FINAL Am28F512A 512 Kilobit (64 K x 8-Bit) CMOS 12.0 Volt, Bulk Erase Flash Memory with Embedded Algorithms DISTINCTIVE CHARACTERISTICS ■ High performance — 70 ns maximum access time ■ CMOS low power consumption — 30 mA maximum active current ■ Embedded Erase Electrical Bulk Chip-Erase — Two seconds typical chip-erase including pre-programming ■ Embedded Program — 100 µA maximum standby current — 4 µs typical byte-program including time-out — No data retention power consumption — One second typical chip program ■ Compatible with JEDEC-standard byte-wide 32-Pin EPROM pinouts — 32-pin PDIP — 32-pin PLCC — 32-pin TSOP ■ 100,000 write/erase cycles minimum ■ Write and erase voltage 12.0 V −5% ■ Latch-up protected to 100 mA from -1 V to VCC +1 V ■ Command register architecture for microprocessor/microcontroller compatible write interface ■ On-chip address and data latches ■ Advanced CMOS flash memory technology — Low cost single transistor memory cell ■ Embedded algorithms for completely self-timed write/erase operations GENERAL DESCRIPTION The Am28F512A is a 512 Kbit Flash memory organized as 64 Kbytes of 8 bits each. AMD’s Flash memories offer the most cost-effective and reliable read/write non- volatile random access memory. The Am28F512A is packaged in 32-pin PDIP, PLCC, and TSOP versions. It is designed to be reprogrammed and erased in-system or in standard EPROM programmers. The Am28F512A is erased when shipped from the factory. The standard Am28F512A offers access times as fast as 70 ns, allowing operation of high-speed microprocessors without wait states. To eliminate bus contention, the Am28F512A has separate chip enable (CE#) and output enable (OE#) controls. AMD’s Flash memories augment EPROM functionality with in-circuit electrical erasure and programming. The Am28F512A uses a command register to manage this functionality, while maintaining a JEDEC Flash standard 32-pin pinout. The command register allows for 100% TTL level control inputs and fixed power supply levels during erase and programming. AMD’s Flash technology reliably stores memory contents even after 100,000 erase and program cycles. The AMD cell is designed to optimize the erase and Publication# 18880 Rev: C Amendment/+2 Issue Date: April 1998 programming mechanisms. In addition, the combination of advanced tunnel oxide processing and low internal electr ic fields for erase a nd programmi ng operations produces reliable cycling. The Am28F512A uses a 12.0V± 5% VPP high voltage input to perform the erase and programming functions. The highest degree of latch-up protection is achieved with AMD’s proprietary non-epi process. Latch-up protection is provided for stresses up to 100 milliamps on address and data pins from –1 V to VCC +1 V. Embedded Program The Am28F512A is byte programmable using the Embedded Programming algorithm. The Embedded Programming algorithm does not require the system to time-out or verify the data programmed. The typical room temperature programming time of the Am28F512A is one second. Embedded Erase The entire chip is bulk erased using the Embedded Erase algorithm. The Embedded Erase algorithm automatically programs the entire array prior to electrical erase. The timing and verification of electrical erase are controlled internal to the device. Typical erasure at room temperature is accomplished in two seconds, including programming. AMD’s Am28F512A is entirely pin and software compatible with AMD Am28F020A, Am28F010A, and Am28F256A Flash memories. Comparing Embedded Algorithms with Flasherase and Flashrite Algorithms Embedded Programming Algorithm vs. Flashrite Programming Algorithm Embedded Erase Algorithm vs. Flasherase Erase Algorithm Am28F512A with Embedded Algorithms Am28F512 using AMD Flashrite and Flasherase Algorithms AMD’s Embedded Programming algorithm requires the user to only write a program set-up command and a program command (program data and address). The device automatically times the programming pulse width, verifies the programming, and counts the number of sequences. A status bit, Data# Polling, provides the user with the programming operation status. The Flashrite Programming algorithm requires the user to write a program set-up command, a program command, (program data and address), and a program verify command, followed by a read and compare operation. The user is required to time the programming pulse width in order to issue the program verify command. An integrated stop timer prevents any possibility of overprogramming. AMD’s Embedded Erase algorithm requires the user to only write an erase setup command and erase command. The device automatically pre-programs and verifies the entire array. The device then automatically times the erase pulse width, verifies the erase operation, and counts the number of sequences. A status bit, Data# Polling, provides the user with the erase operation status. The Flasherase Erase algorithm requires the device to be completely programmed prior to executing an erase command. Commands are written to the command register using standard microprocessor write timings. Register contents serve as inputs to an internal state-machine which controls the erase and programming circuitry. During write cycles, the command register internally latches address and data needed for the programming and erase operations. For system design simplification, the Am28F512A is designed to support either WE# or CE controlled writes. During a system write cycle, addresses are latched on the falling edge of WE# or CE# whichever occurs last. Data is latched on the rising edge of WE# or CE# whichever occurs first. 2 Upon completion of this sequence, the data is read back from the device and compared by the user with the data intended to be written; if there is not a match, the sequence is repeated until there is a match or the sequence has been repeated 25 times. To invoke the erase operation, the user writes an erase set-up command, an erase command, and an erase verify command. The user is required to time the erase pulse width in order to issue the erase verify command. An integrated stop timer prevents any possibility of overerasure. Upon completion of this sequence, the data is read back from the device and compared by the user with erased data. If there is not a match, the sequence is repeated until there is a match or the sequence has been repeated 1,000 times. To simplify the following discussion, the WE# pin is used as the write cycle control pin throughout the rest of this text. All setup and hold times are with respect to the WE# signal. AMD’s Flash technology combines years of EPROM and EEPROM experience to produce the highest levels of quality, reliability, and cost effectiveness. The Am28F512A electrically erases all bits simultaneously using Fowler-Nordheim tunneling. The bytes are programmed one byte at a time using the EPROM programming mechanism of hot electron injection. Am28F512A BLOCK DIAGRAM DQ0–DQ7 VCC VSS Input/Output Buffers Erase Voltage Switch VPP To Array WE# State Control Command Register Program Voltage Switch Data Latch Chip Enable Output Enable Logic CE# OE# Embedded Algorithms Program/Erase Pulse Timer Address Latch Low VCC Detector A0–A15 Y-Decoder Y-Gating X-Decoder 524,288 Bit Cell Matrix 18880C-1 PRODUCT SELECTOR GUIDE Family Part Number Am28F512A Speed Options (VCC = 5.0 V ±10%) -70 -90 -120 -150 -200 Max Access Time (ns) 70 90 120 150 200 CE# (E#) Access (ns) 70 90 120 150 200 OE# (G#) Access (ns) 35 35 50 55 55 Am28F512A 3 CONNECTION DIAGRAMS PDIP WE# (W#) A15 3 30 NC A12 4 29 A14 4 3 2 1 32 31 30 5 29 A14 6 28 A13 A5 A4 7 8 27 26 A9 A6 6 27 A8 A5 7 26 A9 A3 9 25 A11 A4 8 25 A11 A2 10 24 OE# (G#) A3 9 24 OE# (G#) A1 11 23 A10 A2 10 23 A10 A0 12 22 A1 11 22 CE# (E#) DQ0 13 21 CE (E#) DQ7 A0 12 21 DQ7 14 15 16 17 18 19 20 DQ0 13 20 DQ6 DQ1 14 19 DQ5 DQ2 15 18 DQ4 VSS 16 17 DQ3 A8 DQ5 DQ6 DQ4 A13 DQ3 28 VSS 5 DQ1 DQ2 A7 A6 A7 18880C-3 18880C-2 Note: Pin 1 is marked for orientation. 4 WE# (W#) NC VCC 31 VCC 32 2 VPP 1 NC NC VPP A12 A15 PLCC Am28F512A CONNECTION DIAGRAMS (Continued) A11 A9 A8 A13 A14 NC WE# VCC VPP NC A15 A12 A7 A6 A5 A4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 OE# A10 CE# D7 D6 D5 D4 D3 VSS D2 D1 D0 A0 A1 A2 A3 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 A11 A9 A8 A13 A14 NC WE# VCC VPP NC A15 A12 A7 A6 A5 A4 32-Pin TSOP—Standard Pinout OE# A10 CE# D7 D6 D5 D4 D3 VSS D2 D1 D0 A0 A1 A2 A3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 32-Pin TSOP—Reverse Pinout 18880C-4 LOGIC SYMBOL 16 8 A0–A15 DQ0–DQ7 CE# (E#) OE# (G#) WE# (W#) 18880C-5 Am28F512A 5 ORDERING INFORMATION Standard Products AMD standard products are available in several packages and operating ranges. The order number (Valid Combination) is formed by a combination of: AM28F512A -70 J C B OPTIONAL PROCESSING Blank = Standard Processing B = Burn-In Contact an AMD representative for more information. TEMPERATURE RANGE C = Commercial (0°C to +70°C) I = Industrial (–40°C to +85°C) E = Extended (–55°C to +125°C) PACKAGE TYPE P = 32-Pin Plastic DIP (PD 032) J = 32-Pin Rectangular Plastic Leaded Chip Carrier (PL 032) E = 32-Pin Thin Small Outline Package (TSOP) Standard Pinout (TS 032) F = 32-Pin Thin Small Outline Package (TSOP) Reverse Pinout (TSR032) SPEED OPTION See Product Selector Guide and Valid Combinations DEVICE NUMBER/DESCRIPTION Am28F512A 512 Kilobit (64 K x 8-Bit) 12.0 Volt CMOS Flash Memory with Embedded Algorithms Valid Combinations AM28F512A-70 AM28F512A-90 AM28F512A-120 AM28F512A-150 PC, PI, PE, JC, JI, JE, EC, EI, EE, FC, FI, FE Valid Combinations Valid Combinations list configurations planned to be supported in volume for this device. Consult the local AMD sales office to confirm availability of specific valid combinations and to check on newly released combinations. AM28F512A-200 6 Am28F512A PIN DESCRIPTION A0–A15 Address Inputs for memory locations. Internal latches hold addresses during write cycles. VCC Power supply for device operation. (5.0 V ± 5% or 10%) VPP Chip Enable active low input activates the chip’s control logic and input buffers. Chip Enable high will deselect the device and operates the chip in stand-by mode. Program voltage input. VPP must be at high voltage in order to write to the command register. The command register controls all functions required to alter the memory array contents. Memory contents cannot be altered when VPP ≤ VCC +2 V. DQ0–DQ7 VSS CE# (E#) Data Inputs during memory write cycles. Internal latches hold data during write cycles. Data Outputs during memory read cycles. NC No Connect-corresponding pin is not connected internally to the die. OE# (G#) Ground. WE# (W#) Write Enable active low input controls the write function of the command register to the memory array. The target address is latched on the falling edge of the Write Enable pulse and the appropriate data is latched on the rising edge of the pulse. Write Enable high inhibits writing to the device. Output Enable active low input gates the outputs of the device through the data buffers during memory read cycles. Output Enable is high during command sequencing and program/erase operations. Am28F512A 7 BASIC PRINCIPLES This section contains descriptions about the device read, erase, and program operations, and write operation status of the Am29FxxxA, 12.0 volt family of Flash devices. References to some tables or figures may be given in generic form, such as “Command Definitions table”, rather than “Table 1”. Refer to the corresponding data sheet for the actual table or figure. The Am28FxxxA family uses 100% TTL-level control inputs to manage the command register. Erase and reprogramming operations use a fixed 12.0 V ± 5% high voltage input. Read Only Memory Without high VPP voltage, the device functions as a read only memory and operates like a standard EPROM. The control inputs still manage traditional read, standby, output disable, and Auto select modes. Command Register The command register is enabled only when high voltage is applied to the VPP pin. The erase and reprogramming operations are only accessed via the register. In addition, two-cycle commands are required for erase and reprogramming operations. The traditional read, standby, output disable, and Auto select modes are available via the register. The device’s command register is written using standard microprocessor write timings. The register controls an internal state machine that manages all device operations. For system design simplification, the device is designed to support either WE# or CE# controlled writes. During a system write cycle, addresses are latched on the falling edge of WE# or CE# whichever occurs last. Data is latched on the rising edge of WE# or CE# whichever occur first. To simplify the following discussion, the WE# pin is used as the write cycle control pin throughout the rest of this text. All setup and hold times are with respect to the WE# signal. AMD now makes programming extremely simple and reliable. The Embedded Programming algorithm requires the user to only write a program setup command and a program command. The device automatically times the programming pulse width, provides the program verify and counts the number of sequences. A status bit, Data# Polling, provides feedback to the user as to the status of the programming operation. DATA PROTECTION The device is designed to offer protection against accidental erasure or programming caused by spurious system level signals that may exist during power transitions. The device powers up in its read only state. Also, with its control register architecture, alteration of the memory contents only occurs after successful completion of specific command sequences. The device also incorporates several features to prevent inadvertent write cycles resulting from V CC power-up and power-down transitions or system noise. Low VCC Write Inhibit To avoid initiation of a write cycle during VCC power-up and power-down, the device locks out write cycles for VCC < VLKO (see DC characteristics section for voltages). When VCC < VLKO, the command register is disabled, all internal program/erase circuits are disabled, and the device resets to the read mode. The device ignores all writes until VCC > VLKO. The user must ensure that the control pins are in the correct logic state when VCC > VLKO to prevent unintentional writes. Write Pulse “Glitch” Protection Noise pulses of less than 10 ns (typical) on OE#, CE# or WE# will not initiate a write cycle. Logical Inhibit Writing is inhibited by holding any one of OE# = VIL, CE#=VIH or WE# = VIH. To initiate a write cycle CE# and WE# must be a logical zero while OE# is a logical one. OVERVIEW OF ERASE/PROGRAM OPERATIONS Embedded Erase Algorithm AMD now makes erasure extremely simple and reliable. The Embedded Erase algorithm requires the user to only write an erase setup command and erase command. The device will automatically pre-program and verify the entire array. The device automatically times the erase pulse width, provides the erase verify and counts the number of sequences. A status bit, Data# Polling, provides feedback to the user as to the status of the erase operation. 8 Embedded Programming Algorithm Power-Up Write Inhibit Power-up of the device with WE# = CE# = VIL and OE# = V IH will not accept commands on the rising edge of WE#. The internal state machine is automatically reset to the read mode on power-up. Am28F512A FUNCTIONAL DESCRIPTION Description Of User Modes Table 1. Am28F512A Device Bus Operations (Notes 7 and 8) CE# (E#) OE# (G#) WE# (W#) VPP (Note 1) A0 A9 I/O Read VIL VIL X VPPL A0 A9 DOUT Standby VIH X X VPPL X X HIGH Z Output Disable VIL VIH VIH VPPL X X HIGH Z Auto-select Manufacturer Code (Note 2) VIL VIL VIH VPPL VIL VID (Note 3) CODE (01h) Auto-select Device Code (Note 2) VIL VIL VIH VPPL VIH VID (Note 3) CODE (AEh) Read VIL VIL VIH VPPH A0 A9 DOUT (Note 4) Standby (Note 5) VIH X X VPPH X X HIGH Z Output Disable VIL VIH VIH VPPH X X HIGH Z Write VIL VIH VIL VPPH A0 A9 DIN (Note 6) Operation Read-Only Read/Write Legend: X = Don’t care, where Don’t Care is either VIL or VIH levels. VPPL = VPP < VCC + 2 V. See DC Characteristics for voltage levels of VPPH. 0 V < An < VCC + 2 V, (normal TTL or CMOS input levels, where n = 0 or 9). Notes: 1. VPPL may be grounded, connected with a resistor to ground, or < VCC + 2.0 V. VPPH is the programming voltage specified for the device. Refer to the DC characteristics. When VPP = VPPL, memory contents can be read but not written or erased. 2. Manufacturer and device codes may also be accessed via a command register write sequence. Refer to Table 2. 3. 11.5 < VID < 13.0 V. Minimum VID rise time and fall time (between 0 and VID voltages) is 500 ns. 4. Read operation with VPP = VPPH may access array data or the Auto select codes. 5. With VPP at high voltage, the standby current is ICC + IPP (standby). 6. Refer to Table 3 for valid DIN during a write operation. 7. All inputs are Don’t Care unless otherwise stated, where Don’t Care is either VIL or VIH levels. In the Auto select mode all addresses except A9 and A0 must be held at VIL. 8. If VCC ≤ 1.0 Volt, the voltage difference between VPP and VCC should not exceed 10.0 volts. Also, the Am28F256 has a VPP rise time and fall time specification of 500 ns minimum. Am28F512A 9 READ-ONLY MODE When VPP is less than VCC + 2 V, the command register is inactive. The device can either read array or autoselect data, or be standby mode. Read The device functions as a read only memory when VPP < VCC + 2 V. The device has two control functions. Both must be satisfied in order to output data. CE# controls power to the device. This pin should be used for specific device selection. OE# controls the device outputs and should be used to gate data to the output pins if a device is selected. Address access time tACC is equal to the delay from stable addresses to valid output data. The chip enable access time tCE is the delay from stable addresses and stable CE# to valid data at the output pins. The output enable access time is the delay from the falling edge of OE# to valid data at the output pins (assuming the addresses have been stable at least tACC - tOE). Output Disable Output from the device is disabled when OE# is at a logic high level. When disabled, output pins are in a high impedance state. Auto Select Flash memories can be programmed in-system or in a standard PROM programmer. The device may be soldered to the circuit board upon receipt of shipment and programmed in-system. Alternatively, the device may initially be programmed in a PROM programmer prior to soldering the device to the board. The Auto select mode allows the reading out of a binary code from the device that will identify its manufacturer and type. This mode is intended for the purpose of automatically matching the device to be programmed with its corresponding programming algorithm. This mode is functional over the entire temperature range of the device. Programming In A PROM Programmer Standby Mode The device has two standby modes. The CMOS standby mode (CE# input held at VCC ± 0.5 V), consumes less than 100 µA of current. TTL standby mode (CE# is held at VIH) reduces the current requirements to less than 1 mA. When in the standby mode the outputs are in a high impedance state, independent of the OE# input. If the device is deselected during erasure, programming, or program/erase verification, the device will draw active current until the operation is terminated. Table 2. To activate this mode, the programming equipment must force VID (11.5 V to 13.0 V) on address A9. Two identifier bytes may then be sequenced from the device outputs by toggling address A0 from VIL to VIH. All other address lines must be held at VIL, and V PP must be less than or equal to VCC + 2.0 V while using this Auto select mode. Byte 0 (A0 = VIL) represents the manufacturer code and byte 1 (A0 = VIH) the device identifier code. For the device the two bytes are given in the table 2 of the device data sheet. All identifiers for manufacturer and device codes will exhibit odd parity with the MSB (DQ7) defined as the parity bit. Am28F512A Auto Select Code Type A0 Code (HEX) Manufacturer Code VIL 01 Device Code VIH AE 10 Am28F512A ERASE, PROGRAM, AND READ MODE When VPP is equal to 12.0 V ± 5%, the command register is active. All functions are available. That is, the device can program, erase, read array or autoselect data, or be standby mode. Write Operations High voltage must be applied to the VPP pin in order to activate the command register. Data written to the register serves as input to the internal state machine. The output of the state machine determines the operational function of the device. The command register does not occupy an addressable memory location. The register is a latch that stores the command, along with the address and data information needed to execute the command. The register is written by bringing WE# and CE# to VIL, while OE# is at VIH. Addresses are latched on the falling edge of WE#, while data is latched on the rising edge of the WE# pulse. Standard microprocessor write timings are used. The device requires the OE# pin to be VIH for write operations. This condition eliminates the possibility for bus contention during programming operations. In order to write, OE# must be VIH, and CE# and WE# must be VIL. If any pin is not in the correct state a write command will not be executed. Table 3. Refer to AC Write Characteristics and the Erase/Programming Waveforms for specific timing parameters. Command Definitions The contents of the command register default to 00h (Read Mode) in the absence of high voltage applied to the V PP pin. The device operates as a read only memory. High voltage on the V PP pin enables the command register. Device operations are selected by writing specific data codes into the command register. Table 3 in the device data sheet defines these register commands. Read Command Memory contents can be accessed via the read command when VPP is high. To read from the device, write 00h into the command register. Standard microprocessor read cycles access data from the memory. The device will remain in the read mode until the command register contents are altered. The command register defaults to 00h (read mode) upon VPP power-up. The 00h (Read Mode) register default helps ensure that inadvertent alteration of the memory contents does not occur during the VPP power transition. Refer to the AC Read Characteristics and Waveforms for the specific timing parameters. Am28F512A Command Definitions First Bus Cycle Second Bus Cycle Operation (Note 1) Address (Note 2) Data (Note 3) Operation (Note 1) Address (Note 2) Data (Note 3) Read Memory (Note 4) Write X 00h/FFh Read RA RD Read Auto select Write X 80h or 90h Read 00h/01h 01h/AEh Embedded Erase Set-up/ Embedded Erase Write X 30h Write X 30h Embedded Program Set-up/ Embedded Program Write X 10h or 50h Write PA PD Reset (Note 4) Write X 00h/FFh Write X 00h/FFh Command Notes: 1. Bus operations are defined in Table 1. 2. RA = Address of the memory location to be read. PA = Address of the memory location to be programmed. Addresses are latched on the falling edge of the WE# pulse. X = Don’t care. 3. RD = Data read from location RA during read operation. PD = Data to be programmed at location PA. Data latched on the rising edge of WE#. 4. Please reference Reset Command section. Am28F512A 11 FLASH MEMORY PROGRAM/ERASE OPERATIONS Embedded Erase Algorithm The automatic chip erase does not require the device to be entirely pre-programmed prior to executing the Embedded set-up erase command and Embedded erase command. Upon executing the Embedded erase command the device automatically will program and verify the entire memory for an all zero data pattern. The system is not required to provide any controls or timing during these operations. When the device is automatically verified to contain an all zero pattern, a self-timed chip erase and verify begin. The erase and verify operation are complete when the data on DQ7 is “1" (see Write Operation Status section) atwhich time the device returns to Read mode. The system is not required to provide any control or timing during these operations. When using the Embedded Erase algorithm, the erase automatically terminates when adequate erase margin has been achieved for the memory array (no erase verify command is required). The margin voltages are internally generated in the same manner as when the standard erase verify command is used. The Embedded Erase Set-Up command is a command only operation that stages the device for automatic electrical erasure of all bytes in the array. Embedded Erase Setup is performed by writing 30h to the command register. To commence automatic chip erase, the command 30h must be written again to the command register. The automatic erase begins on the rising edge of the WE and terminates when the data on DQ7 is “1" (see Write Operation Status section) at which time the device returns to Read mode. Figure 1 and Table 4 illustrate the Embedded Erase algorithm, a typical command string and bus operation. START Apply VPPH Write Embedded Erase Setup Command Write Embedded Erase Command Data# Poll from Device Erasure Completed 18880C-6 Bus Operations Figure 1. Embedded Erase Algorithm Table 4. Embedded Erase Algorithm Command Comments Wait for VPP Ramp to VPPH (see Note) Standby Embedded Erase Setup Command Data = 30h Embedded Erase Command Data = 30h Write Read Data# Polling to Verify Erasure Standby Compare Output to FFh Read Available for Read Operations Note: See AC and DC Characteristics for values of VPP parameters. The VPP power supply can be hard-wired to the device or switchable. When VPP is switched, VPPL may be ground, no connect with a resistor tied to ground, or less than VCC + 2.0 V. Refer to Functional Description. 12 Am28F512A Embedded Programming Algorithm of WE# also begins the programming operation. The system is not required to provide further controls or timings. The device will automatically provide an adequate internally generated program pulse and verify margin. The automatic programming operation is completed when the data on DQ7 is equivalent to data written to this bit (see Write Operation Status section) at which time the device returns to Read mode. The Embedded Program Setup is a command only operation that stages the device for automatic programming. Embedded Program Setup is performed by writing 10h or 50h to the command register. Once the Embedded Setup Program operation is performed, the next WE# pulse causes a transition to an active programming operation. Addresses are latched on the falling edge of CE# or WE# pulse, whichever happens later. Data is latched on the rising edge of WE# or CE#, whichever happens first. The rising edge Figure 2 and Table 5 illustrate the Embedded Program algorithm, a typical command string, and bus operation. START Apply VPPH Write Embedded Setup Program Command Write Embedded Program Command (A/D) Data# Poll Device No Increment Address Last Address Yes Programming Completed 18880C-7 Bus Operations Figure 2. Embedded Programming Algorithm Table 5. Embedded Programming Algorithm Command Comments Wait for VPP Ramp to VPPH (see Note) Standby Write Embedded Program Setup Command Data = 10h or 50h Write Embedded Program Command Valid Address/Data Read Data# Polling to Verify Completion Read Available for Read Operations Note: See AC and DC Characteristics for values of VPP parameters. The VPP power supply can be hard-wired to the device or switchable. When VPP is switched, VPPL may be ground, no connect with a resistor tied to ground, or less than VCC + 2.0 V. Refer to Functional Description. Device is either powered-down, erase inhibit or program inhibit. Am28F512A 13 Write Operation Status Data Polling—DQ7 The device features Data# Polling as a method to indicate to the host system that the Embedded algorithms are either in progress or completed. While the Embedded Programming algorithm is in operation, an attempt to read the device at a valid address will produce the complement of expected Valid data on DQ7. Upon completion of the Embedded Program algorithm an attempt to read the device at a valid address will produce Valid data on DQ7. The Data# Polling feature is valid after the rising edge of the second WE# pulse of the two write pulse sequence. While the Embedded Erase algorithm is in operation, DQ7 will read “0" until the erase operation is completed. Upon completion of the erase operation, the data on DQ7 will read “1.” The Data# Polling feature is valid after the rising edge of the second WE# pulse of the two Write pulse sequence. The Data# Polling feature is only active during Embedded Programming or erase algorithms. See Figures 3 and 4 for the Data# Polling timing specifications and diagrams. Data# Polling is the standard method to check the write operation status, however, an alternative method is available using Toggle Bit. START Read Byte (DQ0–DQ7) Addr = VA DQ7 = Data ? VA = Byte address for programming = XXXXh during chip erase Yes No No DQ5 = 1 ? Yes Read Byte (DQ0–DQ7) Addr = VA DQ7 = Data ? Yes No Fail Pass 18880C-8 Note: DQ7 is rechecked even if DQ5 = “1” because DQ7 may change simultaneously with DQ5 or after DQ5. Figure 3. 14 Data# Polling Algorithm Am28F512A tCH CE# tDF tOE OE# tOEH WE# tCE tOH * DQ7 DQ7# DQ7 = Valid Data High Z tWHWH 3 or 4 DQ0–DQ6 DQ0–DQ6 = Invalid DQ0–DQ7 Valid Data 18880C-9 *DQ7 = Valid Data (The device has completed the Embedded operation.) Figure 4. AC Waveforms for Data# Polling during Embedded Algorithm Operations Am28F512A 15 Toggle Bit—DQ6 The device also features a “Toggle Bit” as a method to indicate to the host system that the Embedded algorithms are either in progress or completed. Successive attempts to read data from the device at a valid address, while the Embedded Program algorithm is in progress, or at any address while the Embedded Erase algorithm is in progress, will result in DQ6 toggling between one and zero. Once the Embedded Program or Erase algorithm is completed, DQ6 will stop toggling to indicate the completion of either Embedded operation. Only on the next read cycle will valid data be obtained. The toggle bit is valid after the rising edge of the first WE# pulse of the two write pulse sequence, unlike Data# Polling which is valid after the rising edge of the second WE# pulse. This feature allows the user to determine if the device is partially through the two write pulse sequence. See Figures 5 and 6 for the Toggle Bit timing specifications and diagrams. START VA = Byte address for programming = XXXXh during chip erase Read Byte (DQ0–DQ7) Addr = VA DQ6 = Toggle ? No Yes No DQ5 = 1 ? Yes Read Byte (DQ0–DQ7) Addr = VA DQ6 = Toggle ? No Yes Pass Fail 18880C-10 Note: DQ6 is rechecked even if DQ5 = “1” because DQ6 may stop toggling at the same time as DQ5 changing to “1”. Figure 5. 16 Toggle Bit Algorithm Am28F512A CE# tOEH WE# OE# * Data DQ0–DQ7 DQ6 = DQ6 Stop Toggling DQ6 = DQ0–DQ7 Valid tOE 18880C-11 Note: *DQ6 stops toggling (The device has completed the Embedded operation.) Figure 6. AC Waveforms for Toggle Bit during Embedded Algorithm Operations DQ5 Power-Up/Power-Down Sequence Exceeded Timing Limits The device powers-up in the Read only mode. Power supply sequencing is not required. Note that if VCC ≤ 1.0 Volt, the voltage difference between VPP and VCC should not exceed 10.0 Volts. Also, the device has a rise VPP rise time and fall time specification of 500 ns minimum. DQ5 will indicate if the program or erase time has exceeded the specified limits. This is a failure condition and the device may not be used again (internal pulse count exceeded). Under these conditions DQ5 will produce a “1.” The program or erase cycle was not successfully completed. Data# Polling is the only operating function of the device under this condition. The CE# circuit will partially power down the device under these conditions (to approximately 2 mA). The OE# and WE# pins will control the output disable functions as described in the Command Definitions table in the corresponding device data sheet. Parallel Device Erasure The Embedded Erase algorithm greatly simplifies parallel device erasure. Since the erase process is internal to the device, a single erase command can be given to multiple devices concurrently. By implementing a parallel erase algorithm, total erase time may be minimized. Note that the Flash memories may erase at different rates. If this is the case, when a device is completely erased, use a masking code to prevent further erasure (over-erasure). The other devices will continue to erase until verified. The masking code applied could be the read command (00h). Reset Command The Reset command initializes the Flash memory device to the Read mode. In addition, it also provides the user with a safe method to abort any device operation (including program or erase). The Reset must be written two consecutive times after the Setup Program command (10h or 50h). This will reset the device to the Read mode. Following any other Flash command, write the Reset command once to the device. This will safely abort any previous operation and initialize the device to the Read mode. The Setup Program command (10h or 50h) is the only command that requires a two-sequence reset cycle. The first Reset command is interpreted as program data. However, FFh data is considered as null data during programming operations (memory cells are only programmed from a logical “1" to “0"). The second Reset command safely aborts the programming operation and resets the device to the Read mode. Memory contents are not altered in any case. Am28F512A 17 This detailed information is for your reference. It may prove easier to always issue the Reset command two consecutive times. This eliminates the need to determine if you are in the Setup Program state or not. In-System Programming Considerations Flash memories can be programmed in-system or in a standard PROM programmer. The device may be soldered to the circuit board upon receipt of shipment and programmed in-system. Alternatively, the device may initially be programmed in a PROM programmer prior to soldering the device to the circuit board. Auto Select Command AMD’s Flash memories are designed for use in applications where the local CPU alters memory contents. In order to correctly program any Flash memories 18 in-system, manufacturer and device codes must be accessible while the device resides in the target system. PROM programmers typically access the signature codes by raising A9 to a high voltage. However, multiplexing high voltage onto address lines is not a generally desired system design practice. The device contains an Auto Select operation to supplement traditional PROM programming methodologies. The operation is initiated by writing 80h or 90h into the command register. Following this command, a read cycle address 0000h retrieves the manufacturer code of 01h (AMD). A read cycle from address 0001h returns the device code (see the Auto Select Code table of the corresponding device data sheet). To terminate the operation, it is necessary to write another valid command, such as Reset (00h or FFh), into the register. Am28F512A ABSOLUTE MAXIMUM RATINGS OPERATING RANGES Storage Temperature . . . . . . . . . . . . –65°C to +150°C Plastic Packages . . . . . . . . . . . . . . . –65°C to +125°C Commercial (C) Devices Ambient Temperature with Power Applied. . . . . . . . . . . . . .–55°C to + 125°C Industrial (I) Devices Voltage with Respect To Ground All pins except A9 and VPP (Note 1) . . . . . . . . . . . . . . . . . . . . . . .–2.0 V to +7.0 V Extended (E) Devices Ambient Temperature (TA) . . . . . . . . . . .0°C to +70°C Ambient Temperature (TA) . . . . . . . . .–40°C to +85°C Ambient Temperature (TA) . . . . . . . .–55°C to +125°C VCC (Note 1). . . . . . . . . . . . . . . . . . . .–2.0 V to +7.0 V VCC Supply Voltages A9 (Note 2). . . . . . . . . . . . . . . . . . . .–2.0 V to +14.0 V VCC . . . . . . . . . . . . . . . . . . . . . . . +4.50 V to +5.50 V VPP (Note 2). . . . . . . . . . . . . . . . . . .–2.0 V to +14.0 V VPP Voltages Output Short Circuit Current (Note 3) . . . . . . 200 mA Read . . . . . . . . . . . . . . . . . . . . . . . . –0.5 V to +12.6 V Notes: 1. Minimum DC voltage on input or I/O pins is –0.5 V. During voltage transitions, inputs may overshoot VSS to –2.0 V for periods of up to 20 ns. Maximum DC voltage on input and I/O pins is V CC + 0.5 V. During voltage transitions, input and I/O pins may overshoot to V CC + 2.0 V for periods up to 20ns. Program, Erase, and Verify . . . . . . +11.4 V to +12.6 V Operating ranges define those limits between which the functionality of the device is guaranteed. 2. Minimum DC input voltage on A9 and VPP pins is –0.5 V. During voltage transitions, A9 and VPP may overshoot VSS to –2.0 V for periods of up to 20 ns. Maximum DC input voltage on A9 and V PP is +13.0 V which may overshoot to 14.0 V for periods up to 20 ns. 3. No more than one output shorted to ground at a time. Duration of the short circuit should not be greater than one second. Stresses above those listed under “Absolute Maximum Ratings” may cause permanent damage to the device. This is a stress rating only; functional operation of the device at these or any other conditions above those indicated in the operational sections of this specification is not implied. Exposure of the device to absolute maximum rating conditions for extended periods may affect device reliability. Am28F512A 19 MAXIMUM OVERSHOOT 20 ns 20 ns +0.8 V –0.5 V –2.0 V 20 ns 18880C-12 Maximum Negative Input Overshoot 20 ns VCC + 2.0 V VCC + 0.5 V 2.0 V 20 ns 20 ns 18880C-13 Maximum Positive Input Overshoot 20 ns 14.0 V 13.5 V VCC + 0.5 V 20 ns 20 ns 18880C-14 Maximum VPP Overshoot 20 Am28F512A DC CHARACTERISTICS over operating range unless otherwise specified TTL/NMOS Compatible Parameter Symbol Parameter Description Test Conditions Min Typ Max Unit ILI Input Leakage Current VCC = VCC Max, VIN = VCC or VSS ±1.0 µA ILO Output Leakage Current VCC = VCC Max, VOUT = VCC or VSS ±1.0 µA ICCS VCC Standby Current VCC = VCC Max, CE# = VIH 0.2 1.0 mA ICC1 VCC Active Read Current VCC = VCC Max, CE# = VIL, OE# = VIH IOUT = 0 mA, at 6 MHz 20 30 mA ICC2 VCC Programming Current CE# = VIL Programming in Progress (Note 4) 20 30 mA ICC3 VCC Erase Current CE# = VIL Erasure in Progress (Note 4) 20 30 mA IPPS VPP Standby Current VPP = VPPL ±1.0 µA IPP1 VPP Read Current IPP2 VPP Programming Current VPP = VPPH Programming in Progress (Note 4) 10 30 mA IPP3 VPP Erase Current VPP = VPPH Erasure in Progress (Note 4) 10 30 mA VIL Input Low Voltage –0.5 0.8 V VIH Input High Voltage 2.0 VCC + 0.5 V VOL Output Low Voltage IOL = 5.8 mA, VCC = VCC Min 0.45 V VOH1 Output High Voltage IOH = –2.5 mA, VCC = VCC Min 2.4 VID A9 Auto Select Voltage A9 = VID 11.5 IID A9 Auto Select Current A9 = VID Max, VCC = VCC Max VPPL VPP during Read-Only Operations Note: Erase/Program are inhibited when VPP = VPPL VPPH VLKO VPP = VPPH 70 200 ±1.0 VPP = VPPL µA V 13.0 V 50 µA 0.0 VCC +2.0 V VPP during Read/Write Operations 11.4 12.6 V Low VCC Lock-out Voltage 3.2 5 3.7 V Notes: 1. Caution: The Am28F512A must not be removed from (or inserted into) a socket when VCC or VPP is applied. If VCC ≤ 1.0 volt, the voltage difference between VPP and VCC should not exceed 10.0 volts. Also, the Am28F512A has a VPP rise time and fall time specification of 500 ns minimum. 2. ICC1 is tested with OE# = VIH to simulate open outputs. 3. Maximum active power usage is the sum of ICC and IPP. 4. Not 100% tested. Am28F512A 21 DC CHARACTERISTICS CMOS Compatible Parameter Symbol Parameter Description Test Conditions Min Typ Max Unit ILI Input Leakage Current VCC = VCC Max, VIN = VCC or VSS ±1.0 µA ILO Output Leakage Current VCC = VCC Max, VOUT = VCC or VSS ±1.0 µA ICCS VCC Standby Current VCC = VCC Max, CE# = VCC + 0.5 V 15 100 µA ICC1 VCC Active Read Current VCC = VCC Max, CE# = VIL, OE# = VIH IOUT = 0 mA, at 6 MHz 20 30 mA ICC2 VCC Programming Current CE# = VIL Programming in Progress (Note 4) 20 30 mA ICC3 VCC Erase Current CE# = VIL Erasure in Progress (Note 4) 20 30 mA IPPS VPP Standby Current VPP = VPPL ±1.0 µA IPP1 VPP Read Current VPP = VPPH 70 200 µA IPP2 VPP Programming Current VPP = VPPH Programming in Progress (Note 4) 10 30 mA IPP3 VPP Erase Current VPP = VPPH Erasure in Progress (Note 4) 10 30 mA VIL Input Low Voltage –0.5 0.8 V VIH Input High Voltage 0.7 VCC VCC + 0.5 V VOL Output Low Voltage 0.45 V VOH1 Output High Voltage VOH2 IOL = 5.8 mA, VCC = VCC Min IOH = –2.5 mA, VCC = VCC Min 0.85 VCC IOH = –100 µA, VCC = VCC Min VCC –0.4 VID A9 Auto Select Voltage A9 = VID IID A9 Auto Select Current A9 = VID Max, VCC = VCC Max VPPL VPPL during Read-Only Operations Note: Erase/Program are inhibited when VPP = VPPL VPPH VLKO V 11.5 13.0 V 50 µA 0.0 VCC + 2.0 V VPP during Read/Write Operations 11.4 12.6 V Low VCC Lock-out Voltage 3.2 5 3.7 V Notes: 1. Caution: The Am28F512A must not be removed from (or inserted into) a socket when VCC or VPP is applied. If VCC ≤ 1.0 volt, the voltage difference between VPP and VCC should not exceed 10.0 volts. Also, the Am28F512A has a VPP rise time and fall time specification of 500 ns minimum. 2. ICC1 is tested with OE# = VIH to simulate open outputs. 3. Maximum active power usage is the sum of ICC and IPP. 4. Not 100% tested. 22 Am28F512A 25 ICC Active in mA 20 15 10 55°C 0°C 25°C 70°C 125°C 5 0 0 1 2 3 4 5 6 Frequency in MHz 7 8 9 10 11 12 18880B-15 Figure 7. Am28F512A—Average ICC Active vs. Frequency VCC = 5.5 V, Addressing Pattern = Minmax Data Pattern = Checkerboard TEST CONDITIONS Table 6. 5.0 V Test Specifications Test Condition 2.7 kΩ Device Under Test Output Load Output Load Capacitance, CL (including jig capacitance) CL 6.2 kΩ Note: Diodes are IN3064 or equivalent Unit 30 100 ≤10 pF ns 0.0–3.0 0.45–2.4 V Input timing measurement reference levels 1.5 0.8, 2.0 V Output timing measurement reference levels 1.5 0.8, 2.0 V 18880C-16 23 All others 1 TTL gate Input Rise and Fall Times Input Pulse Levels Figure 8. -70 Test Setup Am28F512A SWITCHING TEST WAVEFORMS 3V 2.4 V 2.0 V 2.0 V Test Points 1.5 V Test Points 1.5 V 0.8 V 0.8 V 0V 0.45 V Output Input Output Input AC Testing (all speed options except -70): Inputs are driven at 2.4 V for a logic “1” and 0.45 V for a logic “0”. Input pulse rise and fall times are ≤10 ns. AC Testing for -70 devices: Inputs are driven at 3.0 V for a logic “1” and 0 V for a logic “0”. Input pulse rise and fall times are ≤10 ns. 18880C-17 SWITCHING CHARACTERISTICS over operating range unless otherwise specified AC Characteristics—Read Only Operation Parameter Symbols Am28F512A Speed Options JEDEC Standard -70 -90 -120 -150 -200 Unit tAVAV tRC Read Cycle Time (Note 2) Min 70 90 120 150 200 ns tELQV tCE Chip Enable Access Time Max 70 90 120 150 200 ns tAVQV tACC Address Access Time Max 70 90 120 150 200 ns tGLQV tOE Output Enable Access Time Max 35 35 50 55 55 ns tELQX tLZ Chip Enable to Output in Low Z (Note 2) Min 0 0 0 0 0 ns tEHQZ tDF Chip Disable to Output in High Z (Note 1) Max 20 20 30 35 35 ns tGLQX tOLZ Output Enable to Output in Low Z (Note 2) Min 0 0 0 0 0 ns tGHQZ tDF Output Disable to Output in High Z (Note 2) Max 20 20 30 35 35 ns tAXQX tOH Output Hold from first of Address, CE#, or OE# Change (Note 2) Min 0 0 0 0 0 ns VCC Setup Time to Valid Read (Note 2) Min 50 50 50 50 50 µs tVCS Parameter Description Notes: 1. Guaranteed by design not tested. 2. Not 100% tested. Am28F512A 24 AC Characteristics—Write/Erase/Program Operations Parameter Symbols Am28F512A Speed Options JEDEC Standard -70 -90 -120 -150 -200 Unit tAVAV tWC Write Cycle Time (Note 4) Min 70 90 120 150 200 ns tAVWL tAS Address Setup Time Min 0 0 0 0 0 ns tWLAX tAH Address Hold Time Min 45 45 50 60 75 ns tDVWH tDS Data Setup Time Min 45 45 50 50 50 ns tWHDX tDH Data Hold Time Min 10 10 10 10 10 ns Output Enable Hold Time for Embedded Algorithm only Min 10 10 10 10 10 µs Read Recovery Time before Write Min 0 0 0 0 0 µs tOEH tGHWL Parameter Description tELWLE tCSE Chip Enable Embedded Algorithm Setup Time Min 20 20 20 20 20 ns tWHEH tCH Chip Enable Hold Time Min 0 0 0 0 0 ns tWLWH tWP Write Pulse Width Min 45 45 50 60 60 ns tWHWL tWPH Write Pulse Width HIGH Min 20 20 20 20 20 ns tWHWH3 Embedded Programming Operation (Note 2) Min 14 14 14 14 14 µs tWHWH4 Embedded Erase Operation (Note 2) Typ 5 5 5 5 5 ms tVPEL VPP Setup Time to Chip Enable LOW (Note 4) Min 100 100 100 100 100 ns tVCS VCC Setup Time to Chip Enable LOW (Note 4) Min 50 50 50 50 50 µs tVPPR VPP Rise Time 90% VPPH (Note 4) Min 500 500 500 500 500 ns tVPPF VPP Fall Time 10% VPPL (Note 4) Min 500 500 500 500 500 ns tLKO VCC < VLKO to Reset (Note 4) Min 100 100 100 100 100 ns Notes: 1. Read timing characteristics during read/write operations are the same as during read-only operations. Refer to AC Characteristics for Read Only operations. 2. Embedded Program Operation of 14 µs consists of 10 µs program pulse and 4 ms write recovery before read. This is the minimum time for one pass through the programming algorithm. 3. Embedded Erase Operation of 5 sec consists of 4 sec array pre-programming time and one sec array erase time. This is a typical time for one embedded erase operation. 4. Not 100% tested. 25 Am28F512A KEY TO SWITCHING WAVEFORMS WAVEFORM INPUTS OUTPUTS Steady Changing from H to L Changing from L to H Don’t Care, Any Change Permitted Changing, State Unknown Does Not Apply Center Line is High Impedance State (High Z) SWITCHING WAVEFORMS Power-up, Standby Device and Address Selection Outputs Enabled Data Valid Standby, Power-down Addresses Stable Addresses tAVAV (tRC) CE# (E#) tEHQZ (tDF) OE# (G#) tWHGL tGHQZ (tDF) WE# (W#) tGLQV (tOE) tELQV (tCE) tGLQX (tOLZ) tVCS High Z tAXQX (tOH) tELQX (tLZ) Output Valid Data (DQ) High Z tAVQV (tACC) 5.0 V VCC 0V 18880C-18 Figure 9. AC Waveforms for Read Operations Am28F512A 26 SWITCHING WAVEFORMS Embedded Erase Setup Embedded Erase tWC tAS Erase Data# Polling Read Standby Addresses CE# tRC tAH tGHWL OE# tWHWH3 OR 4 tWP WE# tWPH tCSE Data 30h tVCS tDF tDH tOE 30h DQ7# DQ7# tDS tOH VCC tCE VPP tVPEL 18880C-19 Note: DQ7# is the complement of the data written to the device. Figure 10. 27 AC Waveforms for Embedded Erase Operation Am28F512A SWITCHING WAVEFORMS Embedded Program Setup Embedded Program Addresses Data# Polling PA tWC Read PA tRC tAS CE# tAH tGHWL OE# tWHWH3 OR 4 tWP WE# tCSE tDH Data tWPH tDF tOE DIN 50h DQ7# DQ7# DOUT tVCS tOH tDS VCC tCE VPP tVPEL 18880C-20 Notes: 1. DIN is data input to the device. 2. DQ7# is the complement of the data written to the device. 3. DOUT is the data written to the device. Figure 11. AC Waveforms for Embedded Programming Operation Am28F512A 28 AC CHARACTERISTICS—WRITE/ERASE/PROGRAM OPERATIONS Alternate CE# Controlled Writes Parameter Symbols Am28F512A Speed Options JEDEC Standard Parameter Description -70 -90 -120 -150 -200 Unit tAVAV tWC Write Cycle Time (Note 4) Min 70 90 120 150 200 ns tAVEL tAS Address Setup Time Min 0 0 0 0 0 ns tELAX tAH Address Hold Time Min 45 45 50 60 75 ns tDVEH tDS Data Setup Time Min 45 45 50 50 50 ns tEHDX tDH Data Hold Time Min 10 10 10 10 10 ns tOEH Output Enable Hold Time for Embedded Algorithm only Min 10 10 10 10 10 µs tGHEL Read Recovery Time before Write Min 0 0 0 0 0 ns tWLEL tWS WE# Setup Time by CE# Min 0 0 0 0 0 ns tEHWK tWH WE# Hold Time Min 0 0 0 0 0 ns tELEH tCP Write Pulse Width Min 65 65 70 80 80 ns tEHEL tCPH Write Pulse Width HIGH Min 20 20 20 20 20 ns tEHEH3 Embedded Programming Operation (Note 2) Min 14 14 14 14 14 µs tEHEH4 Embedded Erase Operation (Note 3) Typ 5 5 5 5 5 sec tVPEL VPP Setup Time to Chip Enable LOW (Note 4) Min 100 100 100 100 100 ns tVCS VCC Setup Time to Chip Enable LOW (Note 4) Min 50 50 50 50 50 µs tVPPR VPP Rise Time 90% VPPH (Note 4) Min 500 500 500 500 500 ns tVPPF VPP Fall Time 10% VPPL (Note 4) Min 500 500 500 500 500 ns tLKO VCC < VLKO to Reset (Note 4) Min 100 100 100 100 100 ns Notes: 1. Read timing characteristics during read/write operations are the same as during read-only operations. Refer to AC Characteristics for Read Only operations. 2. Embedded Program Operation of 14 µs consists of 10 µs program pulse and 4 ms write recovery before read. This is the minimum time for one pass through the programming algorithm. 3. Embedded Erase Operation of 5 sec consists of 4 sec array pre-programming time and one sec array erase time. This is a typical time for one embedded erase operation. 4. Not 100% tested. 29 Am28F512A SWITCHING WAVEFORMS Embedded Program Setup Embedded Program Addresses PA PA tWC tAS CE# OE# Data# Polling tAH tGHEL tCPH tEHEH3 OR 4 tCP WE# tDH tWS Data 50h DIN DQ7# DQ7# DOUT tDS VCC VPP tVPEL 18880C-21 Notes: 1. DIN is data input to the device. 2. DQ7# is the complement of the data written to the device. 3. DOUT is the data written to the device. Figure 12. AC Waveforms for Embedded Programming Operation Using CE# Controlled Writes Am28F512A 30 ERASE AND PROGRAMMING PERFORMANCE Limits Typ (Note 1) Max (Note 2) Unit Chip Erase Time 1 10 sec Excludes 00h programming prior to erasure Chip Programming Time 1 7 sec Excludes system-level overhead Parameter Min Write/Erase Cycles 100,000 Comments Cycles 14 Byte Programming Time µs 96 (Note 3) ms Notes: 1. 25°C, 12 V VPP. 2. Maximum time specified is lower than worst case. Worst case is derived from the Embedded Algorithm internal counter which allows for a maximum 6000 pulses for both program and erase operations. Typical worst case for program and erase is significantly less than the actual device limit. 3. Typical worst case = 84 µs. DQ5 = “1” only after a byte takes longer than 96 ms to program. LATCHUP CHARACTERISTICS Parameter Min Max Input Voltage with respect to VSS on all pins except I/O pins (Including A9 and VPP) –1.0 V 13.5 V Input Voltage with respect to VSS on all pins I/O pins –1.0 V VCC + 1.0 V –100 mA +100 mA Current Includes all pins except VCC. Test conditions: VCC = 5.0 V, one pin at a time. PIN CAPACITANCE Parameter Symbol CIN Parameter Description Test Conditions Typ Max Unit Input Capacitance VIN = 0 8 10 pF COUT Output Capacitance VOUT = 0 8 12 pF CIN2 VPP Input Capacitance VPP = 0 8 12 pF Note: Sampled, not 100% tested. Test conditions TA = 25°C, f = 1.0 MHz. DATA RETENTION Parameter Test Conditions Min Unit 150°C 10 Years 125°C 20 Years Minimum Pattern Data Retention Time 31 Am28F512A PHYSICAL DIMENSIONS PD032—32-Pin Plastic DIP (measured in inches) 1.640 1.670 .600 .625 17 32 .009 .015 .530 .580 Pin 1 I.D. .630 .700 16 .045 .065 0° 10° .005 MIN .140 .225 16-038-S_AG PD 032 EC75 5-28-97 lv SEATING PLANE .090 .110 .120 .160 .016 .022 .015 .060 PL032—32-Pin Plastic Leaded Chip Carrier (measured in inches) .447 .453 .485 .495 .009 .015 .585 .595 .042 .056 .125 .140 Pin 1 I.D. .080 .095 .547 .553 SEATING PLANE .400 REF. .490 .530 .013 .021 .050 REF. .026 .032 TOP VIEW SIDE VIEW Am28F512A 16-038FPO-5 PL 032 DA79 6-28-94 ae 32 PHYSICAL DIMENSIONS TS032—32-Pin Standard Thin Small Outline Package (measured in millimeters) 0.95 1.05 Pin 1 I.D. 1 7.90 8.10 0.50 BSC 0.05 0.15 18.30 18.50 19.80 20.20 0.08 0.20 0.10 0.21 1.20 MAX 0° 5° 0.50 0.70 33 Am28F512A 16-038-TSOP-2 TS 032 DA95 3-25-97 lv PHYSICAL DIMENSIONS TSR032—32-Pin Reversed Thin Small Outline Package (measured in millimeters) 0.95 1.05 Pin 1 I.D. 1 7.90 8.10 0.50 BSC 0.05 0.15 18.30 18.50 19.80 20.20 0.08 0.20 0.10 0.21 1.20 MAX 0° 5° 16-038-TSOP-2 TSR032 DA95 3-25-97 lv 0.50 0.70 DATA SHEET REVISION SUMMARY FOR AM28F512A Revision C Revision C+2 Deleted -75, -95, and -250 speed options. Matched formatting to other current data sheets. Product Selector Guide Revision C+1 Connection Diagrams Programming In A PROM Programmer On standard TSOP, corrected pin 9 to VPP and pin 10 to NC. On reverse TSOP, corrected pin 23 to NC and pin 24 to VPP. Deleted the paragraph “(Refer to the AUTO SELECT paragraph in the ERASE, PROGRAM, and READ MODE section for programming the Flash memory device in-system).” Corrected maximum access time for -200 to 200 ns. Trademarks Copyright © 1998 Advanced Micro Devices, Inc. All rights reserved. ExpressFlash is a trademark of Advanced Micro Devices, Inc. AMD, the AMD logo, and combinations thereof are registered trademarks of Advanced Micro Devices, Inc. Product names used in this publication are for identification purposes only and may be trademarks of their respective companies. Am28F512A 34