_n shall use the same pins as the second data bus CE_n pins (marked as CE0_1_n and CE1_1_n). Note that CE0_n, CE1_n, CE2_n, and CE3_n may all use the first data bus and the first set of control signals (RE0_n, CLE0_n, ALE0_n, WE0_n, and WP0_n) if the device does not support independent data buses. Implementations may tie the data lines and control signals (RE_n, CLE, ALE, WE_n, WP_n, and DQS) together for the two independent 8-bit data buses externally to the device. 2.15. Bus Width Requirements All targets per device shall use the same data bus width. All targets shall either have an 8-bit bus width or a 16-bit bus width. Note that devices that support the source synchronous interface shall have an 8-bit bus width. When the host supports a 16-bit bus width, only data is transferred at the 16-bit width. All address and command line transfers shall use only the lower 8-bits of the data bus. During 32 command transfers, the host may place any value on the upper 8-bits of the data bus. During address transfers, the host shall set the upper 8-bits of the data bus to 00h. 2.16. 2.16.1. Ready/Busy (R/B_n) Requirements Power-On Requirements Once VCC and VccQ reach the VCC minimum and VccQ minimum values, respectively, listed in Table 5 and power is stable, the R/B_n signal shall be valid after RB_valid_Vcc and shall be set to one (Ready) within RB_device_ready, as listed in Table 11. R/B_n is undefined until 50 µs has elapsed after VCC has started to ramp. The R/B_n signal is not valid until both of these conditions are met. Parameter RB_valid_Vcc Raw NAND 10 µs EZ NAND 250 µs RB_device_ready 1 ms 2 ms Table 11 R/B_n Power-on Requirements During power-on, VccQ shall be less than or equal to Vcc at all times. Figure 15 shows VccQ ramping after Vcc, however, they may ramp at the same time. 50 µs (min) VccQ RB_valid_Vcc Vcc >= 0 µs (min) R/B_n Vcc ramp starts RB_device_ready VccQ = VccQ_min Reset (FFh) is issued Vcc = Vcc_min Figure 15 Undefined R/B_n Power-On Behavior Ready/Busy is implemented as an open drain circuit, thus a pull-up resistor shall be used for termination. The combination of the pull-up resistor and the capacitive loading of the R/B_n circuit determines the rise time of R/B_n. 33 2.16.2. R/B_n and SR[6] Relationship R/B_n shall reflect the logical AND of the SR[6] (Status Register bit 6) values for all LUNs on the corresponding target. For example, R/B3_n is the logical AND of the SR[6] values for all LUNs on CE3_n. Thus, R/B_n reflects whether any LUN is busy on a particular target. 2.17. Write Protect When cleared to zero, the WP_n signal disables Flash array program and erase operations. This signal shall only be transitioned while there are no commands executing on the device. After modifying the value of WP_n, the host shall not issue a new command to the device for at least tWW delay time. Figure 16 describes the tWW timing requirement, shown with the start of a Program command. The transition of the WP_n signal is asynchronous and unrelated to any CLK transition in the source synchronous data interface. The bus shall be idle for tWW time after WP_n transitions from zero to one before a new command is issued by the host, including Program. The bus shall be idle for tWW time after WP_n transitions from one to zero before a new command is issued by the host. As defined for Page Program Cycle Type CMD ADDR ADDR ADDR ADDR ADDR 80h C1 C2 R1 R2 R3 tWW DQx SR[6] WP_n Bus shall be idle Figure 16 Write Protect timing requirements, example 34 3. Memory Organization Figure 17 shows an example of a Target memory organization. In this case, there are two logical units where each logical unit has two planes. Page 0 Page 0 Page 0 Page 1 Page 1 Page 1 Page 1 Page P Page P Page P Page P Block 0 Block 1 Block 0 Block 1 Page 0 Page 0 Page 0 Page 0 Page 1 Page 1 Page 1 Page 1 Page P Page P Page P Page P Block 2 Block 3 Block 2 Block 3 Page 0 Page 0 Page 0 Page 0 Page 1 Page 1 Page 1 Page 1 Page P Page P Page P Page P Logical Unit 1 Logical Unit 0 Page 0 Block B Block B+1 Block B Block B+1 Page Register Page Register Page Register Page Register Plane Address 0 Plane Address 1 Plane Address 0 Plane Address 1 Figure 17 Target memory organization A device contains one or more targets. A target is controlled by one CE_n signal. A target is organized into one or more logical units (LUNs). A logical unit (LUN) is the minimum unit that can independently execute commands and report status. Specifically, separate LUNs may operate on arbitrary command sequences in parallel. For example, it is permissible to start a Page Program operation on LUN 0 and then prior to the operation’s completion to start a Read command on LUN 1. See multiple LUN operation restrictions in section 3.1.3. A LUN contains at least one page register and a Flash array. The number of page registers is dependent on the number of multi-plane operations supported for that LUN. The Flash array contains a number of blocks. 35 A block is the smallest erasable unit of data within the Flash array of a LUN. There is no restriction on the number of blocks within the LUN. A block contains a number of pages. A page is the smallest addressable unit for read and program operations. A page consists of a number of bytes or words. The number of user data bytes per page, not including the spare data area, shall be a power of two. The number of pages per block shall be a multiple of 32. Each LUN shall have at least one page register. A page register is used for the temporary storage of data before it is moved to a page within the Flash array or after it is moved from a page within the Flash array. If EZ NAND is supported then a buffer exists in the EZ NAND controller that provides for temporary storage of data that may then be transferred to or from the page register within each LUN. The byte or word location within the page register is referred to as the column. There are two mechanisms to achieve parallelism within this architecture. There may be multiple commands outstanding to different LUNs at the same time. To get further parallelism within a LUN, multi-plane operations may be used to execute additional dependent operations in parallel. 3.1. Addressing There are two address types used: the column address and the row address. The column address is used to access bytes or words within a page, i.e. the column address is the byte/word offset into the page. The least significant bit of the column address shall always be zero in the source synchronous data interface, i.e. an even number of bytes is always transferred. The row address is used to address pages, blocks, and LUNs. When both the column and row addresses are required to be issued, the column address is always issued first in one or more 8-bit address cycles. The row addresses follow in one or more 8-bit address cycles. There are some functions that may require only row addresses, like Block Erase. In this case the column addresses are not issued. For both column and row addresses the first address cycle always contains the least significant address bits and the last address cycle always contains the most significant address bits. If there are bits in the most significant cycles of the column and row addresses that are not used then they are required to be cleared to zero. The row address structure is shown in Figure 18 with the least significant row address bit to the right and the most significant row address bit to the left. MSB LUN Address LSB Block Address Figure 18 Page Address Row Address Layout The number of blocks and number of pages per block is not required to be a power of two. In the case where one of these values is not a power of two, the corresponding address shall be rounded to an integral number of bits such that it addresses a range up to the subsequent power of two value. The host shall not access upper addresses in a range that is shown as not supported. For example, if the number of pages per block is 96, then the page address shall be rounded to 7 bits such that it can address pages in the range of 0 to 127. In this case, the host shall not access pages in the range from 96 to 127 as these pages are not supported. 36 The page address always uses the least significant row address bits. The block address uses the middle row address bits and the LUN address uses the most significant row address bit(s). 3.1.1. Multi-plane Addressing The multi-plane address comprises the lowest order bits of the block address as shown in Figure 19. The following restrictions apply to the multi-plane address when executing a multi-plane command sequence on a particular LUN: The plane address bit(s) shall be distinct from any other multi-plane operation in the multi-plane command sequence. The page address shall be the same as any other multi-plane operations in the multiplane command sequence. MSB LSB LUN Address Block Address Page Address Plane Address bit(s) Figure 19 Plane Address Location 3.1.1.1. Multi-plane Block Address Restrictions The device may indicate multi-plane block address restrictions. The specific cases are: No restriction: All block address bits may be different between two plane addresses. Full restriction: All block address bits (other than the plane address bits) shall be the same between two plane addresses. Lower bit XNOR restriction: If the XNOR of the lowest plane address bits (bit 0) is one between two plane addresses, then there is a full restriction between these two plane addresses. If the XNOR of the lower plane address bits is zero between two plane addresses, then there is no restriction between these two plane addresses. Table 12 illustrates the three types of restrictions for a four plane operation. Restriction Type No restriction Plane Address 0 Block A Plane Address 1 Block B Plane Address 2 Block C Plane Address 3 Block D XNOR restriction Block A Block B Block A+2 Block B+2 Full restriction Block A Block A+1 Block A+2 Block A+3 Table 12 Four plane address restriction Table 13 describes whether there is a lower bit XNOR restriction between two plane addresses A and B, based on their plane address bits for a 4 plane implementation. If there is a lower bit XNOR restriction, then the block addresses (other than the plane address bits) shall be the same between multi-plane addresses A and B. 37 Multi-plane Address bits A 00b Multi-plane Address bits B 01b Lower Bit XNOR 0 XNOR 1 = 0 XNOR Restriction Between A and B No 00b 10b 0 XNOR 0 = 1 Yes 00b 11b 0 XNOR 1 = 0 No 01b 10b 1 XNOR 0 = 0 No 01b 11b 1 XNOR 1 = 1 Yes 10b 11b 0 XNOR 1 = 0 No Table 13 4-way lower bit XNOR restriction 3.1.2. Logical Unit Selection Logical units within one target share a single data bus with the host. The host shall ensure that only one LUN is selected for data output to the host at any particular point in time to avoid bus contention. The host selects a LUN for future data output by issuing a Read Status Enhanced command to that LUN. The Read Status Enhanced command shall deselect the output path for all LUNs that are not addressed by the command. The page register selected for output within the LUN is determined by the previous Read (Cache) commands issued, and is not impacted by Read Status Enhanced. 3.1.3. Multiple LUN Operation Restrictions LUNs are independent entities. A multiple LUN operation is one in which two or more LUNs are simultaneously processing commands. This implies that R/B_n is cleared to zero when the subsequent LUN operation is issued. When a Page Program command (80h) is issued on any LUN that is not preceded by an 11h command, all idle LUNs may clear their page registers if the program page register clear enhancement is not supported or enabled. Thus, the host should not begin a Page Program command on a LUN while a Read Page operation is either ongoing or has completed but the data has not been read from another LUN, as the contents of the page register for the Read operation are lost. A Read Page can be issued to one LUN while a Page Program is ongoing within a second LUN without any restriction. If the program page register clear enhancement is enabled, this restriction does not apply. When issuing a Page Program command (80h), the host should not select another LUN until after all data has been input and a 10h or 15h command has been issued. In the case of multi-plane operations, all data input for all multi-plane addresses should be completed prior to selecting another LUN. When issuing Reads to multiple LUNs, the host shall take steps to avoid issues due to column address corruption. Specifically, if the column addresses in Reads issued to multiple LUNs are different, then the host shall issue a Change Read Column before starting to read out data from a newly selected LUN. If the column addresses are the same, then no Change Read Column is necessary. If a multiple LUN operation has been issued, then the next status command issued shall be Read Status Enhanced. Read Status Enhanced causes LUNs that are not selected to turn off their 38 output buffers. This ensures that only the LUN selected by the Read Status Enhanced command responds to a subsequent data output cycle. After a Read Status Enhanced command has been completed, the Read Status command may be used until the next multiple LUN operation is issued. When the host has issued Read Page commands to multiple LUNs at the same time, the host shall issue Read Status Enhanced before reading data from either LUN. This ensures that only the LUN selected by the Read Status Enhanced command responds to a data output cycle after being put in data output mode with a 00h command, and thus avoiding bus contention. If the host issues a Change Read Column (Enhanced) for any LUN that Read Page commands are outstanding for, then the host shall issue a Change Read Column (Enhanced) prior to transferring data from any subsequent LUN that is part of the multiple LUN read sequence. An example sequence is shown below: 1) Read Page command issued to LUN 0 2) Read Page command issued to LUN 1 3) Read Status Enhanced selects LUN 0 4) Change Read Column (Enhanced) issued to LUN 0 (if needed by host) 5) Data transferred from LUN 0 6) Read Status Enhanced selects LUN 1 7) Change Read Column (Enhanced) issued to LUN 1 (required if action 4 taken) 8) Data transferred from LUN 1 The host may substitute Change Read Column Enhanced for the Read Status Enhanced / Change Read Column sequence if all LUNs are not busy. 3.2. Factory Defect Mapping The Flash array is not presumed to be pristine, and a number of defects may be present that renders some blocks unusable. Block granularity is used for mapping factory defects since those defects may compromise the block erase capability. 3.2.1. Device Requirements If a block is defective and 8-bit data access is used, the manufacturer shall mark the block as defective by setting the first byte in the defect area, as shown in Figure 20, of the first or last page of the defective block to a value of 00h. If a block is defective and 16-bit data access is used, the manufacturer shall mark the block as defective by setting the first word in the defect area of the first or last page of the defective block to a value of 0000h. # of data bytes - 1 Byte 0 1 2 Area Figure 20 # of data bytes + # of spare bytes - 1 # of data bytes … … n/a Defect Area Area marked in factory defect mapping 39 3.2.2. Host Requirements The host shall not erase or program blocks marked as defective by the manufacturer, and any attempt to do so yields indeterminate results. Figure 21 outlines the algorithm to scan for factory mapped defects. This algorithm should be performed by the host to create the initial bad block table prior to performing any erase or programming operations on the target. The initial state of all pages in non-defective blocks is FFh (or FFFFh for 16-bit access) for all page addresses, although some bit errors may be present if they are correctable via the required ECC reported to the host. A defective block is indicated by a byte value equal to 00h for 8-bit access or a word value equal to 0000h for 16-bit access being present at the first byte/word location in the defect area of either the first page or last page of the block. The host shall check the first byte/word of the defect area of both the first and last past page of each block to verify the block is valid prior to any erase or program operations on that block. NOTE: Over the lifetime use of a NAND device, the defect area of defective blocks may encounter read disturbs that cause values to change. The manufacturer defect markings may change value over the lifetime of the device, and are expected to be read by the host and used to create a bad block table during initial use of the part. for (i=0; i<NumLUNs; i++) { for (j=0; j<BlocksPerLUN; j++) { Defective=FALSE; ReadPage(lun=i; block=j; page=0; DestBuff=Buff); if (Buff[PageSize] == 00h) // Value checked for is 0000h for 16-bit access Defective=TRUE; ReadPage(lun=i; block=j; page=PagesPerBlock-1; DestBuff=Buff); if (Buff[PageSize] == 00h) // Value checked for is 0000h for 16-bit access Defective=TRUE; if (Defective) MarkBlockDefective(lun=i; block=j); } } Figure 21 Factory defect scanning algorithm 3.3. Extended ECC Information Reporting The device may report extended ECC information in the extended parameter page. The required ECC correctability is closely related to other device parameters, like the number of valid blocks and the number of program/erase cycles supported. Extended ECC information allows the device to specify multiple valid methods for using the device. 40 Table 14 defines the extended ECC information block. Byte 0 1 2-3 4-5 6-7 Table 14 Definition Number of bits ECC correctability Codeword size Bad blocks maximum per LUN Block endurance Reserved Extended ECC Information Block Definition The definition of each field follows in the subsequent sections. 3.3.1. Byte 0: Number of bits ECC correctability This field indicates the number of bits that the host should be able to correct per codeword. The codeword size is reported in byte 1. With this specified amount of error correction by the host, the target shall achieve the block endurance specified in bytes 4-5. When the specified amount of error correction is applied by the host and the block endurance is followed, then the maximum number of bad blocks specified in bytes 2-3 shall not be exceeded by the device. All used bytes in the page shall be protected by host controller ECC including the spare bytes if the ECC requirement reported in byte 0 has a value greater than zero. When this value is cleared to zero, the target shall return valid data if the ECC Information Block is valid (the Codeword size is non-zero). 3.3.2. Byte 1: Codeword size The number of bits of ECC correctability specified in byte 0 is based on a particular ECC codeword size. The ECC codeword size is specified in this field as a power of two. The minimum value that shall be reported is 512 bytes (a value of 9). If a value of 0 is reported then this ECC Information Block is invalid and should not be used. 3.3.3. Byte 2-3: Bad blocks maximum per LUN This field contains the maximum number of blocks that may be defective at manufacture and over the life of the device per LUN. The maximum rating assumes that the host is following the block endurance requirements and the ECC requirements reported in this extended ECC information block. 3.3.4. Byte 4-5: Block endurance This field indicates the maximum number of program/erase cycles per addressable page/block. This value assumes that the host is using the ECC correctability reported in byte 0. The block endurance is reported in terms of a value and a multiplier according to the following multiplier equation: value x 10 . Byte 4 comprises the value. Byte 5 comprises the multiplier. For example, a block endurance of 75,000 cycles would be reported as a value of 75 and a multiplier 3 of 3 (75 x 10 ). The value field shall be the smallest possible; for example 100,000 shall be 5 reported as a value of 1 and a multiplier of 5 (1 x 10 ). 41 3.4. Discovery and Initialization 3.4.1. CE_n Discovery There may be up to four chip enable (CE_n) signals on a package, one for each separately addressable target. To determine the targets that are connected, the procedure outlined in this section shall be followed for each distinct CE_n signal. CE_n signals shall be used sequentially on the device; CE0_n is always connected and CE_n signals shall be connected in a numerically increasing order. The host shall attempt to enumerate targets connected to all host CE_n signals. The discovery process for a package that supports independent dual data buses includes additional steps to determine which data bus the target is connected to. The LGA and 100-ball BGA package with 8-bit data access are the packages within ONFI that have a dual data bus option. 3.4.1.1. Single Data Bus Discovery The CE_n to test is first pulled low by the host to enable the target if connected, while all other CE_n signals are pulled high. The host shall then issue the Reset (FFh) command to the target. Following the reset, the host should then issue a Read ID command to the target. If the ONFI signature is returned by the Read ID command with address 20h, then the corresponding target is connected. If the ONFI signature is not returned or any step in the process encountered an error/timeout, then the CE_n is not connected and no further use of that CE_n signal shall be done. 3.4.1.2. Dual Data Bus Discovery The CE_n to test is first pulled low by the host to enable the target if connected, while all other CE_n signals are pulled high. The host shall then issue the Reset (FFh) command to the target. Following the reset, the host should then issue a Read ID command with address 20h to the target. If the ONFI signature is returned by the Read ID command, then the corresponding target is connected. If the ONFI signature is not returned (or any step in the process encountered an error/timeout), then the second 8-bit data bus should be probed. The host shall issue the Reset (FFh) command to the target using the second 8-bit data bus. Following the reset, the host should then issue a Read ID command with address 20h to the target on the second 8-bit data bus. If the ONFI signature is returned by the Read ID command, then the corresponding target is connected and is using the second 8-bit data bus. After discovering that the target is using the second 8-bit data bus, all subsequent commands to that target shall use the second 8-bit data bus including Read Parameter Page. If after this point a valid ONFI signature is not discovered or further errors were encountered, then the CE_n is not connected and no further use of that CE_n signal shall be done. 3.4.2. Target Initialization To initialize a discovered target, the following steps shall be taken. The initialization process should be followed for each connected CE_n signal, including performing the Read Parameter Page (ECh) command for each target. Each chip enable corresponds to a unique target with its own independent properties that the host shall observe and subsequently use. The host should issue the Read Parameter Page (ECh) command. This command returns information that includes the capabilities, features, and operating parameters of the device. 42 When the information is read from the device, the host shall check the CRC to ensure that the data was received correctly and without error prior to taking action on that data. If the CRC of the first parameter page read is not valid (refer to section 5.7.1.4), the host should read redundant parameter page copies. The host can determine whether a redundant parameter page is present or not by checking if the first four bytes contain at least two bytes of the parameter page signature. If the parameter page signature is present, then the host should read the entirety of that redundant parameter page. The host should then check the CRC of that redundant parameter page. If the CRC is correct, the host may take action based on the contents of that redundant parameter page. If the CRC is incorrect, then the host should attempt to read the next redundant parameter page by the same procedure. The host should continue reading redundant parameter pages until the host is able to accurately reconstruct the parameter page contents. The host may use bit-wise majority or other ECC techniques to recover the contents of the parameter page from the parameter page copies present. When the host determines that a parameter page signature is not present (refer to section 5.7.1.1), then all parameter pages have been read. The Read ID and Read Parameter Page commands only use the lower 8-bits of the data bus. The host shall not issue commands that use a word data width on x16 devices until the host determines the device supports a 16-bit data bus width in the parameter page. After successfully retrieving the parameter page, the host has all information necessary to successfully communicate with that target. If the host has not previously mapped defective block information for this target, the host should next map out all defective blocks in the target. The host may then proceed to utilize the target, including erase and program operations. 43 4. Data Interface and Timing 4.1. Data Interface Types ONFI supports two different data interface types: asynchronous and source synchronous. The asynchronous data interface is the traditional NAND interface that uses RE_n to latch data read, WE_n to latch data written, and does not include a clock. The source synchronous data interface includes a clock that indicates where commands and addresses should be latched and a data strobe that indicates where data should be latched. On power-up, the device shall operate in asynchronous data interface timing mode 0. After the host determines that the source synchronous data interface is supported in the parameter page, the host may select a source synchronous timing mode by using Set Features with a Feature Address of 01h. Refer to section 5.26.1. The source synchronous data interface uses a DDR protocol. Thus, an even number of bytes is always transferred. The least significant bit of the column address shall always be zero in the source synchronous data interface. If the least significant bit of the column address is set to one in the source synchronous data interface then the results are indeterminate. 4.1.1. Signal Function Reassignment The function of some signals is different when using the asynchronous data interface versus when using the source synchronous data interface. When source synchronous is selected, the function of RE_n and WE_n is modified and DQS is enabled. WE_n becomes the clock signal (CLK) when in source synchronous mode. CLK shall be enabled with a valid clock period whenever a command cycle, address cycle, or data cycle is occurring. CLK shall maintain the same frequency while CE_n is driven to zero. Refer to section 4.2.3. RE_n becomes the write/read direction signal (W/R_n) when in source synchronous mode. This signal indicates the owner of the DQ data bus and the DQS signal. The host shall only transition W/R_n when ALE and CLE are latched to zero. The I/O bus is renamed to the DQ bus in the source synchronous interface. A strobe signal for the DQ data bus is used in source synchronous mode, called DQS (DQ strobe). DQS is bi-directional and is used for all data transfers. DQS is not used for command or address cycles. The latching edge of DQS is center aligned to the valid data window for data transfers from the host to the device (writes). The latching edge of DQS is aligned to the transition of the DQ bus for data transfers from the device to the host (reads). DQS should be pulled high by the host and shall be ignored by the device when operating in the asynchronous data interface. When W/R_n changes from one to zero, the host shall tri-state the DQ bus and the DQS signal and then the device shall drive DQS to zero. When W/R_n changes from zero to one, the device shall tri-state the DQ bus and the DQS signal. DQS and the DQ bus should be driven high by the host during idle when no data operations are outstanding and W/R_n is set to one. There is a turn-around time whenever W/R_n changes its value where the DQS signal is tri-stated (as neither the host nor the device is driving the signal), see section 4.3.2.6. 44 Symbol Source synchronous ALE CE_n CLE DQ[7:0] DQS W/R_n CLK WP_n R/B_n Asynchronous ALE CE_n CLE I/O[7:0] — RE_n WE_n WP_n R/B_n Table 15 Type Input Input Input I/O I/O Input Input Input Output Description Address latch enable Chip enable Command latch enable Data inputs/outputs Data strobe Read enable / (Write / Read_n direction) Write enable / Clock Write protect Ready / Busy_n Signal Reassignments between Data Interface Types 4.1.2. Bus State ALE and CLE are used to determine the current bus state in asynchronous and source synchronous data interfaces. Table 16 describes the bus state for asynchronous. Note that in asynchronous the value 11b for ALE/CLE is undefined. CE_n ALE CLE WE_n RE_n 1 0 0 0 0 0 0 X 0 0 1 0 0 1 X 0 1 0 0 0 1 X 1 0 0 0 1 X X 1 1 1 1 0 X Table 16 Asynchronous Bus State Standby Idle Command cycle Address cycle Data input cycle Data output cycle Undefined ALE/CLE value and asynchronous bus state Table 17 describes the bus state for source synchronous. In source synchronous the value 11b for ALE/CLE is used for data transfers. The bus state lasts for an entire CLK period, starting with the rising edge of CLK. Thus, for data cycles there are two data input cycles or two data output cycles per bus state. The idle bus state is used to terminate activity on the DQ bus after a command cycle, an address cycle, or a stream of data. The value of CE_n shall only change when the source synchronous bus state is idle (i.e. ALE and CLE are both cleared to zero) and no data is being transmitted during that clock period. 45 CE_n ALE CLE W/R_n CLK 1 X X X 0 0 0 1 0 0 0 0 0 0 1 1 0 1 0 1 0 1 1 1 0 1 1 0 0 0 1 0 0 1 0 0 X Rising edge to rising edge Rising edge to rising edge Rising edge to rising edge Rising edge to rising edge Rising edge to rising edge Rising edge to rising edge Rising edge to rising edge Rising edge to rising edge Source Synchronous Bus State Standby 1 Idle Bus Driving 1 Command cycle Address cycle Data input cycle 2 Data output cycle 2 Reserved Reserved NOTE: 1. When W/R_n is cleared to ‘0’, the device is driving the DQ bus and DQS signal. When W/R_n is set to ‘1’ then the DQ and DQS signals are not driven by the device. 2. There are two data input/output cycles from the rising edge of CLK to the next rising edge of CLK. Table 17 ALE/CLE value and source synchronous bus state 4.1.2.1. Pausing Data Input/Output The host may pause data input or data output by inserting Idle cycles. In the asynchronous data interface, pausing data input or data output is done by maintaining WE_n or RE_n at a value of one, respectively. In the source synchronous data interface, pausing data input or data output is done by clearing ALE and CLE both to zero. The host may continue data transfer by setting ALE and CLE both to one after the applicable tCAD time has passed. 4.1.3. Source Synchronous and Repeat Bytes The source synchronous interface uses DDR to achieve a high data transfer rate. However, certain configuration and settings commands are not often used and do not require a high data transfer rate. Additionally, these commands typically are not serviced by the pipeline used for data transfers. To avoid adding unnecessary complexity and requirements to implementations for these commands, the data is transferred using single data rate. Specifically, the same data byte is repeated twice and shall conform to the timings required for the source synchronous data interface. The data pattern in these cases is D0 D0 D1 D1 D2 D2 etc. The receiver (host or device) shall only latch one copy of each data byte. CLK should not be stopped during data input for these commands. The receiver is not required to wait for the repeated data byte before beginning internal actions. 46 The commands that repeat each data byte twice in the source synchronous data interface are: Set Features, Read ID, Get Features, Read Status, and Read Status Enhanced. 4.1.4. Data Interface / Timing Mode Transitions 4.1.4.1. Asynchronous to Source Synchronous To transition from an asynchronous timing mode to a source synchronous timing mode, the procedure described in this section shall be followed. The Set Features command is used to change the data interface and timing mode. The Set Features command (EFh), Feature Address, and the four parameters are entered using the previously selected timing mode in the asynchronous data interface. When issuing the Set Features command, the host shall drive the DQS signal high. After the fourth parameter, P4, is entered until the tITC time has passed the host shall not issue any commands to the device. Prior to issuing any new commands to the device, the host shall transition CE_n high. When CE_n is high, the host selects the new CLK rate. After issuing the Set Features command and prior to transitioning CE_n high, the host shall observe the following requirements: ALE and CLE shall be cleared to zero RE_n / W/R_n shall be set to one WE_n / CLK shall be set to one DQS shall be set to one 4.1.4.2. Source Synchronous to Source Synchronous To transition from a source synchronous timing mode to another source synchronous timing mode, the procedure described in this section shall be followed. The Set Features command is used to change the timing mode. The Set Features command (EFh), Feature Address, and the four parameters are entered using the previously selected timing mode in the source synchronous data interface. After the fourth parameter, P4, is entered until the tITC time has passed the host shall not issue any commands to the device. Prior to issuing any new commands to the device, the host shall transition CE_n high. When CE_n is high, the host selects the new CLK rate. After issuing the Set Features command and prior to transitioning CE_n high, the host shall observe the following requirements: ALE and CLE shall be cleared to zero W/R_n shall be set to one CLK shall continue running at the previously selected speed grade 4.1.4.3. Source Synchronous to Asynchronous To transition from a source synchronous timing mode to an asynchronous timing mode, the procedure described in this section shall be followed. To transition from the source synchronous data interface to the asynchronous data interface, the Reset (FFh) command shall be used. After the Reset is issued, the host shall not issue any commands to the device until after the tITC time has passed. Note that after the tITC time has passed, only status commands may be issued by the host until the Reset completes. The host shall transition to the asynchronous data interface. Then the host shall issue the Reset (FFh) command described in the previous paragraph using asynchronous timing mode 0, thus the host transitions to the asynchronous data interface prior to issuing the Reset (FFh). A device in any timing mode is required to recognize a Reset (FFh) command issued in asynchronous timing mode 0. After issuing the Reset (FFh) and prior to transitioning CE_n high, the host shall observe the following requirements: ALE and CLE shall be cleared to zero 47 RE_n / W/R_n shall be set to one WE_n / CLK shall be set to one After CE_n has been pulled high and then transitioned low again, the host should issue a Set Features to select the appropriate asynchronous timing mode. 4.2. Timing Parameters All timing parameters are from a host perspective. For example, the “Minimum WE_n pulse width” is the minimum allowed WE_n pulse width that the host is permitted to present to the device while still assuring correct operation of the device. The behavior of the device when the required host minimum and maximum times are not adhered to is undefined. Note that the host needs to account for channel effects in meeting the specified timings with the device. 4.2.1. General Timings This section describes timing parameters that apply regardless of the data interface type being used. For execution of the first Read Parameter Page command, prior to complete initialization, a tR value of 200 microseconds and tCCS value of 500 ns shall be used. For page reads, including execution of additional Read Parameter Page commands after initialization is complete, the value for tR and tCCS contained in the parameter page shall be used. There are three maximums listed for tRST in the asynchronous and source synchronous data interfaces. The target is allowed a longer maximum reset time when a program or erase operation is in progress. The maximums correspond to: 1. The target is not performing an erase or program operation. 2. The target is performing a program operation. 3. The target is performing an erase operation. Table 18 defines the array timing parameters. The array timing parameter values are either returned in the parameter page (tR, tPROG, tBERS, and tCCS) or they are statically defined in Table 19. . Parameter Description 1 tBERS Block erase time tCCS Change Column setup time 1 tPLEBSY 1 tPLPBSY 1 Busy time for multi-plane erase operation Busy time for multi-plane program operation tPLRBSY Busy time for multi-plane read operation tPCBSY Program cache busy time 1 tPROG Page program time 1 tR Page read time 1 tRCBSY Read cache busy time NOTE: 1. Measured from the falling edge of SR[6] to the rising edge of SR[6]. Table 18 Array Timing Parameter Descriptions 48 There are “short” busy times associated with cache operations (tRCBSY, tPCBSY) and multiplane operations (tPLEBSY, tPLPBSY, and tPLRBSY). Typical and maximum times for these busy times are listed in Table 19. Parameter Typical Maximum tPLEBSY 500 ns tBERS tPLPBSY 500 ns tPROG tPLRBSY 500 ns tR tPCBSY 3 µs tPROG tRCBSY 3 µs tR NOTE: 1. Typical times for tPCBSY and tRCBSY are the recommended interval at which the host should consider polling status. Device busy time may be longer than the typical value. Table 19 Cache and Multi-plane Short Busy Times 4.2.2. Asynchronous Table 20 defines the descriptions of all timing parameters. Table 23 and Table 24 define the requirements for timing modes 0, 1, 2, 3, 4, and 5. Timing mode 0 shall always be supported and the device operates in this mode at power-on. A host shall only begin use of a more advanced timing mode after determining that the device supports that timing mode in Read Parameter Page. The host shall use EDO data output cycle timings, as defined in section 4.3.1.5, when running with a tRC value less than 30 ns. 49 Parameter tADL Description 3 ALE to data loading time tALH ALE hold time tALS ALE setup time tAR ALE to RE_n delay 3 tCCS WE_n high to RE_n low, value specified in parameter page tCEA CE_n access time tCEH CE_n high hold time tCH CE_n hold time tCHZ 2 CE_n high to output hi-Z tCLH CLE hold time tCLR CLE to RE_n delay tCLS CLE setup time tCOH CE_n high to output hold tCS CE_n setup time tDH Data hold time tDS Data setup time 1 tFEAT 2 tIR 1 tITC Busy time for Set Features and Get Features Output hi-Z to RE_n low Interface and Timing Mode Change time tRC RE_n cycle time tREA RE_n access time tREH RE_n high hold time tRHOH RE_n high to output hold tRHW 2 tRHZ RE_n high to WE_n low RE_n high to output hi-Z tRLOH RE_n low to output hold tRP RE_n pulse width tRR Ready to RE_n low (data only) Device reset time, measured from the falling edge of R/B_n to the rising edge of R/B_n. tRST 4 tWB WE_n high to SR[6] low tWC WE_n cycle time tWH WE_n high hold time 3 tWHR WE_n high to RE_n low tWP WE_n pulse width tWW WP_n transition to WE_n low NOTE: 1. Measured from the falling edge of SR[6] to the rising edge of SR[6]. 2. Refer to Appendix E for measurement technique. 3. tADL is used for Program operations. tWHR is used for Read ID, Read Status, and Read Status Enhanced commands. tCCS is used for commands that modify the column address and thus impact the data pipeline; these commands include Change Read Column and Change Write Column. 4. Commands (including Read Status / Read Status Enhanced) shall not be issued until 50 after tWB is complete. Table 20 Asynchronous Timing Parameter Descriptions The testing conditions that shall be used to verify that a device complies with a particular asynchronous timing mode are listed in Table 21 for devices that support the asynchronous data interface only and do not support driver strength settings. Parameter Value Input pulse levels 0.0 V to VccQ Input rise and fall times 5 ns Input and output timing levels VccQ / 2 Output load for VccQ of 3.3V CL = 50 pF Output load for VccQ of 1.8V CL = 30 pF Table 21 Testing Conditions for Asynchronous Only Devices The testing conditions that shall be used to verify compliance with a particular timing mode for devices that support driver strength settings are listed in Table 22. This includes all devices that support the source synchronous data interface. It also includes devices that only support the asynchronous data interface that support driver strength settings. Parameter Value Positive input transition VIL (DC) to VIH (AC) Negative input transition VIH (DC) to VIL (AC) Minimum input slew rate tIS = 1.0 V/ns Input timing levels VccQ / 2 Output timing levels VccQ / 2 Driver strength Output capacitive load Nominal 1 CL = 5 pF EZ NAND output reference load 50 Ohm to Vtt NOTE: 1. Assumes small propagation delay from output to CL. Table 22 Testing Conditions for Devices that Support Driver Strength Settings NAND Package Rtt = 50 Ohms Output Vtt = 0.5 x VccQ 51 Figure 22 EZ NAND Output Reference Load 52 Parameter Mode 0 Mode 1 Mode 2 Unit 100 50 35 ns Min Max Min Max Min Max tADL 200 — 100 — 100 — ns tALH 20 — 10 — 10 — ns tALS 50 — 25 — 15 — ns tAR 25 — 10 — 10 — ns tCEA — 100 — 45 — 30 ns tCEH 20 — 20 — 20 — ns tCH 20 — 10 — 10 — ns tCHZ — 100 — 50 — 50 ns tCLH 20 — 10 — 10 — ns tCLR 20 — 10 — 10 — ns tCLS 50 — 25 — 15 — ns tCOH 0 — 15 — 15 — ns tCS 70 — 35 — 25 — ns tDH 20 — 10 — 5 — ns tDS 40 — 20 — 15 — ns tFEAT — 1 — 1 — 1 µs tIR 10 — 0 — 0 — ns tITC — 1 — 1 — 1 µs tRC 100 — 50 — 35 — ns tREA — 40 — 30 — 25 ns tREH 30 — 15 — 15 — ns tRHOH 0 — 15 — 15 — ns tRHW 200 — 100 — 100 — ns tRHZ — 200 — 100 — 100 ns tRLOH 0 — 0 — 0 — ns tRP 50 — 25 — 17 — ns tRR 40 — 20 20 — 5000 — 2 tRST — 250000 — tWB — 200 — — — 10/30/ 500 150/ 150/ 500 100 ns tRST (raw NAND) — 10/30/ 500 150/ 150/ 500 100 tWC 100 — 45 — 35 — ns tWH 30 — 15 — 15 — ns tWHR 120 — 80 — 80 — ns tWP 50 — 25 — 17 — ns tWW 100 — 100 — 100 — (EZ NAND) — — µs µs ns ns NOTE: 1. To easily support EDO capable devices, tCHZ and tRHZ maximums are higher in modes 1, 2, and 3 than typically necessary for a non-EDO capable device. 2. If the reset is invoked using a Reset (FFh) command then the EZ NAND device has 250 ms to complete the reset operation regardless of the timing mode. If the reset is invoked using LUN Reset (FAh) then the values are as shown. 53 Table 23 Asynchronous Timing Modes 0, 1, and 2 54 Parameter Mode 4 (EDO capable) Mode 3 30 Mode 5 (EDO capable) Unit 20 ns 25 Min Max Min Max Min Max 70 — 70 — ns 100 — tALH 5 — 5 — 5 — ns tALS 10 — 10 — 10 — ns tAR 10 — 10 — 10 — ns tCEA — 25 — 25 — 25 ns tCEH 20 — 20 — 20 — ns tCH 5 — 5 — 5 — ns tCHZ — 50 — 30 — 30 ns tCLH 5 — 5 — 5 — ns tCLR 10 — 10 — 10 — ns tCLS 10 — 10 — 10 — ns tCOH 15 — 15 — 15 — ns tCS 25 — 20 — 15 — ns tDH 5 — 5 — 5 — ns tDS 10 — 10 — 7 — ns tFEAT — 1 — 1 — 1 µs tIR 0 — 0 — 0 — ns tITC — 1 — 1 — 1 µs tRC 30 — 25 — 20 — ns tREA — 20 — 20 — 16 ns tREH 10 — 10 — 7 — ns tRHOH 15 — 15 — 15 — ns tRHW 100 — 100 — 100 — ns tRHZ — 100 — 100 — 100 ns tRLOH 0 — 5 — 5 — ns tRP 15 — 12 — 10 — ns tRR 20 20 20 — 2 tRST — 150/15 0/500 — — — 10/30/ 500 150/ 150/ 500 100 ns tRST (raw NAND) — 10/30/ 500 20 — ns tADL tWB — 100 — — 10/30/ 500 150/ 150/ 500 100 tWC 30 — 25 — tWH 10 — 10 — 7 — ns tWHR 60 — 60 — 60 — ns tWP 15 — 12 — 10 — ns tWW 100 — 100 — 100 — (EZ NAND) — — — µs µs ns ns NOTE: 1. To easily support EDO capable devices, tCHZ and tRHZ maximums are higher in modes 1, 2, and 3 than typically necessary for a non-EDO capable device. 2. If the reset is invoked using a Reset (FFh) command then the EZ NAND device has 250 ms to complete the reset operation regardless of the timing mode. If the reset is invoked using LUN Reset (FAh) then the values are as shown. 55 Table 24 Asynchronous Timing Modes 3, 4, and 5 4.2.3. Source Synchronous All source synchronous timing parameters are referenced to the rising edge of CLK or the latching edge of DQS. Note that R/B_n and WP_n are always asynchronous signals. If CLK is a different frequency than those described in the source synchronous timing modes, then the host shall meet the setup and hold requirements for the next fastest timing mode. For parameters measured in clocks (e.g. tDSH), the parameter is measured starting from a latching edge of CLK or DQS, respectively. 56 Parameter Description tAC Access window of DQ[7:0] from CLK tADL 5 Address cycle to data loading time tCAH Command, Address, Data delay (command to command, address to address, command to address, address to command, command/address to start of data) Command/address DQ hold time tCALH W/R_n, CLE and ALE hold time tCADf, tCADs tCALS W/R_n, CLE and ALE setup time tCAS 5 tCCS Command/address DQ setup time Change Column setup time, value specified in parameter page tCEH CE_n high hold time tCH CE_n hold time tCK(avg) 2 Absolute clock period, measured from rising edge to the next consecutive rising edge tCK(abs) 3 tCKH(abs) tCKL(abs) Average clock cycle time, also known as tCK Clock cycle high 3 Clock cycle low tCKWR Data output end to W/R_n high tCS CE_n setup time tDH Data DQ hold time tDPZ Data input pause setup time tDQSCK 4 tDQSD Access window of DQS from CLK W/R_n low to DQS/DQ driven by device tDQSH DQS input high pulse width tDQSHZ 4 W/R_n high to DQS/DQ tri-state by device tDQSL DQS input low pulse width tDQSQ DQS-DQ skew, DQS to last DQ valid, per access tDQSS Data input to first DQS latching transition tDS Data DQ setup time tDSC DQS cycle time tDSH DQS falling edge to CLK rising – hold time tDSS DQS falling edge to CLK rising – setup time tDVW Output data valid window 1 tFEAT Busy time for Set Features and Get Features tHP Half-clock period 57 1 tITC Interface and Timing Mode Change time tJIT(per) The deviation of a given tCK(abs) from tCK(avg) tQH DQ-DQS hold, DQS to first DQ to go non-valid, per access tQHS Data hold skew factor tRHW Data output cycle to command, address, or data input cycle tRR Ready to data output cycle (data only) Device reset time, measured from the falling edge of R/B_n to the rising edge of R/B_n. tRST 6 tWB 5 tWHR CLK rising edge to SR[6] low Command, address or data input cycle to data output cycle tWPRE DQS write preamble tWPST DQS write postamble tWRCK W/R_n low to data output cycle tWW WP_n transition to command cycle NOTE: 1. Measured from the falling edge of SR[6] to the rising edge of SR[6]. 2. tCK(avg) is the average clock period over any consecutive 200 cycle window. 3. tCKH(abs) and tCKL(abs) include static offset and duty cycle jitter. 4. Refer to Appendix E for measurement technique. 5. tADL is used for Program operations. tWHR is used for Read ID, Read Status, and Read Status Enhanced commands. tCCS is used for commands that modify the column address and thus impact the data pipeline; these commands include Change Read Column and Change Write Column. 6. Commands (including Read Status / Read Status Enhanced) shall not be issued until after tWB is complete. Table 25 Source Synchronous Timing Parameter Descriptions The device may be configured with multiple driver strengths with the Set Features command. There is a 50 Ohm, 35 Ohm, 25 Ohm, and 18 Ohm setting that the device may support. Support for all four driver strength settings is required for devices that support the source synchronous data interface. A device that only supports the asynchronous data interface may support all or a subset of driver strength settings. Devices that support driver strength settings shall comply with the output driver requirements in this section. Setting Driver Strength 18 Ohms 2.0x = 18 Ohms 25 Ohms 1.4x = 25 Ohms 35 Ohms 1.0x = 35 Ohms 50 Ohms 0.7x = 50 Ohms 18 Ohms 2.0x = 18 Ohms 25 Ohms 1.4x = 25 Ohms 35 Ohms 1.0x = 35 Ohms 50 Ohms 0.7x = 50 Ohms Table 26 VccQ 3.3V 1.8V I/O Drive Strength Settings The impedance values correspond to several different VccQ values are defined in Table 28 for 3.3V VccQ and Table 29 for 1.8V VccQ. The test conditions that shall be used to verify the impedance values is specified in Table 27. 58 Condition Temperature (TA) VccQ (3.3V) VccQ (1.8V) Process Minimum Impedance TOPER (Min) degrees Celsius 3.6V 1.95V Fast-fast Nominal Impedance 25 degrees Celsius 3.3V 1.8V Typical Maximum impedance TOPER (Max) degrees Celsius 2.7V 1.7V Slow-slow Table 27 Testing Conditions for Impedance Values 59 RON = 18 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal Minimum Unit 0.2 x VccQ 28.7 16.2 7.0 Ohms 0.5 x VccQ 36.0 18.0 9.0 Ohms 0.8 x VccQ 50.0 21.0 11.8 Ohms 0.2 x VccQ 50.0 21.0 11.8 Ohms 0.5 x VccQ 36.0 18.0 9.0 Ohms 0.8 x VccQ 28.7 14.0 7.0 Ohms Minimum Unit RON = 25 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal 0.2 x VccQ 40.0 22.3 9.3 Ohms 0.5 x VccQ 50.0 25.0 12.6 Ohms 0.8 x VccQ 68.0 29.0 16.3 Ohms 0.2 x VccQ 68.0 29.0 16.3 Ohms 0.5 x VccQ 50.0 25.0 12.6 Ohms 0.8 x VccQ 40.0 19.0 9.3 Ohms RON = 35 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal Minimum Unit 0.2 x VccQ 58.0 32.0 12.8 Ohms 0.5 x VccQ 70.0 35.0 18.0 Ohms 0.8 x VccQ 95.0 40.0 23.0 Ohms 0.2 x VccQ 95.0 40.0 23.0 Ohms 0.5 x VccQ 70.0 35.0 18.0 Ohms 0.8 x VccQ 58.0 32.0 12.8 Ohms RON = 50 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal Minimum Unit 0.2 x VccQ 80.0 45.0 18.4 Ohms 0.5 x VccQ 100.0 50.0 25.0 Ohms 0.8 x VccQ 136.0 57.0 32.0 Ohms 0.2 x VccQ 136.0 57.0 32.0 Ohms 0.5 x VccQ 100.0 50.0 25.0 Ohms 0.8 x VccQ 80.0 45.0 18.4 Ohms Table 28 Impedance Values for 3.3V VccQ 60 RON = 18 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal Minimum Unit 0.2 x VccQ 34.0 13.5 7.5 Ohms 0.5 x VccQ 31.0 18.0 9.0 Ohms 0.8 x VccQ 44.0 23.5 11.0 Ohms 0.2 x VccQ 44.0 23.5 11.0 Ohms 0.5 x VccQ 31.0 18.0 9.0 Ohms 0.8 x VccQ 34.0 13.5 7.5 Ohms RON = 25 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal Minimum Unit 0.2 x VccQ 47.0 19.0 10.5 Ohms 0.5 x VccQ 44.0 25.0 13.0 Ohms 0.8 x VccQ 61.5 32.5 16.0 Ohms 0.2 x VccQ 61.5 32.5 16.0 Ohms 0.5 x VccQ 44.0 25.0 13.0 Ohms 0.8 x VccQ 47.0 19.0 10.5 Ohms RON = 35 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal Minimum Unit 0.2 x VccQ 66.5 27.0 15.0 Ohms 0.5 x VccQ 62.5 35.0 18.0 Ohms 0.8 x VccQ 88.0 52.0 22.0 Ohms 0.2 x VccQ 88.0 52.0 22.0 Ohms 0.5 x VccQ 62.5 35.0 18.0 Ohms 0.8 x VccQ 66.5 27.0 15.0 Ohms RON = 50 Ohms Description R_pulldown R_pullup VOUT to VssQ Maximum Nominal Minimum Unit 0.2 x VccQ 95.0 39.0 21.5 Ohms 0.5 x VccQ 90.0 50.0 26.0 Ohms 0.8 x VccQ 126.5 66.5 31.5 Ohms 0.2 x VccQ 126.5 66.5 31.5 Ohms 0.5 x VccQ 90.0 50.0 26.0 Ohms 0.8 x VccQ 95.0 39.0 21.5 Ohms Table 29 Impedance Values for 1.8V VccQ The pull-up and pull-down impedance mismatch is defined in Table 30. Impedance mismatch is the absolute value between pull-up and pull-down impedances. Both are measured at the same temperature and voltage. The testing conditions that shall be used to verify the impedance mismatch requirements are: VccQ = VccQ(min), VOUT = VccQ x 0.5, and T A is across the full operating range. 61 Maximum Minimum Unit 18 Ohms 6.3 0.0 Ohms 25 Ohms 8.8 0.0 Ohms 35 Ohms 12.3 0.0 Ohms 50 Ohms 17.5 0.0 Ohms Output Impedance Table 30 Pull-up and Pull-down Impedance Mismatch The input slew rate requirements that the device shall comply with are defined in Table 31. The testing conditions that shall be used to verify the input slew rate are listed in Table 32. Description Timing Mode 2 3 0 1 Input slew rate (min) 0.5 0.5 0.5 Input slew rate (max) 4.5 4.5 4.5 Derating factor for setup times, address and command TBD TBD Derating factor for hold times, address and command TBD Derating factor for setup times, data input and output Derating factor for hold times, data input and output 5 0.5 0.5 0.5 V/ns 4.5 4.5 4.5 V/ns TBD TBD TBD TBD ns per 100 mV TBD TBD TBD TBD TBD ns per 100 mV TBD TBD TBD TBD TBD TBD ns per 100 mV TBD TBD TBD TBD TBD TBD ns per 100 mV Table 31 Input Slew Rate Requirements Parameter Value Positive input transition VIL (DC) to VIH (AC) Negative input transition VIH (DC) to VIL (AC) Table 32 Unit 4 Testing Conditions for Input Slew Rate The output slew rate requirements that the device shall comply with are defined in Table 33 and Table 34 for a single LUN per 8-bit data bus. The testing conditions that shall be used to verify the output slew rate are listed in Table 35. Output Slew Rate Min Max Unit Normative or Recommended 18 Ohms 1.5 10.0 V/ns Normative 25 Ohms 1.5 9.0 V/ns Normative 35 Ohms 1.2 7.0 V/ns Normative 50 Ohms 1.0 5.5 V/ns Recommended Description Table 33 Output Slew Rate Requirements for 3.3V VccQ 62 Output Slew Rate Min Max Unit Normative or Recommended 18 Ohms 1.0 5.5 V/ns Normative 25 Ohms 0.85 5.0 V/ns Normative 35 Ohms 0.75 4.0 V/ns Normative 50 Ohms 0.60 4.0 V/ns Recommended Description Table 34 Output Slew Rate Requirements for 1.8V VccQ Parameter Value VOL(DC) 0.3 * VccQ VOH(DC) 0.7 * VccQ VOL(AC) 0.2 * VccQ VOH(AC) 0.8 * VccQ Positive output transition VOL (DC) to VOH (AC) Negative output transition 1 tRISE 1 tFALL VOH (DC) to VOL (AC) Time during falling edge from VOH(DC) to VOL(AC) Output slew rate rising edge [VOH(AC) – VOL(DC)] / tRISE Output slew rate falling edge [VOH(DC) – VOL(AC)] / tFALL Time during rising edge from VOL(DC) to VOH(AC) Output load Refer to Table 22 NOTE: 1. Refer to Figure 23. 2. Output slew rate is verified by design and characterization; it may not be subject to production test. 3. The minimum slew rate is the minimum of the rising edge and the falling edge slew rate. The maximum slew rate is the maximum of the rising edge and the falling edge slew rate. Table 35 Figure 23 Testing Conditions for Output Slew Rate tRISE and tFALL Definition for Output Slew Rate 63 The testing conditions that shall be used to verify that a device complies with a particular source synchronous timing mode are listed in Table 36. The test conditions are the same regardless of the number of LUNs per Target. Parameter Value Positive input transition VIL (DC) to VIH (AC) Negative input transition VIH (DC) to VIL (AC) Minimum input slew rate tIS = 1.0 V/ns Input timing levels VccQ / 2 Output timing levels VccQ / 2 Driver strength Nominal Output capacitive load 1 CL = 5 pF NOTE: 1. Assumes small propagation delay from output to CL. Table 36 Testing Conditions for Source Synchronous Timing Modes The input capacitance requirements are defined in Table 37. The testing conditions that shall be used to verify the input capacitance requirements are: temperature of 25 degrees Celsius, VIN = 0V, and a CLK frequency of 100 MHz. The capacitance delta values measure the pin-to-pin capacitance for all LUNs within a package, including across data buses. The capacitance delta values change based on the number of LUNs per x8 data bus. All targets that share an I/O should report an equivalent Typical capacitance in order to meet the capacitance delta requirements. If targets with different Typical capacitance values share an I/O bus, then the values in these tables are not applicable and detailed topology and signal integrity analysis needs to be completed by the implementer to determine the bus speed that is achievable. EZ NAND has a controller stacked in the NAND package and the LUNs are physically connected to the controller and not to the external data bus. EZ NAND values are defined separately from the values specified in Table 37. For the Typical ranges in Table 38 and Table 39, EZ NAND is specified separately since ASIC capacitance tends to be higher. Parameter Number of LUNs per x8 data bus 1 2 4 8 Typical Variance ±1.0 ±1.5 ±2.0 ±4.0 Delta 1.4 1.7 2.0 4.0 Unit pF pF NOTE: 1. Typical capacitance values for pin groups CCK, CIN, and CIO are specified in the parameter page. The allowable range for Typical capacitance values is specified in Table 38 for CLK and input pins and Table 39 for I/O pins. 2. Input capacitance for CE_n and WP_n shall not exceed the maximum capacitance value for CLK and input pins. However, CE_n and WP_n do not have to meet delta or variance requirements. Table 37 Input Capacitance, Minimum and Maximums The Typical capacitance values shall be constrained to the ranges defined in Table 38 for CLK and input pins and Table 39 for I/O pins for devices in a BGA package. Capacitance is shared for 64 LUNs that share the same 8-bit data bus in the same package, thus the ranges are specific to the number of LUNs per data bus. Parameter Min Source Synchronous Timing Modes 0-5 Typ Typ Low High Max Typ Variance Unit 1 LUN per x8 data bus 2.0 3.0 5.4 6.4 ±0.5 pF 2 LUNs per x8 data bus 3.5 5.0 7.7 9.2 ±1.0 pF 4 LUNs per x8 data bus 5.0 7.0 12.3 14.3 ±2.0 pF 8 LUNs per x8 data bus 6.0 10.0 23.3 27.3 ±4.0 pF EZ NAND 3.0 5.0 6.5 8.5 ±2.0 pF Table 38 Input Capacitance for CLK and input pins, Typical Ranges Parameter Min Source Synchronous Timing Modes 0-5 Typ Typ Low High Max Typ Variance Unit 1 LUN per x8 data bus 3.0 4.0 5.7 6.7 ±0.5 pF 2 LUNs per x8 data bus 5.0 6.5 8.5 10.0 ±1.0 pF 4 LUNs per x8 data bus 8.9 10.9 14.9 16.9 ±2.0 pF 8 LUNs per x8 data bus 16.7 20.7 28.7 32.7 ±4.0 pF EZ NAND 3.0 5.0 6.5 8.5 ±2.0 pF Table 39 Input Capacitance for I/O pins, Typical Ranges NOTE: Capacitance ranges are not defined for the TSOP package due to the varying TSOP package construction techniques and bond pad locations. For the TSOP package compared to the BGA package, the input capacitance delta values do not apply, input capacitance values for I/O pins is similar, and input capacitance values for input pins could be significantly higher. For higher speed applications, the BGA-63 or BGA-100 packages are recommended due to their lower and more consistent input capacitance and input capacitance delta values. Table 40 describes the standard source synchronous timing modes. The host is not required to have a clock period that exactly matches any of the clock periods listed for the standard timing modes. The host shall meet the setup and hold times for the timing mode selected. If the host selects timing mode n using Set Features, then its clock period shall be faster than the clock period of timing mode n-1 and slower than or equal to the clock period of timing mode n. For example, if the host selects timing mode 2, then the following equation shall hold: 30 ns > host clock period >= 20 ns 65 If timing mode 0 is selected, then the clock period shall be no slower than 100 ns. The only exception to this requirement is when the host is issuing a Reset (FFh) in asynchronous timing mode 0 (see section 4.1.4.3). Timing parameters that indicate a latency requirement before a data input, data output, address or command cycle shall be satisfied to the rising clock edge after the latency in nanoseconds has elapsed. To calculate the first edge where the associated transition may be made, it is calculated as follows: = RoundUp{[tParam + tCK] / tCK} 66 Parameter Mode 0 50 Mode 1 30 ~20 tAC tADL tCADf tCADs tCAH tCALH tCALS tCAS tCEH tCH tCK(avg) or tCK Mode 2 20 ~33 Mode 3 15 ~50 ~66 Min 3 100 25 45 10 10 10 10 20 10 Max 25 — — — — — — — — — Min 3 100 25 45 5 5 5 5 20 5 Max 25 — — — — — — — — — Min 3 70 25 45 4 4 4 4 20 4 Max 25 — — — — — — — — — Min 3 70 25 45 3 3 3 3 20 3 Max 25 — — — — — — — — — 50 — 30 — 20 — 15 — Mode 4 12 ~83 Min Max 3 70 25 45 2.5 2.5 2.5 2.5 20 2.5 12 tCS tDH tDPZ tDQSCK tDQSD 35 5 1.5 3 0 — — — 25 18 25 2.5 1.5 3 0 Minimum: tCK(avg) + tJIT(per) min Maximum: tCK(avg) + tJIT(per) max 0.57 0.43 0.57 0.43 0.57 0.43 0.57 0.43 0.57 0.43 0.57 0.43 Minimum: RoundUp{[tDQSCK(max) + tCK] / tCK} Maximum: — 15 — 15 — 15 — 1.1 — 1.7 — 1.3 — — 1.5 — 1.5 — 1.5 25 3 25 3 25 3 18 0 18 0 18 0 tDQSH 0.4 0.6 0.4 0.6 0.4 0.6 0.4 0.6 tDQSHZ — 20 — 20 — 20 — tDQSL 0.4 0.6 0.4 0.6 0.4 0.6 tDQSQ tDQSS tDS tDSC tDSH — 0.75 5 50 0.2 5 1.25 — — — — 0.75 3 30 0.2 2.5 1.25 — — — — 0.75 2 20 0.2 1.7 1.25 — — — tCK(abs) tCKH(abs) tCKL(abs) 0.43 0.43 0.57 0.57 0.43 0.43 tCKWR 67 25 — — — — — — — — — — Mode 5 10 ~100 Min Max 3 70 25 45 2 2 2 2 20 2 10 Unit ns MHz 25 — — — — — — — — — ns ns ns ns ns ns ns ns ns ns — ns ns 0.57 0.57 0.43 0.43 0.57 0.57 tCK tCK tCK — — — 25 18 15 0.9 1.5 3 0 — — — 25 18 0.4 0.6 0.4 0.6 20 — 20 — 20 0.4 0.6 0.4 0.6 0.4 0.6 — 0.75 1.5 15 0.2 1.3 1.25 — — — — 0.75 1.1 12 0.2 1.0 1.25 — — — — 0.75 0.9 10 0.2 0.85 1.25 — — — ns ns tDSC ns ns tCK or 4 tDSC ns tCK or 4 tDSC ns tCK ns ns tCK tDSS tDVW tFEAT tHP tITC tJIT(per) tQH tQHS tRHW tRR tRST (raw NAND) 2 tRST (EZ NAND) 0.2 — 0.2 — 0.2 — 1 — 1 — — -0.7 1 0.7 — -0.7 1 0.7 — -0.7 — 100 20 6 — — 10/30/ 500 150/ 150/ 500 100 — — — — — — 100 20 3 — — 10/30/ 500 150/ 150/ 500 100 — — — — — — 100 20 — — — — — — — 0.2 tDVW = tQH – tDQSQ 1 — tHP = min(tCKL, tCKH) 1 — 0.7 -0.6 tQH = tHP – tQHS 2 — — 100 — 20 10/30/ — 500 150/ 150/ — 500 100 — — 80 — 1.5 — 1.5 — 20 — 100 — 0.2 — 0.2 — 1 — 1 — 1 1 0.6 — -0.6 1 0.6 — -0.5 1 0.5 1.5 — — 10/30/ 500 150/ 150/ 500 100 — — — — — — 100 20 1.2 — — 10/30/ 500 150/ 150/ 500 100 — — — — — — 100 20 1.0 — — 10/30/ 500 150/ 150/ 500 100 — — — — — — — — — tCK ns µs ns µs ns ns ns ns ns µs µs tWB — — — — — ns tWHR 80 80 80 80 80 ns tWPRE 1.5 1.5 1.5 1.5 1.5 tCK tWPST 1.5 1.5 1.5 1.5 1.5 tCK tWRCK 20 20 20 20 20 ns tWW 100 100 100 100 100 ns NOTE: 1. tDQSHZ is not referenced to a specific voltage level, but specifies when the device output is no longer driving. 2. tCK(avg) is the average clock period over any consecutive 200 cycle window. 3. tCKH(abs) and tCKL(abs) include static offset and duty cycle jitter. 4. tDQSL and tDQSH are relative to tCK when CLK is running. If CLK is stopped during data input, then tDQSL and tDQSH are relative to tDSC. 5. If the reset is invoked using a Reset (FFh) command then the EZ NAND device has 250 ms to complete the reset operation regardless of the timing mode. If the reset is invoked using Synchronous Reset (FCh) or LUN Reset (FAh) then the values are as shown. Table 40 68 Source Synchronous Timing Modes 4.3. Timing Diagrams 4.3.1. Asynchronous 4.3.1.1. Command Latch Timings The requirements for the R/B_n signal only apply to commands where R/B_n is cleared to zero after the command is issued, as specified in the command definitions. CLE tCLS tCLH tCS tCH CE_n tWP WE_n tALS tALH ALE tDS IO0-7 tDH Command R/B_n tWB Figure 24 69 Command latch timings 4.3.1.2. Address Latch Timings tCLS CLE tCS CE_n tWC tWP WE_n tWH ALE tALS tDS IO0-7 t ALH tDH Address Figure 25 70 Address latch timings 4.3.1.3. Data Input Cycle Timings Data input may be used with CE_n don’t care. However, if CE_n don’t care is used tCS and tCH timing requirements shall be met by the host, and WE_n falling edge shall always occur after the CE_n falling edge (i.e. tCS > tWP). tCLH CLE tCH CE_n tWP tWP WE_n tWP tWH t WC tALS ALE tDS IOx t DH DIN 0 Figure 26 71 tDS tDH DIN 1 Data input cycle timings tDS tDH DIN n 4.3.1.4. Data Output Cycle Timings Data output may be used with CE_n don’t care. However, if CE_n don’t care is used tCEA and tCOH timing requirements shall be met by the host and RE_n shall either remain low or RE_n falling edge shall occur after the CE_n falling edge. tCHZ tCEA CE_n t COH tRP RE_n t RP tRP t RHOH tREH tRR tRHZ tRC R/B_n tRHZ tREA IOx DOUT 0 Figure 27 72 tREA tRHZ DOUT 1 Data output cycle timings tREA tRHZ DOUT n 4.3.1.5. Data Output Cycle Timings (EDO) EDO data output cycle timings shall be used if the host drives tRC less than 30 ns. Data output may be used with CE_n don’t care. However, if CE_n don’t care is used tCEA and tCOH timing requirements shall be met by the host. tCHZ CE_n tCOH tRP RE_n tRR R/B_n t REH tRC tREA tRHZ tREA tRHOH tRLOH IOx DOUT 0 DOUT 1 tCEA Figure 28 73 EDO data output cycle timings DOUT n 4.3.1.6. Read Status Timings tCLR CLE tCLH tCLS tCS tCH tCEA CE_n tWP tCHZ WE_n t WHR tCOH RE_n tRHZ tDS IO7-0 t DH tRHOH 70h Status t IR Figure 29 74 Read Status timings tREA 4.3.1.7. Read Status Enhanced Timings CLE tCLH tCLS tCEA tCH tCS CE_n tWC tWP tWP WE_n tWH t ALH tCHZ tALH tALS tAR tCOH ALE RE_n tRHZ tDS IO7-0 tDH 78h tWHR R1 Figure 30 75 R2 R3 Read Status Enhanced timings t REA tRHOH Status 4.3.2. Source Synchronous For the command, address, data input, and data output diagrams, the tCS timing parameter may consume multiple clock cycles. The host is required to satisfy tCS by the rising edge of CLK shown in the diagrams, and thus needs to pull CE_n low far enough in advance to meet this requirement (which could span multiple clock cycles). 4.3.2.1. Command Cycle Timings tCH CE_n tCS tCALS CLE tCALS tCALH tCALS tCAD ALE tCKL tCKH CLK tCAD starts for next non-idle cycle tCK W/R_n tCALS tCALH tDQSHZ DQS tCAS DQ[7:0] tCAH Command Figure 31 Command cycle timings NOTE: 1. The cycle that tCAD is measured from may be an idle cycle (as shown), another command cycle, an address cycle, or a data cycle. The idle cycle is shown in this diagram for simplicity. 2. ALE and CLE shall be in a valid state when CE_n transitions from one to zero. In the diagram, it appears that tCS and tCALS are equivalent times. However, tCS and tCALS values are not the same, the timing parameter values should be consulted in Table 40. 76 4.3.2.2. Address Cycle Timings tCH CE_n tCS tCALS CLE tCAD tCALS ALE tCALS tCALH tCKL tCKH CLK tCAD starts for next non-idle cycle tCK W/R_n tCALS tCALH tDQSHZ DQS tCAS DQ[7:0] tCAH Address Figure 32 Address cycle timings NOTE: 1. ALE and CLE shall be in a valid state when CE_n transitions from one to zero. In the diagram, it appears that tCS and tCALS are equivalent times. However, tCS and tCALS values are not the same, the timing parameter values should be consulted in Table 40. 77 4.3.2.3. Data Input Cycle Timings Data input cycle timing describes timing for data transfers from the host to the device (i.e. data writes). For the Set Features command, the same data byte is repeated twice. The data pattern in this case is D0 D0 D1 D1 D2 D2 etc. The device shall only latch one copy of each data byte. CLK should not be stopped during data input for the Set Features command. The device is not required to wait for the repeated data byte before beginning internal actions. 78 tCH CE_n tCS CLE tCALS tCALH tCALS tCALH tCAD ALE tCKL tCKH CLK tCAD starts for next non-idle cycle tCK W/R_n tDSH tDQSS tDSS tDSH tDSH tDSS tDSH tDSS DQS tWPRE DQ[7:0] tDQSH tDQSL tDQSH D0 D2 D1 D3 tDQSL tDQSH DN-2 tDH 79 DN tDS tDS Figure 33 DN-1 tWPST Data input cycle timing tDH 4.3.2.4. Data Input Cycle Timings, CLK stopped The host may save power during the data input cycles by holding the CLK signal high (i.e. stopping the CLK). The host may only stop the CLK during data input if the device supports this feature as indicated in the parameter page. Data input cycle timing describes timing for data transfers from the host to the device (i.e. data writes). Figure 34 describes data input cycling with the CLK signal stopped. The values of the ALE, CLE, and W/R_n signals are latched on the rising edge of CLK and thus while CLK is held high these signals are don’t care. Figure 35 shows data input cycling with the CLK signal stopped where the host has optionally paused data input. The host may pause data input if it observes the tDPZ timing parameter for re-starting data input to the device. When re-starting the CLK, the host shall observe the indicated timing parameters in Figure 34 and Figure 35, which include tDSS and tDSH. 80 A B tCH CE_n tCS tCALS CLE tCALS tCALH tCAD tCALS ALE tCALS tCALH tCKL tCKH CLK tCAD starts for next non-idle cycle tCK W/R_n tCALS tCALH tDSS tDSH tDSC tDQSS tDSS DQS tWPRE DQ[7:0] tDQSH tDQSL tDQSH Dn D1 D2 D3 tDQSL tDQSH DN-2 DN-1 tDH 81 DN tDS tDS Figure 34 tWPST Data input cycle timing, CLK stopped tDH A B CE_n tCALS CLE tCALH tCALS ALE tCALH CLK W/R_n tCALS tCALH tDQSS tDSC tDSC tDPZ tDSS tDSH DQS tDQSH tDQSL tDQSH DQ[7:0] DM DM+1 DM+2 DM+3 tDQSH tDQSL tDQSH DM+4 DM+5 DM+6 DM+7 tDQSL tDQSH DN-2 DN-1 DN tDS tDS tDH tDH Figure 35 Data input cycle timing, CLK stopped with data pause 82 4.3.2.5. Data Output Cycle Timings Data output cycle timing describes timing for data transfers from the device to the host (i.e. data reads). The host shall not start data output (i.e. transition ALE/CLE to 11b) until the tDQSD time has elapsed. For the Read ID, Get Features, Read Status, and Read Status Enhanced commands, the same data byte is repeated twice. The data pattern in this case is D0 D0 D1 D1 D2 D2 etc. The host shall only latch one copy of each data byte. A calculated parameter, tCKWR, indicates when W/R_n may be transitioned from a zero to one. This parameter is calculated as: tCKWR(min) = RoundUp{[tDQSCK(max) + tCK] / tCK} 83 tCH CE_n tCS tCAD starts for next non-idle cycle CLE tCALS tCALH tCALS tCALH tCAD ALE tCKH tCKL tHP CLK tHP tHP tHP tCK tHP tHP tCKWR tDQSCK tDQSCK tDQSCK tCALS tWRCK tDQSCK W/R_n tDQSCK tDQSCK tCALS tDQSHZ tDQSD DQS tAC DQ[7:0] tDVW tDVW tDVW D0 D1 tQH tQH tDQSQ Don’t Care Data Transitioning Figure 36 84 tDVW tDVW DN-2 D2 tDQSQ DN-1 DN tDQSQ Device Driving Data output cycle timing tDQSQ tQH tQH 4.3.2.6. W/R_n Behavior Timings Figure 37 describes the ownership transition of the DQ bus and DQS signal. The host owns the DQ bus and DQS signal when W/R_n is one. The device owns the DQ bus and DQS signal when W/R_n is zero. The host shall tri-state the DQ bus and DQS signal whenever W/R_n is zero. When W/R_n transitions from one to zero, the bus ownership is assumed by the device. The device shall start driving the DQS signal low within tDQSD after the transition of W/R_n to zero. When W/R_n transitions from zero to one, the bus ownership is assumed by the host. The device shall tri-state the DQ bus and DQS signal within tDQSHZ after the transition of W/R_n to one. tCALS tCALS CLK (WE_n) W/R_n (RE_n) DQ[7:0] tDQSD tDQSHZ DQS Tri-state Device Driving Figure 37 85 W/R_n timing 4.3.2.7. Satisfying Timing Requirements In some cases there are multiple timing parameters that shall be satisfied prior to the next phase of a command operation. For example, both tDQSD and tCCS shall be satisfied prior to data output commencing for the Change Write Column command. The host and device shall ensure all timing parameters are satisfied. In cases where tADL, tCCS, tRHW, or tWHR are required, then these are the governing parameters (i.e. these parameters are the longest times). Figure 38 and Figure 39 show an example of a Read Status command that includes all the timing parameters for both the command and data phase of the operation. It may be observed that tWHR is the governing parameter prior to the data transfer starting. Also note that the same data byte is transmitted twice (D0, D0) for the Read Status command. A CE_n tCS tCALS CLE tCALS tCALH tWHR tCALS tCAD tCAD ALE tCKL tCKH CLK tCK W/R_n tCALS tCALH tDQSHZ DQS tCAS DQ[7:0] tCAH 70h Figure 38 Read Status including tWHR and tCAD timing requirements NOTE: 1. ALE and CLE shall be in a valid state when CE_n transitions from one to zero. In the diagram, it appears that tCS and tCALS are equivalent times. However, tCS and tCALS values are not the same, the timing parameter values should be consulted in Table 40. 86 A tCH CE_n CLE tCALS tWHR ALE tCALS tHP CLK tHP tCALS tDQSCK tDQSD W/R_n tDQSCK tCALS tDQSHZ DQS tAC DQ[7:0] tDVW tDVW D0 D0 tQH tQH tDQSQ Figure 39 tDQSQ Read Status including tWHR and tCAD timing requirements, continued 87 4.4. Command Examples 4.4.1. Asynchronous This section shows examples of commands using the asynchronous data interface. Figure 40 and Figure 41 show an example of Change Read Column. A CE_n CLE ALE WE_n RE_n DQx Figure 40 05h C1 C2 E0h Asynchronous Data Interface: Change Read Column example 88 A CE_n CLE ALE WE_n tCCS RE_n DQx E0h Figure 41 D0 D1 D2 D3 Asynchronous Data Interface: Change Read Column example, continued 89 Figure 42 shows an example of Change Write Column. CE_n CLE ALE WE_n tCCS RE_n DQx 85h C1 Figure 42 C2 D0 D1 Asynchronous Data Interface: Change Write Column example 90 D2 D3 4.4.2. Source Synchronous This section shows examples of commands using the source synchronous data interface. Figure 43 through Figure 45 show an example of Change Read Column. Figure 45 shows a continuation of data transfer, after the host stops the transfer for a period of time. 91 A CLK CE_n CLE ALE W/R_n DQ[7:0] tCAD 05h tCAD C1 tCAD C2 E0h DQS DQ, DQS driven by host Figure 43 Source Synchronous Data Interface: Change Read Column example, command issue 92 A B tCCS CLK CE_n CLE ALE W/R_n DQ[7:0] E0h D0 D1 D2 D3 DQS DQ, DQS driven by host Figure 44 tDQSD turnaround on DQ, DQS DQ, DQS driven by device tDQSHZ turn-around on DQ, DQS Source Synchronous Data Interface: Change Read Column example, data transfer 93 DQ, DQS driven by host B CLK CE_n CLE ALE W/R_n DQ[7:0] D0 D1 D2 D3 D4 D5 DQS DQ, DQS driven by device Figure 45 tDQSHZ turn-around on DQ, DQS DQ, DQS driven by host Source Synchronous Data Interface: Change Read Column example, data transfer continue 94 Figure 46 shows an example of Change Write Column. tCCS CLK CE_n CLE ALE W/R_n tCAD DQ[7:0] 85h tCAD C1 C2 D0 D1 D2 D3 DQS DQ, DQS driven by host Figure 46 Source Synchronous Data Interface: Change Write Column example 95 5. Command Definition 5.1. Command Set Table 41 outlines the ONFI command set. The value specified in the first command cycle identifies the command to be performed. Some commands have a second command cycle as specified in Table 41. Typically, commands that have a second command cycle include an address. 96 Command O/M Read Multi-plane Copyback Read Change Read Column Change Read Column Enhanced Read Cache Random Read Cache Sequential Read Cache End Block Erase Multi-plane Read Status Read Status Enhanced Page Program Multi-plane Page Cache Program Copyback Program Multi-plane Small Data Move 2 Change Write Column 1 st nd 1 Cycle 2 Cycle M O O M O 00h 00h 00h 05h 06h 30h 32h 35h E0h E0h O O 00h 31h 31h O M O M O M O O O O O 3Fh 60h 60h 70h 78h 80h 80h 80h 85h 85h 85h M 85h Acceptable while Accessed LUN is Busy Acceptable while Other LUNs are Busy Y Y Y Y Y Target level commands Y Y D0h D1h Y Y 10h 11h 15h 10h 11h 11h Y Y Y Y Y Y Y Y Y Y Y Y 1 O 85h Y Change Row Address Read ID M 90h Y Read Parameter Page M ECh Y Read Unique ID O EDh Y Get Features O EEh Y Set Features O EFh Y Reset LUN O FAh Y Y Synchronous Reset O FCh Y Y Y Reset M FFh Y Y Y NOTE: 1. Change Write Column specifies the column address only. Change Row Address specifies the row address and the column address. Refer to the specific command definitions. 2. Small Data Move’s first opcode may be 80h if the operation is a program only with no data output. For the last second cycle of a Small Data Move, it is a 10h command to confirm the Program or Copyback operation. Table 41 Command set Reserved opcodes shall not be used by the device, as the ONFI specification may define the use of these opcodes in a future revision. Vendor specific opcodes may be used at the discretion of the vendor and shall never be defined for standard use by ONFI. Future Standardization opcodes are those opcodes already being used commonly in the industry and may be defined for standard use by ONFI for those same purposes. Future Standardization opcodes may be used by compliant ONFI implementations with the common industry usage. Block abstracted NAND opcodes are opcodes used in a BA NAND implementation. 97 Type Vendor Specific Future Standardization Block Abstracted NAND Reserved Opcode 02h – 04h, 08h, 16h – 17h, 19h, 1Dh, 20h – 22h, 25h – 29h, 2Bh, 2Dh – 2Fh, 33h, 36h – 39h, 3B – 3Eh, 40h – 41h, 48h, 4Ch, 53h – 55h, 68h, 72h – 75h, 84h, 87h – 89h, 91h – BFh, CFh, F1-F4h 23h – 24h, 2Ah, 2Ch, 34h, 3Ah, 65h, 71h, 79h – 7Bh, 81h, 8Ch C0h - CEh 01h, 07h, 09h – 0Fh, 12h-14h, 18h, 1Ah – 1Ch, 1Eh – 1Fh, 42h – 47h, 49h – 4Bh, 4Dh – 52h, 56h – 5Fh, 62h – 64h, 66h – 67h, 69h – 6Fh, 76h – 77h, 7Ch – 7Fh, 82h – 83h, 86h, 8Ah – 8Bh, 8Dh – 8Fh, D2h –DFh, E1h – EBh, F0h, F5h – F9h, FBh, FDh – FEh Table 42 Opcode Reservations 98 5.2. Command Descriptions The command descriptions in section 5 are shown in a format that is agnostic to the data interface being used (when the command may be used in either data interface). An example of the agnostic command description for Change Write Column is shown in Figure 47. The agnostic command examples shall be translated to a command description for the particular data interface selected. The command description for Change Write Column in the asynchronous data interface is shown in Figure 48. The command description for Change Write Column in the source synchronous data interface is shown in Figure 49. Note that the timing parameters defined in section 4 shall be observed for each command (e.g. the tCAD timing parameter for the source synchronous data interface). Cycle Type CMD ADDR ADDR DIN DIN Dk Dk+1 tCCS DQx 85h C1 C2 SR[6] Figure 47 99 Agnostic command description CE_n CLE ALE WE_n tCCS RE_n IOx 85h C1 Figure 48 100 C2 D0 Asynchronous data interface command description D1 tCCS CLK CE_n CLE ALE W/R_n tCAD DQ[7:0] 85h tCAD C1 C2 DQS Figure 49 Source synchronous data interface command description 101 D0 D1 5.3. Reset Definition The Reset function puts the target in its default power-up state and places the target in the asynchronous data interface. This command shall only be issued when the host is configured to the asynchronous data interface. The device shall also recognize and execute the Reset command when it is configured to the source synchronous data interface. The R/B_n value is unknown when Reset is issued; R/B_n is guaranteed to be low tWB after the Reset is issued. Note that some feature settings are retained across Reset commands (as specified in section 5.26). As part of the Reset command, all LUNs are also reset. The command may be executed with the target in any state, except during power-on when Reset shall not be issued until R/B_n is set to one. Figure 50 defines the Reset behavior and timings. CLE ALE WE_n RE_n IOx FFh DQS tWB tRST R/B_n Figure 50 Reset timing diagram 5.4. Synchronous Reset Definition The Synchronous Reset command resets the target and all LUNs. The command may be executed with the target in any state. Figure 51 defines the Synchronous Reset behavior and 102 timings. The R/B_n value is unknown when Synchronous Reset is issued; R/B_n is guaranteed to be low tWB after the Synchronous Reset is issued. This command shall be supported by devices that support the source synchronous data interface. This command is only accepted in source synchronous operation. The host should not issue this command when the device is configured to the asynchronous data interface. The target shall remain in the source synchronous data interface following this command. CLK CLE ALE W/R_n DQ[7:0] FCh DQS tWB tRST R/B_n Figure 51 Synchronous Reset timing diagram 5.5. Reset LUN Definition The Reset LUN command is used to reset a particular LUN. This command is accepted by only the LUN addressed as part of the command. The command may be executed with the LUN in any state. Figure 52 defines the Reset LUN behavior and timings. The SR[6] value is unknown when Reset LUN is issued; SR[6] is guaranteed to be low tWB after the Reset LUN command is issued. This command does not affect the data interface configuration for the target. Reset LUN should be used to cancel ongoing command operations, if desired. When there are issues with the target, e.g. a hang condition, the Reset (FFh) or Synchronous Reset (FCh) commands should be used. 103 Cycle Type CMD ADDR ADDR ADDR DQ[7:0] FAh R1 R2 R3 tWB tRST SR[6] Figure 52 Reset LUN timing diagram 5.6. Read ID Definition The Read ID function identifies that the target supports the ONFI specification. If the target supports the ONFI specification, then the ONFI signature shall be returned. The ONFI signature is the ASCII encoding of ‘ONFI’ where ‘O’ = 4Fh, ‘N’ = 4Eh, ‘F’ = 46h, and ‘I’ = 49h. Reading beyond four bytes yields indeterminate values. Figure 53 defines the Read ID behavior and timings. When issuing Read ID in the source synchronous data interface, each data byte is received twice. The host shall only latch one copy of each data byte. See section 4.3.2.5. For the Read ID command, only addresses of 00h and 20h are valid. To retrieve the ONFI signature an address of 20h shall be entered (i.e. it is not valid to enter an address of 00h and read 36 bytes to get the ONFI signature). Cycle Type CMD ADDR DOUT DOUT DOUT DOUT 4Fh 4Eh 46h 49h tWHR DQ[7:0] 90h 20h R/B_n Figure 53 Read ID timing diagram for ONFI signature The Read ID function can also be used to determine the JEDEC manufacturer ID and the device ID for the particular NAND part by specifying an address of 00h. Figure 54 defines the Read ID behavior and timings for retrieving the JEDEC manufacturer ID and device ID. Reading beyond the first two bytes yields values as specified by the manufacturer. 104 Cycle Type CMD ADDR DOUT DOUT MID DID tWHR DQ[7:0] 90h 00h R/B_n Figure 54 Read ID timing diagram for manufacturer ID MID Manufacturer ID for manufacturer of the part, assigned by JEDEC. DID Device ID for the part, assigned by the manufacturer. The Read ID command may be issued using either the asynchronous or source synchronous data interfaces. The timing parameters for each data interface are shown in Figure 55 and Figure 56. Figure 55 Read ID command using asynchronous data interface 105 Figure 56 Read ID command using source synchronous data interface NOTE: The data bytes in Figure 56 are repeated twice (on the rising and falling edge of CLK). 5.7. Read Parameter Page Definition The Read Parameter Page function retrieves the data structure that describes the target’s organization, features, timings and other behavioral parameters. There may also be additional information provided in an extended parameter page. Figure 57 defines the Read Parameter Page behavior. Values in the parameter page are static and shall not change. The host is not required to read the parameter page after power management events. The first time the host executes the Read Parameter Page command after power-on, timing mode 0 shall be used. If the host determines that the target supports more advanced timing modes, those supported timing modes may be used for subsequent execution of the Read Parameter Page command. The Change Read Column command may be issued following execution of the Read Parameter Page to read specific portions of the parameter page. Read Status may be used to check the status of Read Parameter Page during execution. After completion of the Read Status command, 00h shall be issued by the host on the command line to continue with the data output flow for the Read Parameter Page command. Read Status Enhanced and Change Read Column Enhanced shall not be used during execution of the Read Parameter Page command. 106 A Cycle Type CMD ADDR DOUT DOUT DOUT DOUT DOUT DOUT DQ[7:0] ECh 00h P00 P10 ... P01 P11 ... tRR tWB tR R/B_n A Cycle Type DQ[7:0] DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT P0n P1n ... E00 E10 ... E01 E11 ... R/B_n Figure 57 Read Parameter Page command timing P0k-Pnk The kth copy of the parameter page data structure. See section 5.7.1. Reading bytes beyond the end of the final parameter page copy (or beyond the final extended parameter page copy if supported) returns indeterminate values. E0k-Enk The kth copy of the extended parameter page data structure. See section 5.7.2. Reading bytes beyond the end of the final extended parameter page copy returns indeterminate values. This field is only present when the extended parameter page is supported, as indicated in the Features supported field of the parameter page. 107 5.7.1. Parameter Page Data Structure Definition Table 43 defines the parameter page data structure. For parameters that span multiple bytes, the least significant byte of the parameter corresponds to the first byte. See section 1.4.2.3 for more information on the representation of word and Dword values. Values are reported in the parameter page in units of bytes when referring to items related to the size of data access (as in an 8-bit data access device). For example, the target will return how many data bytes are in a page. For a device that supports 16-bit data access, the host is required to convert byte values to word values for its use. Unused fields should be cleared to 0h by the target. Byte 0-3 4-5 6-7 8-9 O/M Description Revision information and features block M Parameter page signature Byte 0: 4Fh, “O” Byte 1: 4Eh, “N” Byte 2: 46h, “F” Byte 3: 49h, “I” M Revision number 6-15 Reserved (0) 5 1 = supports ONFI version 2.3 4 1 = supports ONFI version 2.2 3 1 = supports ONFI version 2.1 2 1 = supports ONFI version 2.0 1 1 = supports ONFI version 1.0 0 Reserved (0) M Features supported 10-15 Reserved (0) 9 1 = supports EZ NAND 8 1 = supports program page register clear enhancement 7 1 = supports extended parameter page 6 1 = supports multi-plane read operations 5 1 = supports source synchronous 4 1 = supports odd to even page Copyback 3 1 = supports multi-plane program and erase operations 2 1 = supports non-sequential page programming 1 1 = supports multiple LUN operations 0 1 = supports 16-bit data bus width M Optional commands supported 10-15 Reserved (0) 9 1 = supports Reset LUN 8 1 = supports Small Data Move 7 1 = supports Change Row Address 6 1 = supports Change Read Column Enhanced 5 1 = supports Read Unique ID 4 1 = supports Copyback 3 1 = supports Read Status Enhanced 2 1 = supports Get Features and Set Features 1 1 = supports Read Cache commands 108 Byte 10-11 12-13 14 15-31 32-43 44-63 64 65-66 67-79 O/M O O Description 0 1 = supports Page Cache Program command Reserved (0) Extended parameter page length Number of parameter pages Reserved (0) Manufacturer information block M Device manufacturer (12 ASCII characters) M Device model (20 ASCII characters) M JEDEC manufacturer ID O Date code Reserved (0) 116-127 Memory organization block M Number of data bytes per page M Number of spare bytes per page Obsolete – Number of data bytes per partial page Obsolete – Number of spare bytes per partial page M Number of pages per block M Number of blocks per logical unit (LUN) M Number of logical units (LUNs) M Number of address cycles 4-7 Column address cycles 0-3 Row address cycles M Number of bits per cell M Bad blocks maximum per LUN M Block endurance M Guaranteed valid blocks at beginning of target M Block endurance for guaranteed valid blocks M Number of programs per page Obsolete – Partial programming attributes M Number of bits ECC correctability M Number of plane address bits 4-7 Reserved (0) 0-3 Number of plane address bits O Multi-plane operation attributes 6-7 Reserved (0) 5 1 = lower bit XNOR block address restriction 4 1 = read cache supported 3 Address restrictions for cache operations 2 1 = program cache supported 1 1 = no block address restrictions 0 Overlapped / concurrent multi-plane support O EZ NAND support 3-7 Reserved (0) 2 1 = Requires Copyback Adjacency 1 1 = supports Copyback for other planes & LUNs 0 1 = supports enable/disable of automatic retries Reserved (0) 128 Electrical parameters block M I/O pin capacitance, maximum 80-83 84-85 86-89 90-91 92-95 96-99 100 101 102 103-104 105-106 107 108-109 110 111 112 113 114 115 109 Byte 129-130 O/M Description M Asynchronous timing mode support 6-15 Reserved (0) 5 1 = supports timing mode 5 4 1 = supports timing mode 4 3 1 = supports timing mode 3 2 1 = supports timing mode 2 1 1 = supports timing mode 1 0 1 = supports timing mode 0, shall be 1 Obsolete – Asynchronous program cache timing mode support 133-134 M 135-136 M 137-138 M 139-140 M tPROG Maximum page program time (µs) tBERS Maximum block erase time (µs) tR Maximum page read time (µs) tCCS Minimum change column setup time (ns) 141-142 O 143 O 144-145 O Source synchronous timing mode support 6-15 Reserved (0) 5 1 = supports timing mode 5 4 1 = supports timing mode 4 3 1 = supports timing mode 3 2 1 = supports timing mode 2 1 1 = supports timing mode 1 0 1 = supports timing mode 0 Source synchronous features 3-7 Reserved (0) 2 1 = device supports CLK stopped for data input 1 1 = typical capacitance values present 0 tCAD value to use CLK input pin capacitance, typical 146-147 O I/O pin capacitance, typical 148-149 O Input pin capacitance, typical 150 M Input pin capacitance, maximum 151 M 152-153 O 154-155 O Driver strength support 3-7 Reserved (0) 2 1 = supports 18 Ohm drive strength 1 1 = supports 25 Ohm drive strength 0 1 = supports driver strength settings tR Maximum multi-plane page read time (µs) tADL Program page register clear enhancement tADL value (ns) 156-157 O tR Typical page read time for EZ NAND (µs) 131-132 158-163 Reserved (0) 164-165 166-253 254-255 Vendor block M Vendor specific Revision number Vendor specific M Integrity CRC 256-511 512-767 768+ Redundant Parameter Pages M Value of bytes 0-255 M Value of bytes 0-255 O Additional redundant parameter pages 110 Table 43 Parameter page definitions 5.7.1.1. Byte 0-3: Parameter page signature This field contains the parameter page signature. When two or more bytes of the signature are valid, then it denotes that a valid copy of the parameter page is present. Byte 0 shall be set to 4Fh. Byte 1 shall be set to 4Eh. Byte 2 shall be set to 46h. Byte 3 shall be set to 49h. 5.7.1.2. Byte 4-5: Revision number This field indicates the revisions of the ONFI specification that the target complies to. The target may support multiple revisions of the ONFI specification. This is a bit field where each defined bit corresponds to a particular specification revision that the target may support. Bit 0 shall be cleared to zero. Bit 1 when set to one indicates that the target supports the ONFI revision 1.0 specification. Bit 2 when set to one indicates that the target supports the ONFI revision 2.0 specification. Bit 3 when set to one indicates that the target supports the ONFI revision 2.1 specification. Bit 4 when set to one indicates that the target supports the ONFI revision 2.2 specification. Bit 5 when set to one indicates that the target supports the ONFI revision 2.3 specification. Bits 6-15 are reserved and shall be cleared to zero. 5.7.1.3. Byte 6-7: Features supported This field indicates the optional features that the target supports. Bit 0 when set to one indicates that the target’s data bus width is 16-bits. Bit 0 when cleared to zero indicates that the target’s data bus width is 8-bits. The host shall use the indicated data bus width for all ONFI commands that are defined to be transferred at the bus width (x8 or x16). Note that some commands, like Read ID, always transfer data as 8-bit only. If the source synchronous data interface is supported, then the data bus width shall be 8-bits. Bit 1 when set to one indicates that the target supports multiple LUN operations (see section 3.1.3). If bit 1 is cleared to zero, then the host shall not issue commands to a LUN unless all other LUNs on the Target are idle (i.e. R/B_n is set to one). Bit 2 when set to one indicates that the target supports non-sequential page programming operations, such that the host may program pages within a block in arbitrary order. Bit 2 when cleared to zero indicates that the target does not support non-sequential page programming operations. If bit 2 is cleared to zero, the host shall program all pages within a block in order starting with page 0. 111 Bit 3 when set to one indicates that the target supports multi-plane program and erase operations. Refer to section 5.7.1.28. Bit 4 when set to one indicates that there are no even / odd page restrictions for Copyback operations. Specifically, a read operation may access an odd page and then program the contents to an even page using Copyback. Alternatively, a read operation may access an even page and then program the contents to an odd page using Copyback. Bit 4 when cleared to zero indicates that the host shall ensure that Copyback reads and programs from odd page to odd page or alternatively from even page to even page. Bit 5 when set to one indicates that the source synchronous data interface is supported by the target. If bit 5 is set to one, then the target shall indicate the source synchronous timing modes supported in the source synchronous timing mode support field. Bit 5 when cleared to zero indicates that the source synchronous data interface is not supported by the target. Bit 6 when set to one indicates that the target supports multi-plane read operations. Refer to section 5.7.1.28. Bit 7 when set to one indicates the target includes an extended parameter page that is stored in the data bytes following the last copy of the parameter page. If bit 7 is cleared to zero, then an extended parameter page is not supported. Refer to section 5.7.2. This bit shall be cleared to zero for devices that support EZ NAND. Note: This bit was inadvertently specified in the BA NAND specification to show support for BA NAND. If the device supports BA NAND, then the number of bits of ECC correctability should be cleared to 0h in byte 112 of the parameter page. Bit 8 when set to one indicates that the target supports clearing only the page register for the LUN addressed with the Program (80h) command. If bit 8 is cleared to zero, then a Program (80h) command clears the page register for each LUN that is part of the target. At power-on, the device clears the page register for each LUN that is part of the target. Refer to section 5.26.1 for how to enable this feature. Bit 9 when set to one indicates that the target supports EZ NAND. If bit 9 is cleared to zero, then the target does not support EZ NAND and is configured as raw NAND. Bits 10-15 are reserved and shall be cleared to zero. 5.7.1.4. Byte 8-9: Optional commands supported This field indicates the optional commands that the target supports. Bit 0 when set to one indicates that the target supports the Page Cache Program command. If bit 0 is cleared to zero, the host shall not issue the Page Cache Program command to the target. For targets that support EZ NAND, this field shall be cleared to zero indicating that the target does not support the Page Cache Program command. Bit 1 when set to one indicates that the target supports the Read Cache Random, Read Cache Sequential, and Read Cache End commands. If bit 1 is cleared to zero, the host shall not issue the Read Cache Sequential, Read Cache Random, or Read Cache End commands to the target. For targets that support EZ NAND, this field shall be cleared to zero indicating that the target does not support Read Cache Random, Read Cache Sequential, or Read Cache End commands. Bit 2 when set to one indicates that the target supports the Get Features and Set Features commands. If bit 2 is cleared to zero, the host shall not issue the Get Features or Set Features commands to the target. 112 Bit 3 when set to one indicates that the target supports the Read Status Enhanced command. If bit 3 is cleared to zero, the host shall not issue the Read Status Enhanced command to the target. Read Status Enhanced shall be supported if the target has multiple LUNs or supports multi-plane operations. Bit 4 when set to one indicates that the target supports the Copyback Program and Copyback Read commands. If bit 4 is cleared to zero, the host shall not issue the Copyback Program or Copyback Read commands to the target. If multi-plane operations are supported and this bit is set to one, then multi-plane copyback operations shall be supported. Bit 5 when set to one indicates that the target supports the Read Unique ID command. If bit 5 is cleared to zero, the host shall not issue the Read Unique ID command to the target. Bit 6 when set to one indicates that the target supports the Change Read Column Enhanced command. If bit 6 is cleared to zero, the host shall not issue the Change Read Column Enhanced command to the target. Bit 7 when set to one indicates that the target supports the Change Row Address command. If bit 7 is cleared to zero, the host shall not issue the Change Row Address command to the target. Bit 8 when set to one indicates that the target supports the Small Data Move command for both Program and Copyback operations. If bit 8 is cleared to zero, the target does not support the Small Data Move command for Program or Copyback operations. The Small Data Move command is mutually exclusive with overlapped multi-plane support. Refer to section 5.19. When bit 8 is set to one, the device shall support the 11h command to flush any internal data pipeline regardless of whether multi-plane operations are supported. Bit 9 when set to one indicates that the target supports the Reset LUN command. If bit 9 is cleared to zero, the host shall not issue the Reset LUN command. Bits 10-15 are reserved and shall be cleared to zero. 5.7.1.5. Byte 12-13: Extended parameter page length If the target supports an extended parameter page as indicated in the Features supported field, then this field specifies the length of the extended parameter page in multiples of 16 bytes. Thus, a value of 2 corresponds to 32 bytes and a value of 3 corresponds to 48 bytes. The minimum size is 3, corresponding to 48 bytes. 5.7.1.6. Byte 14: Number of parameter pages If the target supports an extended parameter page as indicated in the Features supported field, then this field specifies the number of parameter pages present, including the original and the subsequent redundant versions. As an example, a value of 3 means that there are three parameter pages present and thus the extended parameter page starts at byte 768. The number of extended parameter pages should match the number of parameter pages. 5.7.1.7. Byte 32-43: Device manufacturer This field contains the manufacturer of the device. The content of this field is an ASCII character string of twelve bytes. The device shall pad the character string with spaces (20h), if necessary, to ensure that the string is the proper length. There is no standard for how the manufacturer represents their name in the ASCII string. If the host requires use of a standard manufacturer ID, it should use the JEDEC manufacturer ID (refer to section 5.7.1.9). 113 5.7.1.8. Byte 44-63: Device model This field contains the model number of the device. The content of this field is an ASCII character string of twenty bytes. The device shall pad the character string with spaces (20h), if necessary, to ensure that the string is the proper length. 5.7.1.9. Byte 64: JEDEC manufacturer ID This field contains the JEDEC manufacturer ID for the manufacturer of the device. 5.7.1.10. Byte 65-66: Date code This field contains a date code for the time of manufacture of the device. Byte 65 shall contain the two least significant digits of the year (e.g. a value of 05h to represent the year 2005). Byte 66 shall contain the workweek, where a value of 00h indicates the first week of January. If the date code functionality is not implemented, the value in this field shall be 0000h. 5.7.1.11. Byte 80-83: Number of data bytes per page This field contains the number of data bytes per page. The value reported in this field shall be a power of two. The minimum value that shall be reported is 512 bytes. 5.7.1.12. Byte 84-85: Number of spare bytes per page This field contains the number of spare bytes per page. There are no restrictions on the value. Appendix B lists recommendations for the number of bytes per page based on the page size and the number of bits of ECC correctability for the device. 5.7.1.13. Byte 86-89: Obsolete – Number of data bytes per partial page This field is obsolete. It previously contained the number of data bytes per partial page. 5.7.1.14. Byte 90-91: Obsolete – Number of spare bytes per partial page This field is obsolete. It previously contained the number of spare bytes per partial page. 5.7.1.15. Byte 92-95: Number of pages per block This field contains the number of pages per block. This value shall be a multiple of 32. Refer to section 3.1 for addressing requirements. 5.7.1.16. Byte 96-99: Number of blocks per logical unit This field contains the number of blocks per logical unit. There are no restrictions on this value. Refer to section 3.1 for addressing requirements. 5.7.1.17. Byte 100: Number of logical units (LUNs) This field indicates the number of logical units the target supports. Logical unit numbers are sequential, beginning with a LUN address of 0. This field shall be greater than zero. 114 5.7.1.18. Byte 101: Number of Address Cycles This field indicates the number of address cycles used for row and column addresses. The reported number of address cycles shall be used by the host in operations that require row and/or column addresses (e.g. Page Program). Bits 0-3 indicate the number of address cycles used for the row address. This field shall be greater than zero. Bits 4-7 indicate the number of address cycles used for the column address. This field shall be greater than zero. NOTE: Throughout this specification examples are shown with 2-byte column addresses and 3byte row addresses. However, the host is responsible for providing the number of column and row address cycles in each of these sequences based on the values in this field. 5.7.1.19. Byte 102: Number of bits per cell This field indicates the number of bits per cell in the Flash array. This field shall be greater than zero. For some devices, including implementations supporting EZ NAND, the device may be constructed of Flash arrays of different types. A value of FFh indicates that the number of bits per cell is not specified. 5.7.1.20. Byte 103-104: Bad blocks maximum per LUN This field contains the maximum number of blocks that may be defective at manufacture and over the life of the device per LUN. The maximum rating assumes that the host is following the block endurance requirements and the ECC requirements reported in the parameter page. 5.7.1.21. Byte 105-106: Block endurance This field indicates the maximum number of program/erase cycles per addressable page/block. This value assumes that the host is using at least the minimum ECC correctability reported in the parameter page. A page may be programmed in partial operations subject to the value reported in the Number of programs per page field. However, programming different locations within the same page does not count against this value more than once per full page. The block endurance is reported in terms of a value and a multiplier according to the following multiplier equation: value x 10 . Byte 105 comprises the value. Byte 106 comprises the multiplier. For example, a target with an endurance of 75,000 cycles would report this as a value of 75 and a 3 multiplier of 3 (75 x 10 ). For a write once device, the target shall report a value of 1 and a multiplier of 0. For a read-only device, the target shall report a value of 0 and a multiplier of 0. The value field shall be the smallest possible; for example 100,000 shall be reported as a value of 5 1 and a multiplier of 5 (1 x 10 ). 5.7.1.22. Byte 107: Guaranteed valid blocks at beginning of target This field indicates the number of guaranteed valid blocks starting at block address 0 of the target. The minimum value for this field is 1h. The blocks are guaranteed to be valid for the endurance specified for this area (see section 5.7.1.23) when the host follows the specified number of bits to correct. 115 5.7.1.23. Byte 108-109: Block endurance for guaranteed valid blocks This field indicates the minimum number of program/erase cycles per addressable page/block in the guaranteed valid block area (see section 5.7.1.22). This value requires that the host is using at least the minimum ECC correctability reported in the parameter page. This value is not encoded. If the value is 0000h, then no minimum number of cycles is specified, though the block(s) are guaranteed valid from the factory. 5.7.1.24. Byte 110: Number of programs per page This field indicates the maximum number of times a portion of a page may be programmed without an erase operation. After the number of programming operations specified have been performed, the host shall issue an erase operation to that block before further program operations to the affected page. This field shall be greater than zero. Programming the same portion of a page without an erase operation results in indeterminate page contents. 5.7.1.25. Byte 111: Obsolete – Partial programming attributes This field is obsolete. It previously indicated the attributes for partial page programming that the target supports. 5.7.1.26. Byte 112: Number of bits ECC correctability This field indicates the number of bits that the host should be able to correct per 512 bytes of data. With this specified amount of error correction by the host, the target shall achieve the block endurance specified in the parameter page. When the specified amount of error correction is applied by the host and the block endurance is followed, then the maximum number of bad blocks shall not be exceeded by the device. All used bytes in the page shall be protected by host controller ECC including the spare bytes if the minimum ECC requirement has a value greater than zero. If the recommended ECC codeword size is not 512 bytes, then this field shall be set to FFh. The host should then read the Extended ECC Information that is part of the extended parameter page to retrieve the ECC requirements for this device. When this value is cleared to zero, the target shall return valid data. For targets that support EZ NAND, this field shall be cleared to zero indicating that the target returns corrected data to the host. 5.7.1.27. Byte 113: Multi-plane addressing This field describes parameters for multi-plane addressing. Bits 0-3 indicate the number of bits that are used for multi-plane addressing. This value shall be greater than 0h when multi-plane operations are supported. For information on the plane address location, refer to section 3.1.1. Bits 4-7 are reserved. 5.7.1.28. Byte 114: Multi-plane operation attributes This field describes attributes for multi-plane operations. This byte is mandatory when multiplane operations are supported as indicated in the Features supported field. Bit 0 indicates whether overlapped multi-plane operations are supported. If bit 0 is set to one, then overlapped multi-plane operations are supported. If bit 0 is cleared to zero, then concurrent multi-plane operations are supported. 116 Bit 1 indicates that there are no block address restrictions for the multi-plane operation. If set to one all block address bits may be different between multi-plane operations. If cleared to zero, there are block address restrictions. Refer to bit 5 for the specific block address restrictions required. Bit 2 indicates whether program cache is supported with multi-plane programs. If set to one then program cache is supported for multi-plane program operations. If cleared to zero then program cache is not supported for multi-plane program operations. Note that program cache shall not be used with multi-plane copyback program operations. See bit 3 for restrictions on the multi-plane addresses that may be used. Bit 3 indicates whether the block address bits other than the multi-plane address bits of multiplane addresses may change during either: a) a program cache sequence between 15h commands, or b) a read cache sequence between 31h commands. If set to one and bit 2 is set to one, then the host may change the number of multi-plane addresses and the value of the block address bits (other than the multi-plane address bits) in the program cache sequence. If set to one and bit 4 is set to one, then the host may change the number of multi-plane addresses and the value of the block address bits (other than the multi-plane address bits) in the read cache sequence. If cleared to zero and bit 2 is set to one, then for each program cache operation the block address bits (other than the plane address bits) and number of multi-plane addresses issued to the LUN shall be the same. If cleared to zero and bit 4 is set to one, then for each read cache operation the block address bits (other than the multi-plane address bits) and number of multi-plane addresses issued to the LUN shall be the same. Bit 4 indicates whether read cache is supported with multi-plane reads. If set to one then read cache is supported for multi-plane read operations. If cleared to zero then read cache is not supported for multi-plane read operations. Note that read cache shall not be used with multiplane copyback read operations. Bit 5 indicates the type of block address restrictions required for the multi-plane operation. If set to one then all block address bits (other than the multi-plane address bits) shall be the same if the XNOR of the lower multi-plane address bits between two multi-plane addresses is one. If cleared to zero, all block address bits (other than the multi-plane address bits) shall be the same regardless of the multi-plane address bits between two plane addresses. See section 3.1.1.1 for a detailed definition of interleaved block address restrictions. These restrictions apply to all multiplane operations (Read, Program, Erase, and Copyback Program). Bits 6-7 are reserved. 5.7.1.1. Byte 115: EZ NAND support This field describes the EZ NAND attributes that the target supports. This field is only used if EZ NAND is supported by the target. Bit 0 indicates whether the target supports automatic retries being enabled and disabled explicitly by the host using Set Features. If this bit is set to one, then the host may enable or disable automatic retries using Set Features. If this bit is cleared to zero, then the EZ NAND controller determines whether to perform a retry without host intervention. If the EZ NAND controller executes an automatic retry, the typical page read time (tR) may be exceeded. Bit 1 indicates whether Copyback is supported with destination planes or LUNs that are different from the source. If this bit is set to one, then the plane or LUN of the destination for the Copyback may be different than the source. If this bit is cleared to zero, then the plane or LUN of the destination for a Copyback shall be the same as the source. 117 Bit 2 indicates whether the target requires Copyback Read and Copyback Program command adjacency. If this bit is set to one, then each Copyback Read command issued shall be followed by a Copyback Program command prior to any additional Copyback Read command being issued to the target. I.e., A Copyback Read command shall be explicitly paired with aCopyback Program command). If this bit is cleared to zero, then there may be multiple Copyback Read commands prior to issuing a Copyback Program command. For ONFI 2.3, Copyback Adjacency is required and this bit shall be set to one. Bits 3-7 are reserved. 5.7.1.2. Byte 128: I/O pin capacitance, maximum This field indicates the maximum I/O pin capacitance for the target in pF. This may be used by the host to calculate the load for the data bus. Refer to section 2.11. 5.7.1.3. Byte 129-130: Asynchronous timing mode support This field indicates the asynchronous timing modes supported. The target shall always support asynchronous timing mode 0. Bit 0 shall be set to one. It indicates that the target supports asynchronous timing mode 0. Bit 1 when set to one indicates that the target supports asynchronous timing mode 1. Bit 2 when set to one indicates that the target supports asynchronous timing mode 2. Bit 3 when set to one indicates that the target supports asynchronous timing mode 3. Bit 4 when set to one indicates that the target supports asynchronous timing mode 4. Bit 5 when set to one indicates that the target supports asynchronous timing mode 5. Bits 6-15 are reserved and shall be cleared to zero. 5.7.1.4. Byte 131-132: Obsolete – Asynchronous program cache timing mode support This field is obsolete. It previously indicated the asynchronous timing modes supported for Page Cache Program operations. 5.7.1.5. Byte 133-134: Maximum page program time This field indicates the maximum page program time (tPROG) in microseconds. 5.7.1.6. Byte 135-136: Maximum block erase time This field indicates the maximum block erase time (tBERS) in microseconds. 5.7.1.7. Byte 137-138: Maximum page read time This field indicates the maximum page read time (tR) in microseconds. For devices that support EZ NAND, this value is the tR maximum at the end of life and is related to the uncorrectable bit error rate (UBER) specified for the device. 118 5.7.1.8. Byte 139-140: Minimum change column setup time This field indicates the minimum change column setup time (tCCS) in nanoseconds. After issuing a Change Read Column command, the host shall not read data until a minimum of tCCS time has elapsed. After issuing a Change Write Column command including all column address cycles, the host shall not write data until a minimum of tCCS time has elapsed. The value of tCCS shall always be longer than or equal to tWHR and tADL when the source synchronous data interface is supported. 5.7.1.9. Byte 141-142: Source synchronous timing mode support This field indicates the source synchronous timing modes supported. If the source synchronous data interface is supported by the target, at least one source synchronous timing mode shall be supported. The target shall support an inclusive range of source synchronous timing modes (i.e. if timing mode n-1 and n+1 are supported, then the target shall also support timing mode n). Bit 0 when set to one indicates that the target supports source synchronous timing mode 0. Bit 1 when set to one indicates that the target supports source synchronous timing mode 1. Bit 2 when set to one indicates that the target supports source synchronous timing mode 2. Bit 3 when set to one indicates that the target supports source synchronous timing mode 3. Bit 4 when set to one indicates that the target supports source synchronous timing mode 4. Bit 5 when set to one indicates that the target supports source synchronous timing mode 5. Bits 6-15 are reserved and shall be cleared to zero. 5.7.1.10. Byte 143: Source synchronous features This field describes features and attributes for source synchronous operation. This byte is mandatory when the source synchronous data interface is supported. Bit 0 indicates the tCAD value that shall be used by the host. If bit 0 is set to one, then the host shall use the tCADs (slow) value in source synchronous command, address and data transfers. If bit 0 is cleared to zero, then the host shall use the tCADf (fast) value in source synchronous command, address and data transfers. Bit 1 indicates if the typical CLK, I/O and input pin capacitance values are reported in the parameter page. If bit 1 is set to one, then the typical CLK, I/O and input pin capacitance values are reported in the parameter page. If bit 1 is cleared to zero, then the typical capacitance fields are not used. Bit 2 indicates that the device supports the CLK being stopped during data input, as described in Figure 34. If bit 2 is set to one, then the host may optionally stop the CLK during data input for power savings. If bit 2 is set to one, the host may pause data while the CLK is stopped. If bit 2 is cleared to zero, then the host shall leave CLK running during data input. Bits 3-7 are reserved. 5.7.1.11. Byte 144-145: CLK input pin capacitance, typical This field indicates the typical CLK input pin capacitance for the target. This value applies to the CLK_t and CLK_c signals. This field is specified in 0.1 pF units. For example, a value of 31 corresponds to 3.1 pF. The variance from this value is less than +/- 0.5 pF per LUN per x8 data 119 bus. As an example, if two LUNs are present per x8 data bus than the total variance is less than +/- 1.0 pF. This value is only valid if the typical capacitance values are supported as indicated in the source synchronous features field. Additional constraints on the CLK input pin capacitance are specified in section 4.2.3. 5.7.1.12. Byte 146-147: I/O pin capacitance, typical This field indicates the typical I/O pin capacitance for the target. This field is specified in 0.1 pF units. For example, a value of 31 corresponds to 3.1 pF. The variance from this value is less than +/- 0.5 pF per LUN per x8 data bus. As an example, if two LUNs are present than the total variance is less than +/- 1 pF. This value is only valid if the typical capacitance values are supported as indicated in the source synchronous features field. Additional constraints on the I/O pin capacitance are specified in section 4.2.3. 5.7.1.13. Byte 148-149: Input pin capacitance, typical This field indicates the typical input pin capacitance for the target. This value applies to all inputs except the following: CLK_t, CLK_c, CE_n and WP_n signals. This field is specified in 0.1 pF units. For example, a value of 31 corresponds to 3.1 pF. The variance from this value is less than +/- 0.5 pF per LUN per x8 data bus. As an example, if two LUNs are present than the total variance is less than +/- 1 pF. This value is only valid if the typical capacitance values are supported as indicated in the source synchronous features field. Additional constraints on the input pin capacitance are specified in section 4.2.3. 5.7.1.14. Byte 150: Input pin capacitance, maximum This field indicates the maximum input pin capacitance for the target in pF. This value applies to all inputs, including CLK_t, CLK_c, CE_n, and WP_n. This may be used by the host to calculate the load for the data bus. Refer to section 2.11. 5.7.1.15. Byte 151: Driver strength support This field describes if the target supports configurable driver strengths and its associated features. Bit 0 when set to one indicates that the target supports configurable driver strength settings as defined in Table 26. If this bit is set to one, then the device shall support both the 35 Ohm and 50 Ohm settings. If this bit is set to one, then the device shall power-on with a driver strength at the 35 Ohm value defined in Table 26. If this bit is cleared to zero, then the driver strength at poweron is undefined. This bit shall be set to one for devices that support the source synchronous data interface. Bit 1 when set to one indicates that the target supports the 25 Ohm setting in Table 26 for use in the I/O Drive Strength setting. This bit shall be set to one for devices that support the source synchronous data interface. Bit 2 when set to one indicates that the target supports the 18 Ohm setting in Table 26 for use in the I/O Drive Strength setting. This bit shall be set to one for devices that support the source synchronous data interface. Bits 3-7 are reserved. 5.7.1.16. Byte 152-153: Maximum multi-plane page read time This field indicates the maximum page read time (tR) for multi-plane page reads in microseconds. Multi-plane page read times may be longer than non-multi-plane pages read times. This field 120 shall be supported if the target supports multi-plane reads as indicated in the Features supported field. 5.7.1.17. Byte 154-155: Program page register clear enhancement tADL value This field indicates the ALE to data loading time (tADL) in nanoseconds when the program page register clear enhancement is enabled. If the program page register clear enhancement is disabled, then the tADL value is as defined for the selected timing mode. This increased tADL value only applies to Program (80h) command sequences; it does not apply for Set Features, Copyback, or other commands. 5.7.1.1. Byte 156-157: Typical page read time for EZ NAND This field indicates the typical page read time (tR) in microseconds for devices that support EZ NAND. This field is not used for devices that do not support EZ NAND. For devices that include multiple bits per cell in the NAND used, this value is an average of the tR typical values for the pages (e.g. lower and upper pages). 5.7.1.2. Byte 164-165: Vendor specific Revision number This field indicates a vendor specific revision number. This field should be used by vendors to indicate the supported layout for the vendor specific parameter page area and the vendor specific feature addresses. The format of this field is vendor specific. 5.7.1.3. Byte 166-253: Vendor specific This field is reserved for vendor specific use. 5.7.1.4. Byte 254-255: Integrity CRC The Integrity CRC (Cyclic Redundancy Check) field is used to verify that the contents of the parameter page were transferred correctly to the host. The CRC of the parameter page is a word (16-bit) field. The CRC calculation covers all of data between byte 0 and byte 253 of the parameter page inclusive. The CRC shall be calculated on byte (8-bit) quantities starting with byte 0 in the parameter page. The bits in the 8-bit quantity are processed from the most significant bit (bit 7) to the least significant bit (bit 0). The CRC shall be calculated using the following 16-bit generator polynomial: G(X) = X16 + X15 + X2 + 1 This polynomial in hex may be represented as 8005h. The CRC value shall be initialized with a value of 4F4Eh before the calculation begins. There is no XOR applied to the final CRC value after it is calculated. There is no reversal of the data bytes or the CRC calculated value. 5.7.1.5. Byte 256-511: Redundant Parameter Page 1 This field shall contain the values of bytes 0-255 of the parameter page. Byte 256 is the value of byte 0. The redundant parameter page is used when the integrity CRC indicates that there was an error in bytes 0-255. The redundant parameter page shall be stored in non-volatile media; the target shall not create these bytes by retransmitting the first 256 bytes. 121 5.7.1.6. Byte 512-767: Redundant Parameter Page 2 This field shall contain the values of bytes 0-255 of the parameter page. Byte 512 is the value of byte 0. The redundant parameter page is used when the integrity CRC indicates that there was an error in bytes 0-255 and in the first redundant parameter page. The redundant parameter page shall be stored in non-volatile media; the target shall not create these bytes by retransmitting the first 256 bytes. 5.7.1.7. Byte 768+: Additional Redundant Parameter Pages Bytes at offset 768 and above may contain additional redundant copies of the parameter page. There is no limit to the number of redundant parameter pages that the target may provide. The target may provide additional copies to guard against the case where all three mandatory copies have invalid CRC checks. The host should determine whether an additional parameter page is present by checking the first Dword. If at least two out of four bytes match the parameter page signature, then an additional parameter page is present. 5.7.2. Extended Parameter Page Data Structure Definition The extended parameter page, if present, provides additional information to the host that there was insufficient space to include in the parameter page. The extended parameter page is organized in sections. Each section is a multiple of 16 bytes in length. The section types are specified in Table 44. Section Type 0 1 2 3-255 Section Definition Unused section marker. No section present. Section type and length specifiers. Extended ECC information. Reserved Table 44 Section Type Definitions Section types shall be specified in the extended parameter page in order (other than section type value 0). For example, if section type 12 and section type 15 were both present in the extended parameter page then section type 12 shall precede section type 15. There shall only be one instantiation of each section type. All unused sections shall be marked with a section type value of 0. When software encounters a section type value of 0, this marks the end of the valid sections. Table 45 defines the layout of section type 1. Section type 1 specifies additional sections when more than eight sections are present in the extended parameter page. The length of section type 1 shall be a multiple of 16 bytes. 122 Byte O/M 0 1 2 3 4 5 6 – (end) Description M M O O O O O Table 45 Section 8 type Section 8 length Section 9 type Section 9 length Section 10 type Section 10 length Section 11 – n type & lengths Section Type 1: Additional Section Type and Length Specifiers Table 46 defines the layout of section type 2. Section type 2 specifies extended ECC information. Each extended ECC information block is eight bytes in length. If an extended ECC information block is not specified, then all values in that block shall be cleared to 0h. The length of section type 2 shall be a multiple of 16 bytes. Byte 0-7 8-15 16 – (end) O/M M O O Table 46 Description Extended ECC information block 0 Extended ECC information block 1 Extended ECC information block 2 – n (if present) Section Type 2: Extended ECC Information The definition of the extended ECC information block is specified in section 3.3. Table 47 defines the extended parameter page data structure. For parameters that span multiple bytes, the least significant byte of the parameter corresponds to the first byte. See section 1.4.2.3 for more information on the representation of word and Dword values. Values are reported in the extended parameter page in units of bytes when referring to items related to the size of data access (as in an 8-bit data access device). For example, the target will return how many data bytes are in a page. For a device that supports 16-bit data access, the host is required to convert byte values to word values for its use. Unused fields should be cleared to 0h by the target. 123 Byte O/M 0-1 2-5 6-15 16 17 18 19 20 21 22-31 32 – (end) Description Revision information and features block M Integrity CRC M Extended parameter page signature Byte 0: 45h, “E” Byte 1: 50h, “P” Byte 2: 50h, “P” Byte 3: 53h, “S” Reserved (0) M Section 0 type M Section 0 length M Section 1 type M Section 1 length O Section 2 type O Section 2 length O Section 3 – 7 types & lengths M Section information Table 47 Extended Parameter Page definition 5.7.2.1. Byte 0-1: Integrity CRC The Integrity CRC (Cyclic Redundancy Check) field is used to verify that the contents of the extended parameter page were transferred correctly to the host. The CRC of the extended parameter page is a word (16-bit) field. The CRC calculation covers all of data between byte 2 and the end of the extended parameter page inclusive. The CRC shall be calculated on byte (8-bit) quantities starting with byte 2 in the extended parameter page to the end of the extended parameter page. The bits in the 8-bit quantity are processed from the most significant bit (bit 7) to the least significant bit (bit 0). The CRC shall be calculated using the following 16-bit generator polynomial: G(X) = X16 + X15 + X2 + 1 This polynomial in hex may be represented as 8005h. The CRC value shall be initialized with a value of 4F4Eh before the calculation begins. There is no XOR applied to the final CRC value after it is calculated. There is no reversal of the data bytes or the CRC calculated value. 5.7.2.2. Byte 2-5: Extended parameter page signature This field contains the extended parameter page signature. When two or more bytes of the signature are valid, then it denotes that a valid copy of the extended parameter page is present. Byte 2 shall be set to 45h. Byte 3 shall be set to 50h. Byte 4 shall be set to 50h. Byte 5 shall be set to 53h. 124 5.7.2.3. Byte 16: Section 0 type Section 0 is the first section in the extended parameter page and begins at byte offset 32. This field specifies the type of section 0. Section types are defined in Table 44. 5.7.2.4. Byte 17: Section 0 length Section 0 is the first section in the extended parameter page and begins at byte offset 32. This field specifies the length of section 0. The length is specified in multiples of 16 bytes. Thus, a value of 1 corresponds to 16 bytes and a value of 2 corresponds to 32 bytes. 5.7.2.5. Byte 18: Section 1 type Section 1 is the second section in the extended parameter page and starts immediately following section 0. This field specifies the type of section 1. Section types are defined in Table 44. If section 1 is not present, then the type field shall be cleared to 0. 5.7.2.6. Byte 19: Section 1 length Section 1 is the second section in the extended parameter page and starts immediately following section 0. This field specifies the length of section 1. The length is specified in multiples of 16 bytes. Thus, a value of 1 corresponds to 16 bytes and a value of 2 corresponds to 32 bytes. If section 1 is not present, then the length field shall be cleared to 0. 5.7.2.7. Byte 20: Section 2 type Section 2 is the third section in the extended parameter page and starts immediately following section 1. This field specifies the type of section 2. Section types are defined in Table 44. If section 2 is not present, then the type field shall be cleared to 0. 5.7.2.8. Byte 21: Section 2 length Section 2 is the third section in the extended parameter page and starts immediately following section 1. This field specifies the length of section 2. The length is specified in multiples of 16 bytes. Thus, a value of 1 corresponds to 16 bytes and a value of 2 corresponds to 32 bytes. If section 2 is not present, then the length field shall be cleared to 0. 5.7.2.9. Byte 22-31: Section 3 – 7 types and lengths Bytes 22-31 define the type and lengths for sections 3 – 7 in order, following the same definition and layout as section 0 and 1 type and length definitions. If a section is not present, then the type and length fields for that section shall be cleared to 0. 5.7.2.10. Byte 32 – (end): Section information Section 0 begins at byte offset 32 and is a multiple of 16 bytes. If there are additional sections (section 1, 2, 3, etc), each section starts immediately following the previous section and is a multiple of 16 bytes. 5.8. Read Unique ID Definition The Read Unique ID function is used to retrieve the 16 byte unique ID (UID) for the device. The unique ID when combined with the device manufacturer shall be unique. The UID data may be stored within the Flash array. To allow the host to determine if the UID is without bit errors, the UID is returned with its complement, as shown in Table 48. If the XOR of the UID and its bit-wise complement is all ones, then the UID is valid. 125 Bytes 0-15 16-31 Table 48 Value UID UID complement (bit-wise) UID and Complement To accommodate robust retrieval of the UID in the case of bit errors, sixteen copies of the UID and the corresponding complement shall be stored by the target. For example, reading bytes 3263 returns to the host another copy of the UID and its complement. Read Status Enhanced shall not be used during execution of the Read Unique ID command. Figure 58 defines the Read Unique ID behavior. The host may use any timing mode supported by the target in order to retrieve the UID data. 126 Cycle Type CMD ADDR DOUT DOUT DOUT DOUT DOUT DOUT DQ[7:0] EDh 00h U00 U10 ... U01 U11 ... tRR tWB tR R/B_n Figure 58 U0k-Unk Read Unique ID command timing The kth copy of the UID and its complement. Sixteen copies are stored. Reading beyond 512 bytes returns indeterminate values. 5.9. Block Erase Definition The Block Erase function erases the block of data identified by the block address parameter on the LUN specified. After a successful Block Erase, all bits shall be set to one in the block. SR[0] is valid for this command after SR[6] transitions from zero to one until the next transition of SR[6] to zero. Figure 59 defines the Block Erase behavior and timings. Cycle Type DQ[7:0] CMD ADDR ADDR ADDR CMD 60h R1 R2 R3 D0h tWB tBERS SR[6] Figure 59 Block Erase timing R1-R3 The row address of the block to be erased. R1 is the least significant byte in the row address. 5.10. Read Status Definition In the case of non-multi-plane operations, the Read Status function retrieves a status value for the last operation issued. If multiple multi-plane operations are in progress on a single LUN, then Read Status returns the composite status value for status register bits that are independent per plane address. Specifically, Read Status shall return the combined status value of the independent status register bits according to Table 49. See section 5.13 for status register bit definitions. 127 Status Register bit Bit 0, FAIL Bit 1, FAILC Bit 3, CSP Table 49 Composite status value OR OR OR Composite Status Value When issuing Read Status in the source synchronous data interface, each data byte is received twice. The host shall only latch one copy of each data byte. See section 4.3.2.5. Figure 60 defines the Read Status behavior and timings. Cycle Type DOUT CMD tWHR DQ[7:0] Figure 60 SR 70h SR Read Status timing Status value as defined in section 5.13. 128 The Read Status command may be issued using either the asynchronous or source synchronous data interfaces. The timing parameters for each data interface are shown in Figure 61 and Figure 62. Figure 61 Read Status command using asynchronous data interface 129 Figure 62 Read Status command using source synchronous data interface Note (1): For the asynchronous data interface, status may be continually read by pulsing RE_n or leaving RE_n low. For the source synchronous interface, status may continually be read by leaving ALE/CLE at a value of 11b. 5.11. Read Status Enhanced Definition The Read Status Enhanced function retrieves the status value for a previous operation on the particular LUN and plane address specified. Figure 63 defines the Read Status Enhanced behavior and timings. If the row address entered is invalid, the Status value returned has an indeterminate value. The host uses Read Status Enhanced for LUN selection (refer to section 3.1.2). Note that Read Status Enhanced has no effect on which page register is selected for data output within the LUN. When issuing Read Status Enhanced in the source synchronous data interface, each data byte is received twice. The host shall only latch one copy of each data byte. See section 4.3.2.5. 130 Cycle Type CMD ADDR ADDR DOUT ADDR tWHR DQ[7:0] 78h Figure 63 R1 R2 SR R3 Read Status Enhanced timing R1-R3 Row address of the previous operation to retrieve status for. R1 is the least significant byte. The row address contains both the LUN and plane address to retrieve status for. SR 5.12. Status value as defined in section 5.13. Read Status and Read Status Enhanced required usage In certain sequences only one status command shall be used by the host. This section outlines situations in which a particular status command is required to be used. If a command is issued to a LUN while R/B_n is cleared to zero, then the next status command shall be Read Status Enhanced. Read Status Enhanced causes LUNs that are not selected to turn off their output buffers. This ensures that only the LUN selected by the Read Status Enhanced commands responds to a subsequent toggle of the RE_n input signal. When the host has issued Read Page commands to multiple LUNs at the same time, the host shall issue Read Status Enhanced before reading data from either LUN. Read Status Enhanced causes LUNs that are not selected to turn off their output buffers. This ensures that only the LUN selected by the Read Status Enhanced commands responds to a subsequent toggle of the RE_n input signal after data output is selected with the 00h command. Refer to section 3.1.3 for additional requirements if a Change Read Column (Enhanced) command is used as part of a multiple LUN read sequence. During and after Target level commands, the host shall not issue the Read Status Enhanced command. In these sequences, the host uses Read Status to check for the status value. The only exception to this requirement is if commands were outstanding to multiple LUNs when a Reset was issued. In this case, the Read Status Enhanced command shall be used to determine when each active LUN has completed Reset. 5.13. Status Field Definition The returned status register byte value (SR) for Read Status and Read Status Enhanced has the format described below. If the RDY bit is cleared to zero, all other bits in the status byte (except WP_n) are invalid and shall be ignored by the host. Value Status Register FAIL 7 6 5 4 3 2 1 0 WP_n RDY ARDY VSP CSP R FAILC FAIL If set to one, then the last command failed. If cleared to zero, then the last command was successful. For raw NAND operation, this bit is only valid for 131 program and erase operations. For EZ NAND operation, this bit is valid for read, program, and erase operations. During program cache operations, this bit is only valid when ARDY is set to one. FAILC If set to one, then the command issued prior to the last command failed. If cleared to zero, then the command issued prior to the last command was successful. This bit is only valid for program cache operations. This bit is not valid until after the second 15h command or the 10h command has been transferred in a Page Cache Program sequence. When program cache is not supported, this bit is not used and shall be cleared to zero. For EZ NAND operation, this bit is not used (cache commands are not supported with EZ NAND). CSP Command Specific: This bit has command specific meaning. For EZ NAND read operations, if CSP (Threshold) bit is set to one then the last read operation exceeded the ECC threshold and the host should take appropriate action (e.g. rewrite the data to a new location). When FAIL is set to one, the CSP (Threshold) bit is don’t care. For all other operations, this bit is reserved. ARDY If set to one, then there is no array operation in progress. If cleared to zero, then there is a command being processed (RDY is cleared to zero) or an array operation in progress. When overlapped multi-plane operations or cache commands are not supported, this bit is not used. RDY If set to one, then the LUN or plane address is ready for another command and all other bits in the status value are valid. If cleared to zero, then the last command issued is not yet complete and SR bits 5:0 are invalid and shall be ignored by the host. This bit impacts the value of R/B_n, refer to section 2.16.2. When caching operations are in use, then this bit indicates whether another command can be accepted, and ARDY indicates whether the last operation is complete. WP_n If set to one, then the device is not write protected. If cleared to zero, then the device is write protected. This bit shall always be valid regardless of the state of the RDY bit. R Reserved (0) VSP Vendor Specific 5.14. Read Definition The Read function reads a page of data identified by a row address for the LUN specified. The page of data is made available to be read from the page register starting at the column address specified. Figure 64 defines the Read behavior and timings. Reading beyond the end of a page results in indeterminate values being returned to the host. While monitoring the read status to determine when the tR (transfer from Flash array to page register) is complete, the host shall re-issue a command value of 00h to start reading data. 132 Issuing a command value of 00h will cause data to be returned starting at the selected column address. 133 Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD DOUT DOUT DOUT 00h C1 C2 R1 R2 R3 30h D0 D1 D2 tRR tWB tR SR[6] Figure 64 Read timing C1-C2 Column address of the page to retrieve. C1 is the least significant byte. R1-R3 Row address of the page to retrieve. R1 is the least significant byte. Dn Data bytes read from the addressed page. 134 5.15. Read Cache Definition This command is not supported for EZ NAND. The Read Cache Sequential and Read Cache Random functions permit a page to be read from the page register while another page is simultaneously read from the Flash array for the selected LUN. A Read Page command, as defined in section 5.14, shall be issued prior to the initial Read Cache Sequential or Read Cache Random command in a read cache sequence. A Read Cache Sequential or Read Cache Random command shall be issued prior to a Read Cache End (3Fh) command being issued. The Read Cache (Sequential or Random) function may be issued after the Read function is complete (SR[6] is set to one). The host may enter the address of the next page to be read from the Flash array. Data output always begins at column address 00h. If the host does not enter an address to retrieve, the next sequential page is read. When the Read Cache (Sequential or Random) function is issued, SR[6] is cleared to zero (busy). After the operation is begun SR[6] is set to one (ready) and the host may begin to read the data from the previous Read or Read Cache (Sequential or Random) function. Issuing an additional Read Cache (Sequential or Random) function copies the data most recently read from the array into the page register. When no more pages are to be read, the final page is copied into the page register by issuing the 3Fh command. The host may begin to read data from the page register when SR[6]is set to one (ready). When the 31h and 3Fh commands are issued, SR[6] shall be cleared to zero (busy) until the page has finished being copied from the Flash array. The host shall not issue a Read Cache Sequential (31h) command after the last page of a block is read. If commands are issued to multiple LUNs at the same time, the host shall execute a Read Status Enhanced (78h) command to select the LUN prior to issuing a Read Cache Sequential (31h) or Read Cache End (3Fh) command for that LUN. Figure 65 defines the Read Cache Sequential behavior and timings for the beginning of the cache operations subsequent to a Read command being issued to the target. Figure 66 defines the Read Cache Random behavior and timings for the beginning of the cache operations subsequent to a Read command being issued to the target. In each case, SR[6] conveys whether the next selected page can be read from the page register. 135 As defined for Read Cycle Type DQx CMD CMD DOUT DOUT DOUT CMD DOUT 30h 31h D0 ... Dn 31h D0 tWB tRR tWB tR tRCBSY tRR tWB tRCBSY SR[6] Figure 65 Read Cache Sequential timing, start of cache operations D0-Dn Data bytes/words read from page requested by the original Read or the previous cache operation. 136 As defined for Read Cycle Type DQx A CMD CMD ADDR ADDR ADDR ADDR ADDR CMD DOUT DOUT DOUT 30h 00h C1 C2 R1 R2 R3 31h D0 ... Dn tRR tWB tRR tWB tR tRCBSY SR[6] A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD DOUT 00h C1 C2 R1 R2 R3 31h D0 tRR tWB tRCBSY SR[6] Figure 66 Read Cache Random timing, start of cache operations C1-C2 Column address of the page to retrieve. C1 is the least significant byte. The column address is ignored. R1-R3 Row address of the page to retrieve. R1 is the least significant byte. D0-Dn Data bytes/words read from page requested by the original Read or the previous cache operation 137 Figure 67 defines the Read Cache (Sequential or Random) behavior and timings for the end of cache operations. This applies for both Read Cache Sequential and Read Cache Random. A command code of 3Fh indicates to the target to transfer the final selected page into the page register, without beginning another background read operation. As defined for Read Cache (Sequential or Random) Cycle Type DQx CMD DOUT DOUT DOUT CMD DOUT DOUT DOUT 31h D0 ... Dn 3Fh D0 ... Dn tRR tWB tRR tWB tRCBSY tRCBSY SR[6] Figure 67 Read Cache timing, end of cache operations D0-Dn Data bytes/words read from page requested by the previous cache operation. 138 5.16. Page Program Definition The Page Program command transfers a page or portion of a page of data identified by a column address to the page register. The contents of the page register are then programmed into the Flash array at the row address indicated. SR[0] is valid for this command after SR[6] transitions from zero to one until the next transition of SR[6] to zero. Figure 68 defines the Page Program behavior and timings. Writing beyond the end of the page register is undefined. 139 Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN DIN CMD D0 D1 ... Dn 10h tADL DQx 80h C1 C2 R1 R2 R3 tWB tPROG SR[6] Figure 68 Page Program timing C1-C2 Column address of the starting buffer location to write data to. C1 is the least significant byte. R1-R3 Row address of the page being programmed. R1 is the least significant byte. D0-Dn Data bytes/words to be written to the addressed page. 140 5.17. Page Cache Program Definition This command is not supported for EZ NAND. The Page Cache Program function permits a page or portion of a page of data to be written to the Flash array for the specified LUN in the background while the next page to program is transferred by the host to the page register. After the 10h command is issued, all data is written to the Flash array prior to SR[6] being set to one (ready). SR[0] is valid for this command after SR[5] transitions from zero to one until the next transition. SR[1] is valid for this command after SR[6] transitions from zero to one, and this is not the first operation. Figure 69 and Figure 70 define the Page Cache Program behavior and timings. Note that tPROG at the end of the caching operation may be longer than typical as this time also accounts for completing the programming operation for the previous page. Writing beyond the end of the page register is undefined. If the program page register clear enhancement is supported, then the host may choose to only clear the page register for the selected LUN and plane address when a Program (80h) command is received. In this case, the tADL time may be longer than defined for the selected timing mode, refer to section 5.7.1.17. Refer to section 5.26.1 for details on how to enable this feature. 141 A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN DIN CMD D0 D1 ... Dn 15h tADL DQx 80h C1 C2 R1 R2 R3 tWB tPCBSY SR[6] A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN DIN CMD D0 D1 ... Dn 15h tADL DQx 80h C1 C2 R1 R2 R3 tWB tPCBSY SR[6] Figure 69 142 Page Cache Program timing, start of operations As defined for Page Cache Program Cycle Type CMD A ADDR ADDR ADDR ADDR ADDR DIN DIN DIN DIN CMD D0 D1 ... Dn 15h tADL DQx 80h C1 C2 R1 R2 R3 tWB tPCBSY SR[6] A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN DIN CMD D0 D1 ... Dn 10h tADL DQx 80h C1 C2 R1 R2 R3 tWB tPROG SR[6] Figure 70 143 Page Cache Program timing, end of operations C1-C2 Column address of the starting buffer location to write data to. C1 is the least significant byte. R1-R3 Row address of the page being programmed. R1 is the least significant byte. D0-Dn Data bytes/words to be written to the addressed page. 5.18. Copyback Definition The Copyback function reads a page of data from one location and then moves that data to a second location on the same LUN. If the target supports EZ NAND then the data may be moved to a second location on a different plane or LUN. Refer to the parameter page to determine the destinations support for Copyback. The data read from the first location may be read by the host, including use of Change Read Column. After completing any data read out and issuing Copyback Program, the host may perform data modification using Change Write Column as needed. Figure 71 defines the Copyback behavior and timings. Copyback uses a single page register for the read and program operation. If the target supports EZ NAND, the buffer in the EZ NAND controller is used for read and program operations and the page register within each LUN is not explicitly accessible. When multi-plane addressing is supported, the multi-plane address for Copyback Read and Copyback Program for a non-multi-plane Copyback operation shall be the same. If EZ NAND is supported, this restriction may not apply; refer to the parameter page. Copyback may also have odd/even page restrictions. Specifically, when reading from an odd page, the contents may need to be written to an odd page. Alternatively, when reading from an even page, the contents may need to be written to an even page. Refer to section 5.7.1.3. This revision of the ONFI specification requires Copyback Adjacency for all implementations. In future ONFI revisions, for EZ NAND implementations it is possible that Copyback Read and Copyback Program are not required to be adjacent. I.e., there may be multiple Copyback Read commands prior to a Copyback Program being issued. This requirement also applies to multi-plane Copyback operations. 144 A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1A C2A R1A R2A R3A 35h tWB tR SR[6] A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD 85h C1B C2B R1B R2B R3B 10h tWB tPROG SR[6] Figure 71 Copyback timing C1-C2A Column address of the page to retrieve. C1A is the least significant byte. R1-R3 A Row address of the page to retrieve. R1A is the least significant byte. C1-C2B Column address of the page to program. C1B is the least significant byte. 145 R1-R3 B Row address of the page to program. R1B is the least significant byte. Figure 72 and Figure 73 define Copyback support for data output and data modification. A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD DOUT DOUT DOUT 00h C1A C2A R1A R2A R3A 35h Dj ... Dj+n tRR tWB tR SR[6] A Cycle Type CMD ADDR ADDR CMD DOUT DOUT DOUT Dk Dk+1 Dk+2 tCCS DQx 05h C1C C2C E0h SR[6] Figure 72 C1-C2A Copyback Read with data output Column address of the page to retrieve. C1A is the least significant byte. 146 R1-R3 A Row address of the page to retrieve. R1A is the least significant byte. Dj-(Dj+n) Data bytes read starting at column address specified in C1-C2A. C1-C2C Column address of new location (k) to read out from the page register. C1C is the least significant byte. Dk-Dk+n Data bytes read starting at column address specified in C1-C2C. 147 A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN Di Di+1 tADL DQx 85h C1B C2B CMD ADDR ADDR R1B R2B R3B SR[6] A Cycle Type DIN DIN DIN CMD Dj Dj+1 Dj+2 10h tCCS DQx 85h C1D C2D tWB tPROG SR[6] Figure 73 Copyback Program with data modification C1-C2B Column address of the page to program. C1B is the least significant byte. R1-R3 B Row address of the page to program. R1B is the least significant byte. Di-Di+n Data bytes overwritten in page register starting at column address specified in C1-C2B. C1-C2D Column address of new location (j) to overwrite data at in the page register. C1D is the least significant byte. 148 Dj-Dj+n 5.19. Data bytes overwritten starting at column address specified in C1-C2D. Small Data Move If the Small Data Move command is supported, as indicated in the parameter page, then the host may transfer data to the page register in increments that are less than the page size of the device for both Program and Copyback operations (including multi-plane Program and Copyback operations). The host may also read data out as part of the operation. If the Small Data Move is a program operation with no data output, then the 80h opcode may be used for the first cycles. For Copyback and program operations that include data output, the 85h opcode shall be used for the first cycles. Figure 74 defines the data modification portion of a Program or Copyback Program with small data moves; this sequence may be repeated as necessary to complete the data transfer. Figure 75 defines the final program operation that is used to complete the Program or Copyback Program with small data move operation. The row address (R1B – R3B) shall be the same for all program portions of the sequence destined for the same plane address. The function of the 11h command in a small data move operation is to flush any internal data pipeline in the device prior to resuming data output. 149 A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN CMD Dj ... Dj+n 11h tCCS DQx PRG C1B C2B R1B R2B R3B tWB tIPBSY SR[6] A Cycle Type B CMD ADDR ADDR CMD DOUT DOUT DOUT Dk ... Dk+p tCCS DQx 05h C1C C2C E0h SR[6] Figure 74 Small data moves, data modification PRG Program command, either 80h or 85h. Following any data output, the command shall be 85h. C1-C2B Column address to write to in the page register. C1B is the least significant byte. R1-R3B Row address of the page to program. R1B is the least significant byte. 150 Dj-(Dj+n) Data bytes to update in the page register starting at column address specified in C1-C2B. C1-C2C Column address of the byte/word in the page register to retrieve. C1C is the least significant byte. Dk-(Dk+p) Data bytes read starting at column address specified in C1-C2C. NOTE: If Change Read Column Enhanced is supported, this command may be substituted for Change Read Column in Figure 74. Use of the Change Read Column (Enhanced) command and data output in this flow is optional; this flow may be used to incrementally transfer data for a Program or Copyback Program. B Cycle Type DOUT CMD ADDR ADDR ADDR ADDR ADDR tRHW DQx Dk+p DIN DIN DIN CMD Dl ... Dl+q 10h tCCS PRG C1D C2D R1B R2B R3B tWB tPROG SR[6] Figure 75 Small data moves, end PRG Program command, either 80h or 85h. 85h shall be used if there is any data output as part of the command. C1-C2D Column address to write to in the page register. C1D is the least significant byte. R1-R3B Row address of the page to program. R1B is the least significant byte. Dl-(Dl+q) Data bytes to update in the page register starting at column address specified in C1-C2D. 151 5.20. Change Read Column Definition The Change Read Column function changes the column address from which data is being read in the page register for the selected LUN. Change Read Column shall only be issued when the LUN is in a read idle condition. Figure 76 defines the Change Read Column behavior and timings. The host shall not read data from the LUN until tCCS ns after the E0h command is written to the LUN. Refer to Figure 76. Cycle Type DOUT DOUT CMD ADDR ADDR CMD tRHW DQx Dn Dn+1 DOUT DOUT DOUT Dk Dk+1 Dk+2 tCCS 05h C1 C2 E0h SR[6] Figure 76 Dn Change Read Column timing Data bytes read prior to the column address change. C1-C2 New column address to be set for subsequent data transfers. C1 is the least significant byte. Dk 5.21. Data bytes being read starting with the new addressed column. Change Read Column Enhanced Definition The Change Read Column Enhanced function changes the LUN address, plane address and column address from which data is being read in a page previously retrieved with the Read command. This command is used when independent LUN operations or multi-plane operations are being performed such that the entire address for the new column needs to be given. Figure 77 defines the Change Read Column Enhanced behavior and timings. The Change Read Column Enhanced command shall not be issued by the host unless it is supported as indicated in the parameter page. Change Read Column Enhanced shall not be issued while Target level data output commands (Read ID, Read Parameter Page, Read Unique ID, Get Features) are executing or immediately following Target level commands. Change Read Column Enhanced causes idle LUNs (SR[6] is one) that are not selected to turn off their output buffers. This ensures that only the LUN selected by the Change Read Column Enhanced command responds to subsequent data output. If unselected LUNs are active (SR[6] is zero) when Change Read Column Enhanced is issued, then the host shall issue a Read Status Enhanced (78h) command prior to subsequent data output to ensure all LUNs that are not selected turn off their output buffers. 152 Cycle Type DOUT DOUT CMD ADDR ADDR ADDR ADDR ADDR CMD tRHW DQx Dn Dn+1 06h C1 Figure 77 C2 R1 R2 R3 E0h Change Read Column Enhanced timing Data bytes read prior to the row and column address change. C1-C2 New column address to be set for subsequent data transfers. C1 is the least significant byte. R1-R3 New row address to be set for subsequent data transfers. R1 is the least significant byte. Dk DOUT DOUT Dk Dk+1 Dk+2 tCCS SR[6] Dn DOUT Data bytes being read starting with the new addressed row and column. 153 5.22. Change Write Column Definition The Change Write Column function changes the column address being written to in the page register for the selected LUN. Figure 78 defines the Change Write Column behavior and timings. The host shall not write data to the LUN until tCCS ns after the last column address is written to the LUN. Refer to Figure 76. As defined for Page (Cache) Program Cycle Type DIN As defined for Page (Cache) Program DIN CMD ADDR ADDR DIN DIN DIN Dk Dk+1 Dk+2 tCCS DQx Dn Dn+1 85h C1 C2 SR[6] Figure 78 Change Write Column timing C1-C2 New column address to be set for subsequent data transfers. C1 is the least significant byte. Dn Data bytes being written to previous addressed column Dk Data bytes being written starting with the new addressed column 5.23. Change Row Address Definition The Change Row Address function changes the row and column address being written to for the selected LUN. This mechanism may be used to adjust the block address, page address, and column address for a Program that is in execution. The LUN and plane address shall be the same as the Program that is in execution. Figure 79 defines the Change Row Address behavior and timings. The host shall not write data to the LUN until tCCS ns after the last row address is written to the LUN. Refer to Figure 79. 154 As defined for Page (Cache) Program Cycle Type DIN As defined for Page (Cache) Program DIN CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN Dk Dk+1 Dk+2 tCCS DQx Dn Dn+1 85h C1 C2 R1 R2 R3 SR[6] Figure 79 Change Row Address timing C1-C2 New column address to be set for subsequent data transfers. C1 is the least significant byte. R1-R3 Row address of the page being programmed. The LUN address and plane address shall be the same as the Program in execution. R1 is the least significant byte. Dn Data bytes being written prior to row address change; will be written to new row address Dk Data bytes being written to the new block and page, starting with the newly addressed column 155 5.24. Set Features Definition The Set Features function modifies the settings of a particular feature. For example, this function can be used to enable a feature that is disabled at power-on. Parameters are always transferred on the lower 8-bits of the data bus. Figure 80 defines the Set Features behavior and timings. When issuing Set Features in the source synchronous data interface, each data byte is transmitted twice. The device shall only latch one copy of each data byte. See section 4.3.2.3. Set Features is used to change the timing mode and data interface type. When changing the timing mode, the device is busy for tITC, not tFEAT. During the tITC time the host shall not poll for status. 156 Cycle Type CMD ADDR DIN DIN DIN DIN P1 P2 P3 P4 tADL DQx EFh FA tWB tFEAT R/B_n Figure 80 Set Features timing * NOTE: Busy time is tITC when setting the timing mode. FA Feature address identifying feature to modify settings for. P1-P4 Parameters identifying new settings for the feature specified. P1 P2 P3 P4 Sub feature parameter 1 Sub feature parameter 2 Sub feature parameter 3 Sub feature parameter 4 Refer to section 5.26 for the definition of features and sub feature parameters. 157 5.25. Get Features Definition The Get Features function is the mechanism the host uses to determine the current settings for a particular feature. This function shall return the current settings for the feature (including modifications that may have been previously made with the Set Features function). Parameters are always transferred on the lower 8-bits of the data bus. After reading the first byte of data, the host shall complete reading all desired data before issuing another command (including Read Status or Read Status Enhanced). Figure 81 defines the Get Features behavior and timings. When issuing Get Features in the source synchronous data interface, each data byte is received twice. The host shall only latch one copy of each data byte. See section 4.3.2.5. If Read Status is used to monitor when the tFEAT time is complete, the host shall issue a command value of 00h to begin transfer of the feature data starting with parameter P1. Cycle Type CMD ADDR DOUT DOUT DOUT DOUT DQx EEh FA P1 P2 P3 P4 tRR tWB tFEAT R/B_n Figure 81 FA Get Features timing Feature address identifying feature to return parameters for. P1-P4 Current settings/parameters for the feature identified by argument P1 P1 P2 P3 P4 Sub feature parameter 1 setting Sub feature parameter 2 setting Sub feature parameter 3 setting Sub feature parameter 4 setting Refer to section 5.26 for the definition of features and sub feature parameters. 5.26. Feature Parameter Definitions If the Set Features and Get Features commands are not supported by the Target, then no feature parameters are supported. Additionally, the Target only supports feature parameters defined in ONFI specification revisions that the Target complies with. Feature settings are volatile across power cycles. For each feature setting, whether the value across resets is retained is explicitly stated. Feature Address 00h 01h 02h-0Fh Description Reserved Timing Mode Reserved 158 10h 11h-1Fh 20h-4Fh 50h 51h-5Fh 60h-7Fh 80h-FFh 5.26.1. I/O Drive Strength Reserved for programmable I/O settings Reserved EZ NAND control Reserved Reserved for Block Abstracted NAND Vendor specific Timing Mode This setting shall be supported if the Target complies with ONFI specification revision 1.0. The Data Interface setting is not retained across Reset (FFh); after a Reset (FFh) the Data Interface shall be asynchronous. All other settings for the timing mode are retained across Reset (FFh) and Synchronous Reset (FCh) commands. If the Reset (FFh) command is issued when the Data Interface is configured as source synchronous, then the host shall use the asynchronous data interface with Timing Mode 0 until a new data interface and/or timing mode is selected with Set Features. Hosts shall only set a timing mode that is explicitly shown as supported in the Read Parameter Page. The results of the host using Set Features to transition from the source synchronous data interface to the asynchronous data interface is indeterminate. To transition to the asynchronous data interface, the host should use the Reset (FFh) command. Sub Feature Parameter 7 6 P1 R PC P2 P3 P4 5 4 3 2 1 0 Data Timing Mode Number Interface Reserved (0) Reserved (0) Reserved (0) Timing Mode Number Set to the numerical value of the maximum timing mode in use by the host. Default power-on value is 0h. Data Interface 00b = asynchronous (default power-on value) 01b = source synchronous 10-11b = Reserved PC The Program Clear bit controls the program page register clear enhancement which defines the behavior of clearing the page register when a Program (80h) command is received. If cleared to zero, then the page register(s) for each LUN that is part of the target is cleared when the Program (80h) command is received. If set to one, then only the page register for the LUN and interleave address selected with the Program (80h) command is cleared and the tADL time for Program commands is as reported in the parameter page. Reserved / R Reserved values shall be cleared to zero by the host. Targets shall not be sensitive to the value of reserved fields. 159 5.26.2. I/O Drive Strength This setting shall be supported if the Target supports the source synchronous data interface. The I/O drive strength setting shall be retained across Reset (FFh) and Synchronous Reset (FCh) commands. The power-on default drive strength value is the Nominal (10b) setting. Sub Feature Parameter P1 P2 P3 P4 7 6 5 4 3 2 Reserved (0) Reserved (0) Reserved (0) Reserved (0) 1 0 Drive Strength Drive strength 00b = 18 Ohm 01b = 25 Ohm 10b = 35 Ohm (power-on default) 11b = 50 Ohm Reserved Reserved values shall be cleared to zero by the host. Targets shall not be sensitive to the value of reserved fields. 5.26.3. EZ NAND control This setting shall be supported if the device supports EZ NAND. This feature is used to control settings for the EZ NAND device. Sub Feature Parameter P1 P2 P3 P4 7 6 5 4 3 Reserved (0) Reserved (0) Reserved (0) Reserved (0) 2 1 0 RD Retry Disable (RD) If set to one, then the EZ NAND device shall not automatically perform retries. If cleared to zero, then the EZ NAND device may automatically perform retries during error conditions at its discretion. If automatic retries are disabled, the device may exceed the UBER specified. Automatic retries shall only be disabled if the device supports this capability as indicated in the parameter page. If an EZ NAND controller executes an automatic retry, the typical page read time (tR) may be exceeded. Reserved / R Reserved values shall be cleared to zero by the host. Targets shall not be sensitive to the value of reserved fields. 160 6. Multi-plane Operations A LUN may support multi-plane read, program and erase operations. Multi-plane operations are when multiple commands of the same type are issued to different blocks on the same LUN. Refer to section 5.7.1.28 for addressing restrictions with multi-plane operations. There are two methods for multi-plane operations: concurrent and overlapped. When performing multi-plane operations, the operations/functions shall be the same type. The functions that may be used in multi-plane operations are: Page Program Copyback Read and Program Block Erase Read 6.1. Requirements When supported, the plane address comprises the lowest order bits of the block address as shown in Figure 19. The LUN and page addresses are required to be the same. The block address (other than the plane address bits) may be required to be the same, refer to section 5.7.1.28. For copyback program operations, the restrictions are the same as for a multi-planeprogram operation. However, copyback reads shall be previously issued to the same plane addresses as those in the multi-plane copyback program operations. The reads for copyback may be issued non-multi-plane or multi-plane. If the reads are non-multi-plane then the reads may have different page addresses. If the reads are multi-plane then the reads shall have the same page addresses. Multi-plane operations enable operations of the same type to be issued to other blocks on the same LUN. There are two methods for multi-plane operations: concurrent and overlapped. The concurrent multi-plane operation waits until all command, address, and data are entered for all plane addresses before accessing the Flash array. The overlapped multi-plane operation begins its operation immediately after the command, address and data are entered and performs it in the background while the next multi-plane command, address, and data are entered. The plane address component of each address shall be distinct. A single multi-plane (cached) program operation is shown in Figure 82. Between “Multi-plane Op 1” and “Multi-plane Op n”, all plane addresses shall be different from each other. After the 10h or 15h (cached) command cycle is issued, previously issued plane addresses can be used in future multi-plane operations. 80h 11h <ADDR1> CMD CMD 80h 11h <ADDR2> CMD CMD Multi-plane Op 1 Multi-plane Op 2 Figure 82 … 10h 80h <ADDRn > or 15h CMD CMD Multi-plane Op n Multi-plane Program (Cache) For multi-plane erase operations, the plane address component of each address shall be distinct. A single multi-plane erase operation is shown in Figure 83. Between “Multi-plane Op 1” and “Multi-plane Op n”, all plane addresses shall be different from each other. After the D0h command cycle is issued, previously issued plane addresses can be used in future multi-plane operations. 161 60h D1h <ADDR1> CMD CMD 60h D1h <ADDR2> CMD CMD Multi-plane Op 1 Multi-plane Op 2 Figure 83 … 60h D0h <ADDRn > CMD CMD Multi-plane Op n Multi-plane Erase The plane address component of each address shall be distinct. A single multi-plane read (cache) operation is shown in Figure 84. Between “Multi-plane Op 1” and “Multi-plane Op n”, all plane addresses shall be different from each other. After the 30h or 31h (cached) command cycle is issued, previously issued plane addresses can be used in future multi-plane operations. 00h 32h <ADDR1> CMD CMD 00h 32h <ADDR2> CMD CMD Multi-plane Op 1 Multi-plane Op 2 Figure 84 30h 00h <ADDRn > or 31h CMD CMD … Multi-plane Op n Multi-plane Read (Cache) 6.2. Status Register Behavior Some status register bits are independent per plane address. Other status register bits are shared across the entire LUN. This section defines when status register bits are independent per plane address. This is the same for concurrent and overlapped operations. For multi-plane program and erase operations, the FAIL/FAILC bits are independent per plane address. Table 50 lists whether a bit is independent per plane address or shared across the entire LUN for multi-plane operations. Value Status Register Independent 7 WP_n N Table 50 6 RDY N 5 ARDY N 4 VSP N 3 R N 2 R N 1 FAILC Y 0 FAIL Y Independent Status Register bits 6.3. Multi-plane Page Program The Page Program command transfers a page or portion of a page of data identified by a column address to the page register. The contents of the page register are then programmed into the Flash array at the row address indicated. With a multi-plane operation, multiple programs can be issued back to back to the LUN, with a shorter busy time between issuance of the next program operation. Figure 85 defines the behavior and timings for two multi-plane page program commands. Cache operations may be used when doing multi-plane page program operations, as shown, if supported by the target as indicated in the parameter page. Refer to section 5.7.1.27. 162 A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN DIN CMD D0A D1A ... DnA 11h tADL DQx 80h C1A C2A R1A R2A R3A tWB tPLPBSY SR[6] A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR DIN DIN DIN DIN CMD D0B D1B ... DnB 10h tADL DQx 80h C1B C2B R1B R2B R3B tWB tPROG SR[6] Figure 85 Multi-plane Page Program timing C1A-C2A Column address for page A. C1A is the least significant byte. R1A-R3A Row address for page A. R1A is the least significant byte. D0A-DnA Data to program for page A. C1B-C2B Column address for page B. C1B is the least significant byte. 163 R1B-R3B Row address for page B. R1B is the least significant byte. D0B-DnB Data to program for page B. The row addresses for page A and B shall differ in the plane address bits. Finishing a multi-plane program with a command cycle of 15h rather than 10h indicates that this is a cache operation. The host shall only issue a command cycle of 15h to complete an multi-plane program operation if program cache is supported with multi-plane program operations, as described in section 5.7.1.27. 164 6.4. Multi-plane Copyback Read and Program The Copyback function reads a page of data from one location and then moves that data to a second location. With a multi-plane operation, the Copyback Program function can be issued back to back to the target, with a shorter busy time between issuance of the next Copyback Program. Figure 86, Figure 87, and Figure 88 define the behavior and timings for two Copyback Program operations. The reads for the Copyback Program may or may not be multi-plane. Figure 86 defines the non-multi-plane read sequence and Figure 87 defines the multi-plane read sequence. The plane addresses used for the Copyback Read operations (regardless of multi-plane) shall be the same as the plane addresses used in the subsequent multi-plane Copyback Program operations. If EZ NAND is supported, this restricton may not apply; refer to the parameter page. A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1A C2A R1A R2A R3A 35h DQx tWB tR SR[6] A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1B C2B R1B R2B R3B 35h tWB tR SR[6] Figure 86 Non-multi-plane Copyback Read timing for multi-plane Copyback Program C1A-C2A Column address for source page A. C1A is the least significant byte. R1A-R3A Row address for source page A. R1A is the least significant byte. C1B-C2B Column address for source page B. C1B is the least significant byte. R1B-R3B Row address for source page B. R1B is the least significant byte. 165 The row addresses for all source pages shall differ in their plane address bits. A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1A C2A R1A R2A R3A 32h DQx tWB tPLRBSY SR[6] A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1B C2B R1B R2B R3B 35h tWB tR SR[6] Figure 87 Multi-plane Copyback Read timing for Multi-plane Copyback Program C1A-C2A Column address for source page A. C1A is the least significant byte. R1A-R3A Row address for source page A. R1A is the least significant byte. C1B-C2B Column address for source page B. C1B is the least significant byte. R1B-R3B Row address for source page B. R1B is the least significant byte. The row addresses for all source pages shall differ in their plane address bits. The source page addresses shall be the same for multi-plane reads. 166 A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR CMD 85h C1C C2C R1C R2C R3C 11h DQx tWB tPLPBSY SR[6] A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD 85h C1D C2D R1D R2D R3D 10h tWB tPROG SR[6] Figure 88 Multi-plane Copyback Program C1C-C2C Column address for destination page C. C1C is the least significant byte. R1C-R3C Row address for destination page C. R1C is the least significant byte. C1D-C2D Column address for destination page D. C1D is the least significant byte. R1D-R3D Row address for destination page D. R1D is the least significant byte. The row addresses for all destination pages shall differ in their plane address bits. The page address for all destination addresses for multi-plane copyback operations shall be identical. 167 6.5. Multi-plane Block Erase Figure 89 defines the behavior and timings for a multi-plane block erase operation. Only two operations are shown, however additional erase operations may be issued with a 60h/D1h sequence prior to the final 60h/D0h sequence depending on how many multi-plane operations the LUN supports. Cycle Type DQ[7:0] CMD ADDR ADDR ADDR CMD CMD ADDR ADDR ADDR CMD 60h R1A R2A R3A D1h 60h R1B R2B R3B D0h tWB tWB tPLEBSY SR[6] Figure 89 Multi-plane Block Erase timing R1A-R3A Row address for erase block A. R1A is the least significant byte. R1B-R3B Row address for erase block B. R1B is the least significant byte. 168 tBERS 6.6. Multi-plane Read The Read command reads a page of data identified by a row address for the LUN specified. The page of data is made available to be read from the page register starting at the column address specified. With a multi-plane operation, multiple reads can be issued back to back to the LUN, with a shorter busy time between issuance of the next read operation. Figure 90 defines the behavior and timings for issuing two multi-plane read commands. Figure 91 defines the behavior and timings for reading data after the multi-plane read commands are ready to return data. Cache operations may be used when doing multi-plane read operations, as shown, if supported by the target as indicated in the parameter page. Refer to section 5.7.1.27. Change Read Column Enhanced shall be issued prior to reading data from a LUN. If data is read without issuing a Change Read Column Enhanced, the output received is undefined. A Cycle Type DQx CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1A C2A R1A R2A R3A 32h tWB tPLRBSY SR[6] A Cycle Type DQx B CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1B C2B R1B R2B R3B 30h tRR tWB tR SR[6] Figure 90 Multi-plane Read command issue timing C1A-C2A Column address for page A. C1A is the least significant byte. R1A-R3A Row address for page A. R1A is the least significant byte. C1B-C2B Column address for page B. C1B is the least significant byte. R1B-R3B Row address for page B. R1B is the least significant byte. The row addresses for page A and B shall differ in the plane address bits. 169 B Cycle Type CMD CMD ADDR ADDR ADDR ADDR ADDR CMD DOUT DOUT Dn Dn+1 tCCS DQx 06h 30h C1 C2 R1 R2 R3 E0h tRR tWB tR SR[6] Figure 91 Multi-plane Read data output timing, continued from command issue C1-C2 Column address to read from. C1 is the least significant byte. R1-R3 Row address to read from (specifies LUN and plane address). R1 is the least significant byte. Dn Data bytes read starting with addressed row and column. The row address provided shall specify a LUN and plane address that has valid read data. For Multi-plane Read Cache Sequential operations, the initial Multi-plane Read command issue is followed by a Read Cache confirmation opcode 31h, as shown in Figure 92. 170 As defined for Multi-plane Read D B Cycle Type DQx CMD CMD 30h 31h tRR tWB tR tWB tRCBSY SR[6] Figure 92 Multi-plane Read Cache Sequential command issue timing For Multi-plane Read Cache Random operations, the initial multi-plane Read command issue is followed by another Read Multi-plane command sequence where the last confirmation opcode is 31h, as shown in Figure 93. 171 B C Cycle Type DQx CMD CMD ADDR ADDR ADDR ADDR ADDR CMD 30h 00h C1C C2C R1C R2C R3C 32h tRR tWB tWB tPLRBSY tR SR[6] C Cycle Type DQx D CMD ADDR ADDR ADDR ADDR ADDR CMD 00h C1D C2D R1D R2D R3D 31h tRR tWB tRCBSY SR[6] Figure 93 Multi-plane Read Cache Random command issue timing C1C-C2C Column address for page C. C1C is the least significant byte. R1C-R3C Row address for page C. R1C is the least significant byte. C1D-C2D Column address for page D. C1D is the least significant byte. R1D-R3D Row address for page D. R1D is the least significant byte. 172 The row addresses for page C and D shall differ in the plane address bits. For Multi-plane Read Cache operations, two data output operations follow each Multi-plane Read Cache operation. The individual data output sequences are described in Figure 91. Prior to the last set (i.e. two) data output operations, a Read Cache End command (3Fh) should be issued by the host. 173 7. Behavioral Flows 7.1. Target behavioral flows The Target state machine describes the allowed sequences when operating with the target. If none of the arcs are true, then the target remains in the current state. 7.1.1. Variables This section describes variables used within the Target state machine. tbStatusOut This variable is set to TRUE when a data read cycle should return the status value. The power-on value for this variable is FALSE. tbChgCol This variable is set to TRUE when changing the column using Change Read Column is allowed. The power-on value for this variable is FALSE. tbChgColEnh This variable is set to TRUE when changing the column using Change Read Column Enhanced is allowed. The power-on value for this variable is FALSE. tCopyback This variable is set to TRUE if the Target is issuing a copyback command. The power-on value for this variable is FALSE. tLunSelected This variable contains the LUN that is currently selected by the host. The power-on value for this variable is 0. tLastCmd This variable contains the first cycle of the last command (other than 70h/78h) received by the Target. tReturnState tbStatus78hReq This variable contains the state to return to after status operations. This variable is set to TRUE when the next status operation shall be a 78h command (and not a 70h command). The power-on value for this variable is FALSE. 7.1.2. Idle states 1 The target performs the following actions: 1. R/B_n is cleared to zero. 2. Each LUN shall draw less than 10 mA of power per staggered power-up requirement. 2 T_PowerOnReady 1. Target is ready to accept FFh (Reset) command NOTE: 1. This state is entered as a result of a power-on event when Vcc reaches Vcc_min. 2. This arc shall be taken within 1 millisecond of Vcc reaching Vcc_min. T_PowerOn T_PowerOnReady The target performs the following actions: 1. R/B_n is set to one. 2. Each LUN shall draw less than 10mA of power per staggered power-up requirement. 1. Command cycle FFh (Reset) received T_RST_PowerOn 174 T_Idle tCopyback set to FALSE. tReturnState set to T_Idle. 1. WP_n signal transitioned T_Idle_WP_Transition 2. LUN indicates its SR[6] value transitioned T_Idle_RB_Transition 3. Command cycle received T_Cmd_Decode 1 Decode command received. tbStatusOut is set to FALSE. If R/B_n is set to one and command received is not 70h (Read Status), then tbStatus78hReq is set to FALSE. 1. (Command 80h (Page Program) or command 60h T_Idle (Block Erase) decoded) and WP_n is low 2. Command FFh (Reset) decoded T_RST_Execute 3. Command FCh (Synchronous Reset) decoded T_RST_Execute_Sync 4. Command FAh (Reset LUN) decoded T_RST_Execute_LUN 5. Command 90h (Read ID) decoded T_RID_Execute 6. Command ECh (Read Parameter Page) decoded T_RPP_Execute 7. Command EDh (Read Unique ID) decoded T_RU_Execute 8. Command 80h (Page Program) decoded and WP_n is T_PP_Execute high 9. Command 60h (Block Erase) decoded and WP_n is T_BE_Execute high 10. Command 00h (Read) decoded T_RD_Execute 11. Command EFh (Set Features) decoded T_SF_Execute 12. Command EEh (Get Features) decoded T_GF_Execute 13. Command 70h (Read Status) decoded T_RS_Execute 14. Command 78h (Read Status Enhanced) decoded T_RSE_Execute NOTE: 1. The host shall ensure R/B_n is set to one before issuing Target level commands (Reset, Read ID, Read Parameter Page, Read Unique ID, Set Features, Get Features). T_Cmd_Decode T_Idle_WP_Transition Indicate WP_n value to all LUN state machines. 1. State entered from T_Idle_Rd T_Idle_Rd 2. Else T_Idle 1 T_Idle_RB_Transition R/B_n is set to the AND of all LUN status register SR[6] values. 1. Unconditional tReturnState NOTE: 1. R/B_n may transition to a new value prior to the Target re-entering an idle condition when LUN level commands are in the process of being issued. 175 7.1.3. Idle Read states T_Idle_Rd 1. 2. 3. 4. 5. 6. Wait for read request (data or status) or other action. tReturnState set to T_Idle_Rd. WP_n signal transitioned T_Idle_WP_Transition LUN indicates its SR[6] value transitioned T_Idle_RB_Transition Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status Read request received and (tLastCmd set to 90h or T_Idle_Rd_XferByte EEh) Read request received and (tLastCmd set to ECh or T_Idle_Rd_LunByte EDh) Read request received and tbStatus78hReq set to T_Idle_Rd_LunData 1 FALSE 7. Command cycle 05h (Change Read Column) received T_CR_Execute2 and tbChgCol set to TRUE 8. Command cycle 06h (Change Read Column Enhanced) T_CRE_Execute2 received and tbChgColEnh set to TRUE 9. Command cycle of 31h received and tbStatus78hReq T_Idle_Rd_CacheCmd set to FALSE 10. Command cycle of 3Fh received and tLastCmd set to T_Idle_Rd_CacheCmd 31h and tbStatus78hReq set to FALSE 11. Command cycle received T_Cmd_Decode NOTE: 1. When tbStatus78hReq is set to TRUE, a Read Status Enhanced (78h) command followed by a 00h command shall be issued by the host prior to reading data from a particular LUN. 2. If there are reads outstanding on other LUNs for this target, a Change Read Column (Enhanced) shall be issued before transferring data. Refer to section 3.1.3 that describes multiple LUN operation restrictions. T_Idle_Rd_CacheCmd 1. Unconditional T_Idle_Rd_XferByte 1. Unconditional Set tLastCmd to the command received. Pass command received to LUN tLunSelected T_Idle_Rd Return next byte of data. T_Idle_Rd T_Idle_Rd_LunByte Request byte of data from page register of LUN tLunSelected. 1. Byte received from LUN tLunSelected T_Idle_Rd_XferHost T_Idle_Rd_LunData Request byte (x8) or word (x16) of data from page register of LUN tLunSelected. 1. Byte or word received from LUN tLunSelected T_Idle_Rd_XferHost 176 T_Idle_Rd_XferHost Transfer data byte or word received from LUN tLunSelected to host. 1. tReturnState set to T_RD_StatusOff and tCopyback set T_RD_Copyback to TRUE 2. tReturnState set to T_RD_StatusOff T_Idle_Rd 3. Else tReturnState T_Idle_Rd_Status Request status from LUN tLunSelected. 1. Status from LUN tLunSelected received T_Idle_Rd_StatusEnd 1. Unconditional T_Idle_Rd_StatusEnd Transfer status byte received from LUN tLunSelected to host. tReturnState T_CR_Execute Wait for a column address cycle. 1. Column address cycle received T_CR_Addr T_CR_Addr Store the column address cycle received. 1. More column address cycles required 2. All column address cycles received T_CR_Execute T_CR_WaitForCmd T_CR_WaitForCmd Wait for a command cycle. 1. Command cycle E0h received T_CR_ReturnToData T_CR_ReturnToData Request that LUN tLunSelected select the column in the page register based on the column address received. 1. tReturnState set to T_PP_MplWait T_PP_WaitForDataOut 2. tReturnState set to T_RD_Status_Off T_Idle_Rd 3. Else tReturnState T_CRE_Execute Wait for a column address cycle. 1. Column address cycle received T_CRE_ColAddr Store the column address cycle received. 1. More column address cycles required 2. All column address cycles received 177 T_CRE_ColAddr T_CRE_Execute T_CRE_RowAddrWait T_CRE_RowAddrWait Wait for a row address cycle. 1. Row address cycle received T_CRE_RowAddr T_CRE_RowAddr Store the row address cycle received. 1. More row address cycles required 2. All row address cycles received T_CRE_RowAddrWait T_CRE_WaitForCmd T_CRE_WaitForCmd Wait for a command cycle. 1. Command cycle E0h received T_CRE_ReturnToData T_CRE_ReturnToData The target performs the following actions: 1. Set tLunSelected to LUN selected by row address received. 2. Request that LUN tLunSelected select the column in the page register based on the column address received. 3. Indicate plane address received to tLunSelected for use in data output. 4. Request all idle LUNs not selected turn off their output 1 buffers. 1. tReturnState set to T_PP_MplWait T_PP_WaitForDataOut 2. tReturnState set to T_RD_Status_Off T_Idle_Rd 3. Else tReturnState NOTE: 1. LUNs not selected only turn off their output buffers if they are in an idle condition (SR[6] is one) when Change Read Column Enhanced is received. If LUNs are active (SR[6] is zero) when Change Read Column Enhanced is issued, then the host shall issue a Read Status Enhanced (78h) command prior to subsequent data output to ensure all LUNs that are not selected turn off their output buffers. 7.1.4. Reset command states T_RST_PowerOn 1. Unconditional T_RST_PowerOn_Exec 1. Unconditional The target performs the following actions: 1. tLastCmd set to FFh. 2. tbStatusOut is set to FALSE. 3. The target sends a Reset request to each LUN. T_RST_PowerOn_Exec The target performs the following actions: 1. Target level reset actions are performed. 2. R/B_n is set to zero. T_RST_Perform 178 1 The target performs the following actions: 1. tLastCmd set to FFh. 2. The target selects the asynchronous data interface. 3. The target sends a Reset request to each LUN. 4. Set tbChgCol to FALSE. 5. Set tbChgColEnh to FALSE. 6. Request all LUNs invalidate page register(s). T_RST_Perform T_RST_Execute 1. Unconditional NOTE: 1. This state is entered as a result of receiving a Reset (FFh) command in any other state, except if this is the first Reset after power-on. T_RST_Execute_Sync 1 The target performs the following actions: 1. tLastCmd set to FCh. 2. tbStatusOut is set to FALSE. 3. The target sends a Reset request to each LUN. 4. Set tbChgCol to FALSE. 5. Set tbChgColEnh to FALSE. 6. Request all LUNs invalidate page register(s). T_RST_Perform 1. Unconditional NOTE: 1. This state is entered as a result of receiving a Synchronous Reset (FCh) command in any other state. T_RST_Execute_LUN 1 1. Unconditional The target performs the following actions: 1. tLastCmd set to FAh. 2. tbStatusOut is set to FALSE. 3. Set tbChgCol to FALSE. 4. Set tbChgColEnh to FALSE. 5. Wait for an address cycle. 179 T_RST_LUN_Addr T_RST_LUN_AddrWait Wait for an address cycle. 1. Address cycle received T_RST_LUN_Addr T_RST_LUN_Addr Store the address cycle received. 1. More address cycles required 2. All address cycles received T_RST_LUN_AddrWait T_RST_LUN_Perform T_RST_LUN_Perform The target performs the following actions: 1. The target sends a Reset request to the addressed LUN. 2. R/B_n is cleared to zero. 3. Request the addressed LUN invalidate its page register. 1. Addressed LUN reset actions are complete and T_Idle tbStatusOut is set to FALSE 2. Addressed LUN reset actions are complete and T_Idle_Rd tbStatusOut is set to TRUE T_RST_Perform The target performs the following actions: 1. Target level reset actions are performed. 2. R/B_n is set to zero. 3. tReturnState set to T_RST_Perform. 1. Target and LUN reset actions are complete T_RST_End 2. Command cycle 70h (Read Status) received T_RS_Execute 3. Read request received and tbStatusOut is set to TRUE T_Idle_Rd_Status T_RST_End The target performs the following actions: 1. R/B_n is set to one. 1. tbStatusOut is set to FALSE 2. tbStatusOut is set to TRUE T_Idle T_Idle_Rd 7.1.5. Read ID command states T_RID_Execute The target performs the following actions: 1. tLastCmd set to 90h. 2. Wait for an address cycle. 3. Set tbChgCol to FALSE. 4. Set tbChgColEnh to FALSE. 5. Request all LUNs invalidate page register(s). 1. Address cycle of 00h received T_RID_Addr_00h 2. Address cycle of 20h received T_RID_Addr_20h 180 T_RID_Addr_00h Wait for the read request. 1. Read byte request received 2. Command cycle received T_RID_ManufacturerID T_Cmd_Decode T_RID_ManufacturerID Return the JEDEC manufacturer ID. 1. Read byte request received 2. Command cycle received T_RID_DeviceID T_Cmd_Decode T_RID_DeviceID Return the device ID. 1 1. Unconditional T_Idle_Rd NOTE: 1. Reading bytes beyond the device ID returns vendor specific values. T_RID_Addr_20h Wait for the read request. 1. Read byte request received 2. Command cycle received T_RID_Signature T_Cmd_Decode 1 T_RID_Signature Return next ONFI signature byte. 1. Last ONFI signature byte returned 2. Else NOTE: 1. Reading beyond the fourth byte returns indeterminate values. T_Idle_Rd T_RID_Addr_20h 7.1.6. Read Parameter Page command states T_RPP_Execute The target performs the following actions: 1. tLastCmd set to ECh. 2. Set tbChgCol to TRUE. 3. Set tbChgColEnh to FALSE. 4. Wait for an address cycle. 5. Request all LUNs invalidate page register(s). 6. Target selects LUN to execute parameter page read, sets tLunSelected to the address of this LUN. 1. Address cycle of 00h received T_RPP_ReadParams 181 T_RPP_ReadParams The target performs the following actions: 1. Request LUN tLunSelected clear SR[6] to zero. 2. R/B_n is cleared to zero. 3. Request LUN tLunSelected make parameter page data available in page register. 4. tReturnState set to T_RPP_ReadParams_Cont. 1. Read of page complete T_RPP_Complete 2. Command cycle 70h (Read Status) received T_RS_Execute 3. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status T_RPP_ReadParams_Cont 1. Read of page complete 2. Command cycle 70h (Read Status) received 3. Read request received and tbStatusOut set to TRUE T_RPP_Complete 1. Unconditional T_RPP_Complete T_RS_Execute T_Idle_Rd_Status Request LUN tLunSelected set SR[6] to one. R/B_n is set to one. T_Idle_Rd 7.1.7. Read Unique ID command states T_RU_Execute The target performs the following actions: 1. tLastCmd set to EDh. 2. Set tbChgCol to TRUE. 3. Set tbChgColEnh to FALSE. 4. Request all LUNs invalidate page register(s). 5. Wait for an address cycle. 6. Target selects LUN to execute unique ID read, sets tLunSelected to the address of this LUN. 1. Address cycle of 00h received T_RU_ReadUid T_RU_ReadUid The target performs the following actions: 1. Request LUN tLunSelected clear SR[6] to zero. 2. R/B_n is cleared to zero. 3. Request LUN tLunSelected make Unique ID data available in page register. 4. tReturnState set to T_RU_ReadUid. 1. LUN tLunSelected indicates data available in page T_RU_Complete register 2. Command cycle 70h (Read Status) received T_RS_Execute 3. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status T_RU_Complete 1. Unconditional Request LUN tLunSelected set SR[6] to one. R/B_n is set to one. T_Idle_Rd 182 7.1.8. Page Program and Page Cache Program command states T_PP_Execute The target performs the following actions: 1. tLastCmd set to 80h. 2. If R/B_n is cleared to zero, then tbStatus78hReq is set to TRUE. 3. If the program page register clear enhancement is not supported or disabled, request all LUNs clear their page 1 register(s). 1. Unconditional T_PP_AddrWait NOTE: 1. Idle LUNs may choose to not clear their page register if the Program is not addressed to that LUN. T_PP_Copyback 1. Unconditional If R/B_n is cleared to zero, then tbStatus78hReq is set to TRUE. T_PP_AddrWait T_PP_AddrWait Wait for an address cycle. 1. Address cycle received T_PP_Addr T_PP_Addr Store the address cycle received. 1. More address cycles required 2. All address cycles received T_PP_AddrWait T_PP_LUN_Execute 183 T_PP_LUN_Execute 1. Unconditional The target performs the following actions: 1. tLunSelected is set to the LUN indicated by the row address received. 2. If the program page register clear enhancement is enabled, request LUN tLunSelected clear the page register for the plane address specified. 3. Target issues the Program with associated address to the LUN tLunSelected. T_PP_LUN_DataWait T_PP_LUN_DataWait 1. 2. 3. 4. Wait for data byte/word or command cycle to be received from the host. Data byte/word received from the host T_PP_LUN_DataPass Command cycle of 15h received and tCopyback set to T_PP_Cmd_Pass FALSE Command cycle of 10h or 11h received T_PP_Cmd_Pass Command cycle of 85h received T_PP_ColChg T_PP_LUN_DataPass 1. Unconditional Pass data byte/word received from host to LUN tLunSelected T_PP_LUN_DataWait T_PP_Cmd_Pass Pass command received to LUN tLunSelected 1. Command passed was 11h T_PP_MplWait 2. Command passed was 10h or 15h T_Idle 184 T_PP_MplWait Wait for next Program to be issued. tReturnState set to T_PP_MplWait. 1 T_PP_AddrWait 1. Command cycle of 85h received 2 T_PP_AddrWait 2. Command cycle of 80h received and tCopyback set to FALSE 3. Command cycle of 05h received T_CR_Execute 4. Command cycle of 06h received T_CRE_Execute 5. Command cycle of 70h received T_RS_Execute 6. Command cycle of 78h received T_RSE_Execute 7. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status NOTE: 1. If the 85h is part of a Copyback, Change Row Address, or Small Data Move operation, then the LUN address and plane address shall be the same as the preceding Program operation. If the 85h is part of a Small Data Move operation, then the page address shall also be the same as the preceding Program operation. 2. Address cycles for the Program operation being issued shall have the same LUN address and page address as the preceding Program operation. The plane address shall be different than the one issued in the preceding Program operation. T_PP_ColChg Wait for column address cycle. 1. Address cycle received T_PP_ColChg_Addr T_PP_ColChg_Addr Store the address cycle received. 1. More column address cycles required 2. All address cycles received T_PP_ColChg T_PP_ColChg_LUN T_PP_ColChg_LUN 1. Unconditional Request that LUN tLunSelected change column address to column address received. T_PP_ColChg_Wait T_PP_ColChg_Wait 1. 2. 3. 4. 5. Wait for an address cycle, data byte/word, or command cycle to be received from the host Address cycle received T_PP_RowChg_Addr Data byte/word received from the host T_PP_LUN_DataPass Command cycle of 15h received and tCopyback set to T_PP_Cmd_Pass FALSE Command cycle of 10h or 11h received T_PP_Cmd_Pass Command cycle of 85h received T_PP_ColChg T_PP_RowChg Wait for row address cycle. 1. Address cycle received 185 T_PP_RowChg_Addr T_PP_RowChg_Addr Store the address cycle received. 1. More row address cycles required 2. All address cycles received T_PP_RowChg_LUN T_PP_RowChg T_PP_RowChg_LUN Request that LUN tLunSelected change row address to row address received. 1 1. Unconditional T_PP_LUN_DataWait NOTE: 1. The LUN address and plane address shall be the same as previously specified for the Program operation executing. T_PP_WaitForDataOut Wait for read request (data or status) or other action. tReturnState set to T_PP_WaitForDataOut. 1. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status 2. Read request received and tbStatus78hReq set to T_Idle_Rd_LunData 1 FALSE 3. Command cycle of 70h received T_RS_Execute 4. Command cycle of 78h received T_RSE_Execute 5. Command cycle of 00h received T_RD_Execute 6. Command cycle received T_PP_MplWait NOTE: 1. When tbStatus78hReq is set to TRUE, a Read Status Enhanced (78h) command followed by a 00h command shall be issued by the host prior to reading data from a particular LUN. 7.1.9. Block Erase command states T_BE_Execute The target performs the following actions: 1. tLastCmd set to 60h. 2. If R/B_n is cleared to zero, then tbStatus78hReq is set to TRUE. 3. Wait for a row address cycle. 1. Address cycle received T_BE_Addr T_BE_Addr Store the row address cycle received. 1. More address cycles required 2. All address cycles received T_BE_LUN_Execute 1. Unconditional T_BE_Execute T_BE_LUN_Execute tLunSelected is set to the LUN indicated by the row address received. Target issues the Erase with associated row address to the LUN tLunSelected. T_BE_LUN_Confirm 186 T_BE_LUN_Confirm Wait for D0h or D1h command cycle. 1. Command cycle of D0h or D1h received T_BE_Cmd_Pass T_BE_Cmd_Pass Pass command received to LUN tLunSelected 1. Command passed was D1h T_BE_MplWait 2. Command passed was D0h T_Idle T_BE_MplWait Wait for next Erase to be issued. tReturnState set to T_BE_MplWait. 1. Command cycle of 60h received T_BE_Execute 2. Command cycle of 70h received T_RS_Execute 3. Command cycle of 78h received T_RSE_Execute 4. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status 187 7.1.10. Read command states T_RD_Execute 1. tbStatusOut set to TRUE 2. Else T_RD_StatusOff T_RD_AddrWait T_RD_StatusOff tbStatusOut set to FALSE. tReturnState set to T_RD_StatusOff. 1. Address cycle received T_RD_Addr 2. Read request received and tLastCmd set to 80h T_PP_WaitForDataOut 3. Read request received and tLastCmd set to EEh T_Idle_Rd_XferHost 4. Read request received T_Idle_Rd_LunData 5. Command cycle of 05h received T_CR_Execute 6. Command cycle of 06h received T_CRE_Execute T_RD_AddrWait tLastCmd set to 00h. Set tbChgCol to TRUE. Set tbChgColEnh to TRUE. If R/B_n is cleared to zero, then tbStatus78hReq is set to TRUE. Wait for an address cycle. 1. Address cycle received T_RD_Addr T_RD_Addr Store the address cycle received. 3. More address cycles required 4. All address cycles received T_RD_LUN_Execute T_RD_AddrWait T_RD_LUN_Execute The target performs the following actions: 1. tLunSelected is set to the LUN indicated by the row address received. 2. Issues the Read Page with address to LUN tLunSelected. 3. Requests all idle LUNs not selected to turn off their output buffers. 1 1. Unconditional T_RD_LUN_Confirm NOTE: 1. LUNs not selected will only turn off their output buffers if they are in an Idle state. If other LUNs are active, the host shall issue a Read Status Enhanced (78h) command to ensure all LUNs that are not selected turn off their output buffers prior to issuing the Read (00h) command. T_RD_LUN_Confirm Wait for 30h, 31h, 32h, or 35h to be received. 1. Command cycle of 30h, 31h, 32h, or 35h received T_RD_Cmd_Pass T_RD_Cmd_Pass Pass command received to LUN tLunSelected 1. Command passed was 35h T_RD_Copyback 2. Command passed was 30h, 31h, or 32h T_Idle_Rd 188 T_RD_Copyback tCopyback set to TRUE. tReturnState set to T_RD_Copyback. 1. Command cycle of 00h received T_RD_Execute 2. Command cycle of 05h received T_CR_Execute 3. Command cycle of 06h received T_CRE_Execute 4. Command cycle of 85h received T_PP_Copyback 5. Command cycle of 70h received T_RS_Execute 6. Command cycle of 78h received T_RSE_Execute 7. LUN indicates its SR[6] value transitions T_Idle_RB_Transition 8. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status 9. Read request received T_Idle_Rd_LunData 189 7.1.11. Set Features command states T_SF_Execute The target performs the following actions: 1. tLastCmd set to EFh. 2. Request all LUNs invalidate page register(s). 3. Wait for an address cycle. 1. Address cycle received T_SF_Addr T_SF_Addr 1. Unconditional Store the feature address received. T_SF_WaitForParams T_SF_WaitForParams Wait for data byte to be received. 1. Data byte written to target T_SF_StoreParam T_SF_StoreParam Store parameter received. 1. More parameters required 2. All parameters received T_SF_WaitForParams T_SF_Complete T_SF_Complete The target performs the following actions: 1. Request LUN tLunSelected clear SR[6] to zero. 2. R/B_n is cleared to zero. 3. Finish Set Features command. 4. tReturnState set to T_SF_Complete. 1. Set Features command complete T_SF_UpdateStatus 2. Command cycle 70h (Read Status) received T_RS_Execute 3. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status T_SF_UpdateStatus The target performs the following actions: 1. Request LUN tLunSelected set SR[6] to one. 2. R/B_n is set to one. 1. tbStatusOut is set to FALSE T_Idle 2. tbStatusOut is set to TRUE T_Idle_Rd 190 7.1.12. Get Features command states T_GF_Execute The target performs the following actions: 1. tLastCmd set to EEh. 2. Request all LUNs invalidate page register(s). 3. Set tbChgCol to FALSE. 4. Set tbChgColEnh to FALSE. 5. Wait for an address cycle. 1. Address cycle received T_GF_Addr T_GF_Addr 1. Unconditional Store the feature address received. T_GF_RetrieveParams T_GF_RetrieveParams The target performs the following actions: 1. Request LUN tLunSelected clear SR[6] to zero. 2. R/B_n is cleared to zero. 3. Retrieve parameters. 4. tReturnState set to T_GF_RetrieveParams. 1. Parameters are ready to be transferred to the host T_GF_Ready 2. Command cycle 70h (Read Status) received T_RS_Execute 3. Read request received and tbStatusOut set to TRUE T_Idle_Rd_Status T_GF_Ready 1. Unconditional 7.1.13. Request LUN tLunSelected set SR[6] to one. R/B_n is set to one. T_Idle_Rd Read Status command states T_RS_Execute 1 T_RS_Perform 1. tbStatus78hReq is set to FALSE NOTE: 1. When tbStatus78hReq is set to TRUE, issuing a Read Status (70h) command is illegal. T_RS_Perform The target performs the following actions: 1. tbStatusOut is set to TRUE. 2. Indicate 70h command received to LUN tLunSelected. 1. tReturnState set to T_Idle T_Idle_Rd 2. Else tReturnState 191 7.1.14. Read Status Enhanced command states 1 tbStatus78hReq is set to FALSE. tbStatusOut is set to TRUE. Wait for a row address cycle. 1. Row address cycle received T_RSE_Addr NOTE: 1. The host should not issue Read Status Enhanced following a Target level command (Reset, Read ID, Read Parameter Page, Read Unique ID, Set Features, Get Features). The status value read from the LUN selected with Read Status Enhanced may not correspond with the LUN selected during the Target level command. T_RSE_Execute T_RSE_Addr Store the row address cycle received. 1. More row address cycles required 2. All row address cycles received T_RSE_Select T_RSE_Execute T_RSE_Select The target performs the following actions: 1. Set tLunSelected to LUN selected by row address received. 2. Indicate 78h command and row address received to all LUNs. 1. tReturnState set to T_Idle T_Idle_Rd 2. Else tReturnState 192 7.2. LUN behavioral flows The LUN state machine describes the allowed sequences when operating with the LUN. If none of the arcs are true, then the LUN remains in the current state. 7.2.1. Variables This section describes variables used within the LUN state machine. lunStatus This variable contains the current LUN status register value contents. for this variable is 00h. The power on value lunFail[] This array contains the FAIL and FAILC bits for each interleave address. For example, lunFail[3][1] contains the FAILC bit for plane address 3. The power on value for each variable in this array is 00b. lunLastConfirm This variable contains the last confirm command cycle (30h, 31h, 32h, 35h, 10h, 15h, 11h, D0h, D1h). The power on value for this variable is FFh. lunOutputMpl This variable contains the plane address requested for data output. The power on value for this variable is 0h. lunReturnState This variable contains the state to return to after status operations. The power on value for this variable is L_Idle. lunStatusCmd This variable contains the last status command received. variable is 70h. The power on value for this lunStatusMpl This variable contains the plane address indicated in a previous 78h command. The power on value for this variable is 0h. lunbInterleave This variable is set to one when the LUN is performing a multi-plane operation. The power on value for this variable is FALSE. lunbMplNextCmd This variable is set to TRUE when the LUN is ready to receive the next multi-plane command. lunEraseAddr[] This variable contains the block addresses of erases that have been suspended. 7.2.2. Idle command states 1 L_Idle lunReturnState is set to L_Idle. 1. Target request received L_Idle_TargetRequest NOTE: 1. This state is entered as a result of a power-on event when Vcc reaches Vcc_min. 193 L_Idle_TargetRequest If Target indicates an address, the address is stored by the LUN. 1. Target requests LUN perform a Reset L_RST_Execute 2. Target indicates WP_n value L_WP_Update 3. Target requests SR register update L_SR_Update 4. Target requests status or status command received L_Status_Execute 5. Target indicates plane address for use in data output L_Idle_Mpl_DataOutAdd r 6. Target indicates output buffer should be turned off L_Idle 7. Target requests page register clear L_Idle_ClearPageReg 8. Target requests page register invalidate L_Idle_InvalidPageReg 9. Target indicates Program request for this LUN L_PP_Execute 10. Target indicates Erase request for this LUN L_BE_Execute 11. Target indicates Erase Resume request for this LUN L_ER_Execute 12. Target indicates Read Page request for this LUN L_RD_Addr 13. Target indicates Read Parameter Page request L_Idle_RdPp 14. Target indicates Read Unique ID request L_Idle_RdUid L_WP_Update 1. Unconditional Set lunStatus[7] to the WP_n value indicated by the target. lunReturnState L_SR_Update 1. Unconditional Update lunStatus as indicated by the target. L_Idle_Mpl_DataOutAddr 1. Unconditional Set lunOutputMpl to plane address indicated by the target. lunReturnState L_Idle_ClearPageReg 1. Unconditional Set page register to all ones value. L_Idle_InvalidPageReg 1. Unconditional Invalidate page register. L_Idle_RdPp lunReturnState lunReturnState lunReturnState The LUN performs the following actions: 1. LUN reads parameter page data into the page register. 2. lunReturnState set to L_Idle_RdPp_Cont. 1. Parameter page data transferred to page register L_Idle_RdPp_End 2. Target requests status or status command received L_Status_Execute 194 L_Idle_RdPp_Cont 1. Parameter page data transferred to page register 2. Target requests status or status command received L_Idle_RdPp_End 1. Unconditional L_Idle_RdPp_End L_Status_Execute LUN indicates to Target that parameter page data is in page register. L_Idle_Rd L_Idle_RdUid The LUN performs the following actions: 1. LUN reads Unique ID data into the page register. 2. lunReturnState set to L_Idle_RdUid. 1. Unique ID data transferred to page register L_Idle_RdUid_End 2. Target requests status or status command received L_Status_Execute L_Idle_RdUid_End 1. Unconditional LUN indicates to Target that Unique ID data is in page register. L_Idle_Rd 7.2.3. Idle Read states L_Idle_Rd 1. 2. 3. 4. 5. lunReturnState is set to L_Idle_Rd. Background read operation complete Target requests column address be selected Read request received from Target Command cycle 31h (Read Cache Sequential) received Command cycle 3Fh (Read Cache End) received and lunLastConfirm is 31h 6. Target request received L_Idle_Rd_Finish 1. Unconditional L_Idle_Rd_Finish L_Idle_Rd_ColSelect L_Idle_Rd_Xfer L_RD_Cache_Next L_RD_Cache_Xfer_End L_Idle_TargetRequest L_Idle_Rd Set lunStatus[5] to one. L_Idle_Rd_Xfer Return to the target the next byte (x8) or word (x16) of data from page register based on Target requested. Increments column address. 1. lunReturnState set to L_PP_Mpl_Wait L_PP_Mpl_Wait 2. Unconditional L_Idle_Rd L_Idle_Rd_ColSelect Select the column in the page register based on the column address received from the target. 1. lunReturnState set to L_PP_Mpl_Wait L_PP_Mpl_Wait 2. Unconditional L_Idle_Rd 195 7.2.4. Status states L_Status_Execute 1. Target requests status value 2. Target indicates 78h was received 3. Target indicates 70h was received L_Status_Value 1. lunbInterleave set to TRUE and lunStatusCmd set to 70h 2. lunbInterleave set to TRUE and lunStatusCmd set to 78h 3. lunbInterleave set to FALSE L_Status_Enhanced 1. LUN in row address indicated matches this LUN 2. Else L_Status_Value L_Status_Enhanced L_Status_Legacy L_Status_Mpl_Comp L_Status_Mpl_Addr L_Status_Lun L_Status_Record_78h L_Status_Output_Off L_Status_Record_78h lunStatusCmd is set to 78h and lunStatusMpl is set to plane address indicated by Target. The LUN turns on its output buffer. 1. lunReturnState set to L_Idle and (lunLastConfirm set to L_Idle_Rd 30h, 31h, 32h, or 35h) 2. Else lunReturnState L_Status_Output_Off LUN turns off its output buffer. 1. lunReturnState set to L_Idle_Rd 2. Else L_Idle lunReturnState L_Status_Legacy 1. Unconditional lunStatusCmd is set to 70h. lunReturnState L_Status_Mpl_Comp The LUN composes the status value to return as shown: status[7:2] = lunStatus[7:2] status[1] = for all x, OR of lunFail[x][1] status[0] = for all x, OR of lunFail[x][0] Return status to the Target. 1. Unconditional L_Status_Mpl_Addr lunReturnState The LUN composes the status value to return as shown: status[7:2] = lunStatus[7:2] status[1:0] = lunFail[lunStatusMpl][1:0] 196 Return status to the Target. 1. Unconditional L_Status_Lun 1. Unconditional lunReturnState lunReturnState Return lunStatus to the Target. 7.2.5. Reset states 1 The LUN performs the following actions: 1. lunStatus[6] is cleared to zero. 2. lunStatus[6] value is indicated to the Target. 3. Perform reset of the LUN. 4. lunbInterleave is set to FALSE. 5. lunReturnState is set to L_RST_Execute. 1. Reset of the LUN is complete L_RST_Complete 2. Target requests status or status command received L_Status_Execute NOTE: 1. This state is entered as a result of receiving an indication from the Target state machine to perform a Reset in any other state. L_RST_Execute L_RST_Complete 1. Unconditional The LUN performs the following actions: 1. lunStatus[1:0] are cleared to 00b. 2. For all plane addresses x, clear lunFail[x][1:0] to 00b. 3. lunStatus[6] is set to one. 4. lunStatus[6] value is indicated to the Target. 5. Indicate to the Target state machine that Reset for this LUN is complete. L_Idle 7.2.6. Block Erase command states L_BE_Execute 1. Unconditional lunbInterleave set to FALSE. L_BE_WaitForCmd Wait for a command cycle. 1. Command cycle D0h received 2. Command cycle D1h received L_BE_Erase L_BE_WaitForCmd L_BE_Erase L_BE_Mpl The LUN performs the following actions: 1. lunStatus[6] is cleared to zero. 2. If lunbInterleave is TRUE, lunStatus[5] is cleared to zero. 3. lunStatus[6] value is indicated to the Target. 4. lunLastConfirm set to D0h. 5. Erase the requested block and any previously requested blocks if lunbInterleave is set to TRUE and concurrent 197 interleaving is supported. 1. Unconditional L_BE_Erase_Wait lunReturnState set to L_BE_Erase_Wait. 1. Erase of requested block(s) complete and lunbInterleave set to TRUE 2. Erase of requested block complete 3. Target requests page register clear 4. Target requests status or status command received L_BE_Mpl 1. Unconditional 1. Unconditional L_BE_Mpl_Overlap 1. Unconditional L_BE_Mpl_Sts L_BE_Sts L_Idle_ClearPageReg L_Status_Execute The LUN performs the following actions in the order specified: 1. lunbInterleave set to TRUE. 2. lunLastConfirm set to D1h. 3. lunStatus[6:5] is cleared to 00b. lunStatus[6] value is indicated to the Target. 4. LUN begins erasing block specified if overlapped is supported. 5. lunbMplNextCmd is set to FALSE. 6. LUN prepares to receive the next block to erase. L_BE_Mpl_Wait L_BE_Mpl_Wait lunReturnState set to L_BE_Mpl_Wait. 1. An overlapped multi-plane Erase completed 2. Ready to receive the next Erase command and lunbMplNextCmd is set to FALSE 3. Target indicates Erase request for this LUN and lunbMplNextCmd is set to TRUE 4. Target requests status or status command received L_BE_Mpl_NextCmd L_BE_Erase_Wait L_BE_Mpl_Overlap L_BE_Mpl_NextCmd L_BE_WaitForCmd L_Status_Execute The LUN performs the following actions in the order specified: 1. lunbMplNextCmd is set to TRUE. 2. If no array operations are in progress, lunStatus[5] is set to one. 3. lunStatus[6] is set to one. lunStatus[6] value is indicated to the Target. L_BE_Mpl_Wait The LUN performs the following actions in the order specified for the overlapped multi-plane operation that completed: 1. mplComplete set to plane address of completed operation 2. lunFail[mplComplete][0] is set to program status of operation. If all array operations are complete, lunStatus[5] is set to one. lunReturnState 198 L_BE_Sts 1. Unconditional L_BE_Mpl_Sts 1. Unconditional The LUN performs the following actions in the order specified: 1. lunStatus[0] is set to erase status. 2. lunStatus[6] is set to one. lunStatus[6] value is indicated to the Target. L_Idle The LUN performs the following actions in the order specified for each multi-plane operation that completed: 1. mplComplete set to interleave address of completed operation. 2. lunFail[mplComplete][0] is set to erase status value. lunStatus[6:5] is set to 11b and lunStatus[6] value is indicated to the Target. L_Idle 7.2.7. Read command states If caching is not supported, then all actions for status bit 5 are ignored. L_RD_Addr 1. Unconditional The LUN performs the following actions in the order specified: 1. Records address received from the target. 2. If multi-plane addressing is supported, selects the correct page register based on the plane address. 3. Selects the column in the page register based on the column address received. L_RD_WaitForCmd L_RD_WaitForCmd lunbInterleave set to FALSE. Wait for a command cycle. 1. Command cycle 30h or 35h received L_RD_ArrayRead 2. Command cycle 31h received and lunLastConfirm equal L_RD_Cache_Xfer to 30h or 31h 3. Command cycle 32h received L_RD_Mpl_Xfer L_RD_ArrayRead The LUN performs the following actions: 1. lunStatus[6:5] is cleared to 00b. 2. lunStatus[6] value is indicated to the target. 3. lunLastConfirm set to last command cycle (30h or 35h). 4. Read the requested page from the array. If concurrent multiplane operation, read all pages requested from the array. 5. lunReturnState set to L_RD_ArrayRead_Cont. 1. Read of requested page(s) complete L_RD_Complete 2. Target requests status or status command received L_Status_Execute 199 L_RD_ArrayRead_Cont 1. Read of requested page(s) complete 2. Target requests status or status command received L_RD_Complete 1. Unconditional L_RD_Cache_Next 1. Unconditional L_RD_Complete L_Status_Execute lunStatus[6:5] is set to 11b. lunStatus[6] value is indicated to the target. L_Idle_Rd Select the next row address as the sequential increasing row address to the last page read. L_RD_Cache_Xfer L_RD_Cache_Xfer The LUN performs the following actions: 1. lunStatus[6:5] is cleared to 00b. lunStatus[6] value is indicated to the Target. 2. lunLastConfirm set to 31h. 3. Begin background read operation for selected address. 4. lunReturnState set to L_RD_Cache_Xfer. 1. Data available in page register for previous read L_RD_Cache_Sts operation 2. Target requests status or status command received L_Status_Execute L_RD_Cache_Xfer_End The LUN performs the following actions: 1. lunStatus[6] is cleared to zero. 2. lunStatus[6] value is indicated to the target. 3. lunLastConfirm set to 3Fh. 4. lunReturnState set to L_RD_Cache_Xfer_End. 1. Data available in page register for previous read L_RD_Cache_Sts_End operation 2. Target requests status or status command received L_Status_Execute L_RD_Cache_Sts 1. Unconditional L_RD_Cache_Sts_End 1. Unconditional lunStatus[6] is set to one. lunStatus[6] value is indicated to the Target. L_Idle_Rd lunStatus[6:5] is set to 11b. lunStatus[6] value is indicated to the Target. L_Idle_Rd 200 L_RD_Mpl_Xfer The LUN performs the following actions: 1. lunStatus[6:5] is cleared to 00b. 2. lunStatus[6] value is indicated to the target. 3. lunLastConfirm set to 32h. 4. lunbMplNextCmd is set to FALSE. 5. LUN begins reading page specified if overlapped interleaving is supported. 6. Prepare to receive the next page to read. 7. lunReturnState set to L_RD_Mpl_Xfer. 1. Target ready to receive next page to read L_RD_Mpl_Wait 2. Target requests status or status command received L_Status_Execute L_RD_Mpl_Wait lunStatus[6] is set to one. lunStatus[6] value is indicated to the Target. lunReturnState set to L_RD_Mpl_Wait. 1. An overlapped multi-plane Read completed L_RD_Mpl_Overlap 2. Target indicates Read Page request for this LUN L_RD_Addr 3. Target requests status or status command received L_Status_Execute L_RD_Mpl_Overlap 1. Unconditional The LUN performs the following actions in the order specified for the overlapped multi-plane operation that completed: 1. mplComplete set to plane address of completed operation. If all array operations are complete, lunStatus[5] is set to one. lunReturnState 7.2.8. Page Program and Page Cache Program command states If caching or overlapped interleaving is not supported, then all actions for status bit 5 are ignored. If caching is not supported, then all actions for status bit 1 are ignored. L_PP_Execute 1. Unconditional lunbInterleave set to FALSE. L_PP_Addr The LUN performs the following actions in the order specified: 1. Records address received from the Target. 2. If multi-plane addressing is supported, selects the correct page register based on the plane address. 3. Selects the column in the page register based on the column address received. L_PP_WaitForData 1. Unconditional L_PP_WaitForData L_PP_Addr Wait for data to be received. lunReturnState is set to L_PP_WaitForData. 1. Target passes data byte or word to LUN L_PP_AcceptData 2. Command cycle 10h (program execute) received L_PP_Prog 201 3. 4. 5. 6. Command cycle 15h (cache program) received Command cycle 11h (interleave) received Target requests column address be selected Target requests row address be selected L_PP_AcceptData 1. Unconditional L_PP_Prog 1. Unconditional Write the byte (x8) or word (x16) of data into the selected column address in the page register. Increments column address. L_PP_WaitForData The LUN performs the following actions in the order specified: 1. lunStatus[6:5] is cleared to 00h. lunStatus[6] value is indicated to the Target. 2. lunLastConfirm set to 10h. 3. If only one page is specified to be programmed, clear lunbInterleave to FALSE. 4. LUN begins programming page specified and any previous pages specified if lunbInterleave is TRUE and concurrent interleaving is supported. L_PP_Prog_Wait L_PP_Prog_Wait lunReturnState set to L_PP_Prog_Wait. 1. Write of all requested pages are complete and lunbInterleave is set to TRUE 2. Write of requested page is complete and lunbInterleave is cleared to FALSE 3. Target requests status or status command received L_PP_Cache 1. Unconditional L_PP_Cache L_PP_Mpl L_PP_ColSelect L_PP_RowSelect L_PP_Mpl_Sts L_PP_Sts L_Status_Execute The LUN performs the following actions in the order specified: 1. lunStatus[6:5] is cleared to 00b. lunStatus[6] value is indicated to the Target. 2. lunLastConfirm set to 15h. 3. Wait for the page register to become available for data input. 4. Start background program operation. L_PP_Cache_Wait L_PP_Cache_Wait lunReturnState is set to L_PP_Cache_Wait. 1. Page register available for data input L_PP_CacheRdy 2. Target requests status or status command received L_Status_Execute L_PP_CacheRdy The LUN performs the following actions: 1. If lunbInterleave is set to FALSE, then lunStatus[1] is set to the value of lunStatus[0]. 2. If lunbInterleave is set to TRUE, then for all multi-plane addresses, x, lunFail[x][1] is set to the value of lunFail[x][0]. 3. lunStatus[6] is set to one. lunStatus[6] value is indicated to the Target. 202 1. Unconditional L_PP_CacheRdy_Wait L_PP_CacheRdy_Wait lunReturnState set to L_PP_CacheRdy_Wait. 1. Previous cache operation complete and lunbInterleave L_PP_Mpl_Cache_Sts set to TRUE 2. Previous cache operation complete L_PP_Cache_Sts 3. Target indicates Program request for this LUN L_PP_Addr 4. Target requests page register clear L_Idle_ClearPageReg 5. Target requests status or status command received L_Status_Execute L_PP_Mpl 1. Unconditional The LUN performs the following actions in the order specified: 1. lunbInterleave set to TRUE. 2. lunStatus[6:5] is cleared to 00b. lunStatus[6] value is indicated to the Target. 3. lunLastConfirm set to 11h. 4. lunbMplNextCmd is set to FALSE. 5. LUN begins programming page specified if overlapped interleaving is supported. L_PP_Mpl_Wait L_PP_Mpl_Wait lunReturnState set to L_PP_Mpl_Wait. 1. An overlapped multi-plane Program completed 2. A previous cache Program completed 3. LUN is ready to receive the next Program command and lunbMplNextCmd is set to FALSE 4. Target indicates Program request for this LUN and lunbMplNextCmd is set to TRUE 5. Target requests column address be selected 6. Target indicates plane address for use in data output 7. Target requests status or status command received 8. Read request received from Target L_PP_Mpl_NextCmd 1. Unconditional L_PP_Sts L_PP_Mpl_Overlap L_PP_Mpl_Cache_Sts L_PP_Mpl_NextCmd L_PP_Addr L_Idle_Rd_ColSelect L_Idle_Mpl_DataOutAdd r L_Status_Execute L_Idle_Rd_Xfer The LUN performs the following actions in the order specified: 1. lunbMplNextCmd is set to TRUE. 2. If no array operations are in progress, lunStatus[5] is set to one. 3. lunStatus[6] is set to one. lunStatus[6] value is indicated to the Target. L_PP_Mpl_Wait The LUN performs the following actions in the order specified: 1. lunStatus[1] is set to program status of previous operation 2. lunStatus[0] is set to program status of final operation 3. lunStatus[6:5] is set to 11b. 4. lunStatus[6] value is indicated to the Target. 203 1. Unconditional L_PP_Cache_Sts 1. Unconditional L_PP_Mpl_Cache_Sts 1. Unconditional L_PP_Mpl_Overlap 1. Unconditional L_PP_Mpl_Sts 1. Unconditional L_PP_ColSelect 1. Unconditional L_PP_RowSelect 1. Unconditional L_Idle The LUN performs the following actions in the order specified: 1. lunStatus[0] is set to program status. 2. lunStatus[5] is set to one. lunReturnState The LUN performs the following actions in the order specified for all completed cache operations: 1. mplAddr set to interleave address of cache operation. 2. lunFail[mplAddr][0] is set to program status. If all array operations are complete, lunStatus[5] is set to one. lunReturnState The LUN performs the following actions in the order specified for the overlapped multi-plane operation that completed: 1. mplComplete set to interleave address of completed operation 2. lunFail[mplComplete][0] is set to program status of operation. If all array operations are complete, lunStatus[5] is set to one. lunReturnState The LUN performs the following actions in the order specified for each multi-plane operation that completed: 1. mplComplete set to plane address of completed operation 2. lunFail[mplComplete][1] is set to program status of previous operation. 3. lunFail[mplComplete][0] is set to program status of final operation. lunStatus[6:5] is set to 11b and lunStatus[6] value is indicated to the Target. L_Idle Select the column in the page register based on the column address received that the target requested. L_PP_WaitForData Select the block and page to program based on the row address received from the target. L_PP_WaitForData 204 A. SAMPLE CODE FOR CRC-16 (INFORMATIVE) This section provides an informative implementation of the CRC-16 polynomial. The example is intended as an aid in verifying an implementation of the algorithm. int main(int argc, char* argv[]) { // Bit by bit algorithm without augmented zero bytes const unsigned long crcinit = 0x4F4E; // Initial CRC value in the shift register const int order = 16; // Order of the CRC-16 const unsigned long polynom = 0x8005; // Polynomial unsigned long i, j, c, bit; unsigned long crc = crcinit; // Initialize the shift register with 0x4F4E unsigned long data_in; int dataByteCount = 0; unsigned long crcmask, crchighbit; crcmask = ((((unsigned long)1<<(order-1))-1)<<1)|1; crchighbit = (unsigned long)1<<(order-1); // Input byte stream, one byte at a time, bits processed from MSB to LSB printf("Input byte value in hex(eg. 0x30):"); printf("\n"); 205 while(scanf("%x", &data_in) == 1) { c = (unsigned long)data_in; dataByteCount++; for (j=0x80; j; j>>=1) { bit = crc & crchighbit; crc<<= 1; if (c & j) bit^= crchighbit; if (bit) crc^= polynom; } crc&= crcmask; printf("CRC-16 value: 0x%x\n", crc); } printf("Final CRC-16 value: 0x%x, total data bytes: %d\n", crc, dataByteCount); return 0; } 206 B. SPARE SIZE RECOMMENDATIONS (INFORMATIVE) This appendix describes recommendations for the spare bytes per page based on the ECC requirements reported in the parameter page. These recommendations are for raw NAND implementations and do not apply to devices that support EZ NAND. Table 51 lists recommendations for 2KB, 4KB, and 8KB page size devices. Page Size Number of bits ECC correctability Spare Bytes Per Page Recommendation 2048 bytes <= 8 bits 64 bytes 2048 bytes > 8 bits 112 bytes 4096 bytes <= 8 bits 128 bytes 4096 bytes > 8 bits 218 or 224 bytes 8192 bytes 2 8192 bytes <= 8 bits 256 bytes > 8 bits 448 bytes NOTE: 1. The number of bits ECC correctability is based on a 512 byte codeword size. 2. If more correction is required than spare area size allows for with a 512 byte codeword size, it is recommended that the host use a larger ECC codeword size (e.g. 1KB, 2KB, etc). The device manufacturer may provide guidance on the ECC codeword size to use in the extended parameter page. Table 51 Spare Area Size Recommendations for raw NAND The host transfers bytes from the page register in discrete units that include data, metadata, and the ECC check bytes. This discrete unit is recommended to be an even number of bytes for devices that support the source synchronous data interface. As an example, assume the page size is 8192 bytes and the ECC codeword size used is 1KB. Then 1024 bytes of data will be transferred in each discrete unit, resulting in eight discrete units of data being transferred for this page. The spare bytes for this page should be allocated to allow enough storage for the metadata and check bytes, and should also be an even number when divided by eight (i.e. the number of discrete units contained in that page). 207 C. DEVICE SELF-INITIALIZATION WITH PSL (INFORMATIVE) Some devices store configuration information for the Flash array within the Flash array itself. The device loads this information either at power-on or during the first Reset after power-on. Vendors may choose to support PSL as one of the vendor specific pins. If PSL is supported, then it shall have the following behavior: PSL = 0 V: Configuration information is loaded at power-on. The IST current may be up to 15 mA and the time for R/B_n to become one is up to 5 ms. PSL = Vcc or not connected: Configuration information if supported is loaded during the first Reset after power-on. There is no change to the IST current requirement. This corresponds to the normally expected ONFI device operation. If PSL is not supported by the device, then the IST requirement shall be met. Refer to the device vendor’s datasheet to determine if self-initialization at power-on is supported. 208 D. ICC MEASUREMENT METHODOLOGY This section defines the technique used to measure the ICC parameters defined in section 2.10. The common testing conditions that shall be used to measure the DC and Operating Conditions are defined in Table 52. The testing conditions that shall be used to measure the DC and Operating Conditions that are data interface specific are defined in Table 53. Parameter Testing Condition 1. 2. 3. 4. 5. 6. 7. Vcc = Vcc(min) to Vcc(max) VccQ = VccQ(min) to VccQ(max) CE_n = 0 V WP_n = VccQ IOUT = 0 mA Measured across operating temperature range N data input or data output cycles, where N is the number of bytes or words in the page 8. No multi-plane operations 9. Sample sufficient number of times to remove measurement variability 10. Sample an equal ratio of page types that exist in a block. A page type is a group of page addresses and is commonly referred to as upper or lower page (or middle page for 3 bits per cell devices). 11. Choose the first good even/odd block pair beginning at blocks 2-3 General conditions Array preconditioning for ICC1 and ICC3 The array is preconditioned to match the data input pattern for ICC2. Fixed wait time (no R/B_n polling) ICC1: tR = tR(max) ICC2: tPROG = tPROG(max) ICC3: tBERS = tBERS(max) Table 52 Parameter AC Timing Parameters Bus idle data pattern Repeated data pattern (Used for ICC2 and ICC4W ) Array preconditioning for ICC4R Common Testing Conditions for ICC Asynchronous tWC = tWC(min) tRC = tRC(min) tADL = ~tADL(min) tCCS = ~tCCS(min) tRHW = ~tRHW(min) IO[7:0] = FFh IO[15:0] = FFFFh Source Synchronous tCK = tCK(avg) tADL = ~tADL(min) tCCS = ~tCCS(min) tRHW = ~tRHW(min) DQ[7:0] = FFh IO[7:0] = A5h, AAh, 5Ah, 55h IO[15:0] = A5A5h, AAAAh, 5A5Ah, 5555h DQ[7:0] = A5h, AAh, 5Ah, 55h The array is preconditioned to match the following repeating data pattern: IO[7:0] = A5h IO[15:8] = A5A5h The array is preconditioned to match the following repeating data pattern: DQ[7:0] = A5h NOTE: 1. The value of tCK(avg) used should be the minimum tCK(avg) of the timing modes supported for the device. The source synchronous timing modes supported by the device are indicated in the parameter page. Table 53 Data Interface Specific Testing Conditions for ICC 209 The following figures detail the testing procedure for ICC1, ICC2, ICC3, ICC4R, ICC4W, and ICC5. 210 Cycle Type CMD ADDR ADDR ADDR ADDR ADDR CMD IDLE 30h FFh Minimum cycle time between command, address cycles DQx 00h 00h 00h R2 R1 R3 tWB tR(max) R/B_n Repeat to gather sufficient ICC samples to remove measurement variability Figure 94 ICC1 measurement procedure To calculate the active current for ICC1, the following equations may be used. ( ( ( ( ) ) 211 ) ) ( ( ) ( ( ) ( ) ) ( ) ( ) ( ( ) ) ) Cycle Type ADDR CMD ADDR ADDR ADDR IDLE ADDR Minimum cycle time between command, address cycles DQx 00h 80h 00h R2 R1 DIN DIN DIN Cycle(min) tADL FFh R3 IDLE CMD D0 Dn ... FFh 10h tWB tPROG(max) R/B_n Repeat to gather sufficient ICC samples to remove measurement variability Figure 95 ICC2 measurement procedure To calculate the active current for ICC2, the following equations may be used. ( ( ) ( ( ) ( ) ) ( ( ( ( ( ) ( For the source synchronous data interface, the tIO value is calculated as: ) ( ( ( ( ( ) )] ) For the asynchronous interface, the tIO value is calculated as: 212 ) ) ( ) )) ( ( ) ) ) ( ) ) Cycle Type ADDR CMD ADDR ADDR CMD IDLE D0h FFh Min time between command, address cycles DQ[7:0] R1 60h R2 R3 tWB tBERS(max) R/B_n Repeat to gather sufficient ICC samples to remove measurement variability Figure 96 ICC3 measurement procedure To calculate the active current for ICC3, the following equations may be used. ( ( ( ) ( ) 213 ( ) ) ( ) ( ( ( ) ) ) ( ( ) ) ( ( ) ) ) A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR CMD IDLE 30h FFh Minimum cycle time between command, address cycles DQx 00h 00h 00h R1 R2 R3 tWB tR(max) R/B_n A Cycle Type CMD ADDR CMD ADDR Min time between command, address cycles DQx 05h 00h 00h IDLE DOUT DOUT DOUT tCCS E0h FFh tRHW D0 ... Dn R/B_n Repeat to gather sufficient ICC samples to remove measurement variability Figure 97 214 IDLE ICC4R measurement procedure FFh A Cycle Type CMD ADDR ADDR ADDR ADDR ADDR IDLE Minimum cycle time between command, address cycles DQx 80h 00h 00h CMD ADDR ADDR R1 R2 FFh R3 R/B_n A Cycle Type Min time between cmd, addr cycles DQx 85h 00h IDLE DIN DIN Cycle(min) tCCS 00h FFh DIN D0 ... Dn R/B_n Repeat to gather sufficient ICC samples to remove measurement variability Figure 98 215 ICC4W measurement procedure Cycle Type IDLE DQ[7:0] FFh R/B_n Repeat to gather sufficient ICC samples to remove measurement variability Figure 99 216 ICC5 measurement procedure E. MEASURING TIMING PARAMETERS TO/FROM TRI-STATE There are several timing parameters that are measured to or from when: The device is no longer driving the NAND bus or a tri-state (hi-Z) condition The device begins driving from a tri-state (hi-Z) condition These timing parameters include: tDQSD, tDQSHZ, tCHZ, tRHZ, and tIR. See section 4.2. This appendix defines a two point method for measuring timing parameters that involve a tri-state condition. Figure 100 defines a method to calculate the point when the device is no longer driving the NAND bus or begins driving by measuring the signal at two different voltages. The voltage measurement points are acceptable across a wide range (x = 20 mV up to x < 1/4 of VccQ). The figure uses tDQSHZ and tDQSD as examples. However, the method should be used for any timing parameter (asynchronous or source synchronous) that specifies that the device output is no longer driving the NAND bus or specifies that the device begins driving the NAND bus from a tri-state condition. Figure 100 Two point method for measuring timing parameters with tri-state condition 217 F. EZ NAND: END TO END DATA PATH PROTECTION (INFORMATIVE) An EZ NAND implementation may choose to support end to end data protection. In certain topologies with longer trace lengths or discontinuities, it may be helpful to include this capability. However, it is not expected that most EZ NAND implementations need this type of protection at speeds up to 200 MT/s. One potential implementation is to include a two to four byte CRC for each data unit in the page. The data unit is called the CRC sub-page, which is the individual unit that is protected. The capability of end to end data protection should be reported in a vendor specific parameter page bit. The feature may then be configured utilizing a vendor specific Set Features. If there is an error detected as part of end to end data path protection, the error should be reported in the FAIL bit as part of the associated Read or Program operation. Included below is a recommended method for configuring the CRC capabilities as part Set Features. The CRC polynomial to use is not specified. Sub Feature Parameter 7 P1 6 5 4 Reserved (0) P2 P3 P4 3 2 CRC Mode 1 0 CRC Poly Length # of CRC sub-pages Offset of CRC bytes (bits 31:00) Offset of CRC bytes (bits 63:32) CRC Poly Length CRC polynomial length in bytes. Valid values are 2, 3 or 4. CRC Mode 00b = no CRC check or generation 01b = CRC check on Program/store to Flash; CRC check on read from Flash 10b = CRC check and discard on Program/store to Flash; regenerate CRC on read from Flash 11b = Reserved # of CRC sub-pages Specifies the number of CRC sub-pages within a page. The CRC sub-page size is the page size divided by the number of CRC sub-pages. Offset of CRC bytes Specifies the offset of the CRC bytes within each CRC sub-page. E.g., a value of 0h indicates that the CRC is stored at the beginning of the CRC sub-page. 218