EM MICROELECTRONIC - MARIN SA EM4022 Multi Frequency Contactless Identification Device Anti-Collision compatible with BTG's Supertag Category Protocols Description The EM4022 (previously named P4022) chip implements patented anti-collision protocols for both high frequency and low frequency applications. It is even possible to identify transponders with identical codes, thereby making it possible to count identical items. The chip is typically used in “passive” transponder applications, i.e. it does not require a battery power source. Instead, it is powered up by an electromagnetic energy field or beam transmitted by the reader, which is received and rectified to generate a supply voltage for the chip. A preprogrammed code is transmitted to the reader by varying the amount of energy that is reflected back to the reader. This is done by modulating an antenna or coil, thereby effectively varying the load seen by the reader. Typical Applications Access control Animal tagging Asset control Sports event timing Licensing Electronic keys Auto-tolling Features Implements all BTG anti-collision protocols: Fast SWITCH-OFF, SLOW-DOWN, and FREE-RUNNING Can be used to implement low frequency inductive coupled transponders, high frequency RF coupled transponders or bi-frequency transponders Reading 500 transponders in less than one second for high frequency applications Factory programmed 64 bit ID number Data rate options form 4 kbit/s to 64 kbit/s Manchester data encoding Any field frequency: Typically 125 kHz, 13.56 MHz inductive and 100 MHz to 2.54 GHz RF Data transmission done by amplitude modulation Trimmed 110 pF ± 3% on-chip resonant capacitor On-chip oscillator, rectifier and voltage limiter Low power consumption Low voltage operation : down to 1.5 V at ambient temperature ° -40 to +85 C operating temperature range Pin Assignment 11 EM 4 0 2 2 1 10 2 9 3 4 5 Pad N° 1 2 3 4 5 6 7 8 9 10 11 6 7 8 Name XCLK VDD M MTST COIL1 COIL2 VSSTST VSS GAP SI TMC Function external test clock input positive supply connection to antenna test output Coil terminal 1 Coil terminal 2 negative test supply output negative supply GAP input Serial test data input (pull down) Test mode control (pull down) Fig. 1 Copyright 2002, EM Microelectronic-Marin SA 1 www.emmicroelectronic.com EM4022 Typical Operating Configurations Low frequency inductive transponder . VDD M CPX Coil1 L EM4022 Medium frequency applications are those which cannot use the integrated full wave rectifier and where the transponder power is transmitted through a coil. External microwave schottky diodes are required to rectify the carrier wave. An external power storage capacitor can be added to improve reading range. These applications allow higher data rates (64 kbit/s). Where reading rates of 500 transponders per second can be achieved High frequency RF transponder implementation. Coil2 GAP VSS M Coil1 D1 Medium frequency (13.56 MHz) inductive transponder implementation VDD M CPX L EM4022 C CPX D3 Fig. 2 Low frequency applications are those applications that can make use of the on-chip full wave rectifier bridge to rectify the incident energy. These are typically applications that use inductive coupling to transmit energy to the chip. The carrier frequency is typically less than 500 kHz. The design of the on-chip rectifier and resonance capacitor is optimized for frequencies in the order of 125 kHz. Low frequency transponders can be implemented using just a EM4022 chip and an external coil that resonates with the on-chip tuning capacitor at the required carrier frequency. An external power storage capacitor is required to maintain the supply voltage above the integrated power on reset level. In a very strong field, due to the forward resistance of the diode, the GAP input must be limited at VSS-0.3V by a schottky diode (D1) VDD EM4022 D2* Coil2 GAP VSS D1 Fig. 4 D2 in figure 2 and 3 is optional and is only used for GAP enable versions. All diodes are schottky type. High frequency applications are similar to medium frequency applications. These are typically applications that use electromagnetic RF coupling to transmit energy to the chip using carrier frequencies greater than 100 MHz. High frequency transponders can be implemented using a EM4022 chip, two or three microwave diodes and a printed antenna. High frequency RF coupled applications typically have higher reading distances (> 4 m) and Bi-frequency applications are possible by implementing a coil between coil1 and coil2 connections in the high frequency application (fig. 4). D2* GAP VSS D1 Fig. 3 L: C: coil antenna (typical value 1.35 µH). tuning capacitor (typical value 100 pF) Copyright 2002, EM Microelectronic-Marin SA 2 www.emmicroelectronic.com EM4022 Absolute Maximum Ratings Parameter Maximum AC peak current induced on COIL1 and COIL2 Maximum DC voltage induced between M and VSS Maximum DC current supplied into M Power supply Max. voltage other pads Min. voltage other pads Storage temperature Electrostatic discharge maximum to MIL-STD-883C method 3015 Handling Procedures Symbol I COIL Conditions ± 30 mA VM 5V (note1) IM 60 mA (note1) VDD - VSS Vmax Vmax TSTORE VESD -0.3 to VM VDD + 0.3 V VSS - 0.3 V o -55 to +125 C 1000 V note1) whatever is reached first Stresses above these listed maximum ratings may cause permanent damage to the device. Exposure beyond specified operating conditions may affect device reliability or cause malfunction. This device has built-in protection against high static voltages or electric fields; however, due to the unique properties of this device, anti-static precautions should be taken as for any other CMOS component. Unless otherwise specified, proper operation can only occur when all the terminal voltages are kept within the supply voltage range. Operating Conditions Parameter Operating temperature Maximum coil current AC voltage on coil* DC voltage on M* Symbol Min Typ -40 TA -10 ICOIL VCOIL 3.5 VM Max Units o +85 C 10 mA 15 Vpp V * The AC voltage on the coil and the DC voltage at pad M are limited by the on-chip shunt regulator loaded at ICOIL in table 3 Electrical Characteristics O VSUPPLY between 2.0 V and 3.0 V, TA = 25 C, unless otherwise specified. Parameter Supply voltage (VDD - VSS) Regulated voltage Oscillator frequency Power-on reset threshold Power-on reset threshold Power-on reset hysteresis GAP input time constant Symbol VSUPPLY VM FOSC VPONR VPONF VPHYS TGAP Modulation transistor ON resistance Resonance capacitor Supply capacitor Current consumption in modulation state Shunt Regulator current consumption Gap pull-up current consumption Dynamic current consumption RON CR CSUP IMOD ISHUNT IGAP IDYN Test conditions IM = 50 mA VSUPPLY = 3 V VSUPPLY rising VSUPPLY falling Extrapolated with an external capacitor of 64nF VSUPPLY = 3 V f = 100KHz, 100mVpp f = 100KHz, 100mVpp VSUPPLY = 2 V VSUPPLY = 2V VGAP = 0V, VSUPPLY = 2V fOSC = 128KHz, VSUPPLY = 2V Min VPONR +100mV 3.3 92 0.9 0.7 80 106.7 6 1.8 3.5 Typ 4 125 1.4 1.2 160 0.4 4 110 140 9 200 5 5 Max Units VM V 4.7 V 160 kHz 1.8 V 1.6 V 240 mV µs 8 113.3 13 500 7 6.5 Ω pF pF µA nA µA µA Timing Characteristics 1) All timings are derived from the on-chip oscillator. 2) The minimum low frequency GAP width for a single chip is 1 bit at its own clock frequency. The reader must however allow for the spread in clock frequencies possible in a group of tags. Therefore the minimum width of the GAP in MUTE and WAKE-UP signals must be 1.5 bits. High frequency GAPs can be arbitrarily. 3) The maximum GAP width for a single chip is 6 bits at its own clock frequency. The reader must however allow for the spread in clock frequencies possible in a group of tags. Therefore the maximum width of the GAP in MUTE and WAKE-UP signals must be 5 bits. Parameter High frequency GAP width High frequency ACK GAP width High frequency MUTE and WAKE-UP GAP width Low frequency ACK GAP width Low frequency MUTE and WAKE-UP GAP width GAP separation in WAKE-UP signal Copyright 2002, EM Microelectronic-Marin SA Symbol Test conditions THFGAP W HFACK W HFMUTE W LFGAP W LFACK W LFMUTE 3 Min 50 Typ 1.0 1.5 1.5 2 2 2 Max Units ns 6 bit 5 bit 6 bit 5 bit 5 bit www.emmicroelectronic.com EM4022 Calculation example : Below we define typical cases combinations : FOSC = 125 KHz VPHYS = 120 mV IMOD = 9 µA Data rate is 4 KBaud. Power storage capacitor calculation The global current consumption of the device defines the external storage capacitor. When the device modulate, the supply voltage is picked from the supply capacitor and should never decrease under the falling edge of the power on reset (VPONF). If this occurs, the device goes in a reset mode and any data transmission is aborted. The worst case for the storage capacitor calculation is when the device is put in the electromagnetic field. At this moment the supply reaches the VPONR and start to modulate. During modulation the power store in the capacitor must be high enough so that at the end of the modulation the supply is higher than VPORF.. This means that the voltage reduction on the capacitor must be less than the hysteresis of the power on reset (VPHYS). And this when the chip has a supply voltage of around the power on reset threshold The total current consumption from the storage capacitor is defined by the modulation current IMOD, This current is the consumption of the power on reset block, oscillator and the logic which work at a typical frequency of 125KHz. The GAP current is also included in this parameter. The duration where this currents is present for the capacitor calculation, is dependent of the data rate CPx = = I MOD * 128 * 103 FOSC *VHYS * BaudRate 9 * 10− 6 * 128 * 103 = 14.4nF 125 * 103 * 160 * 10− 3 * 4 * 103 Of course, this value can be adapted to the electromagnetic power and to the performances that must be achieved. If a tag is put in a field within a short time, the emitting power must be high enough to charge up the capacitor. The chip integrates a 140pF supply capacitor. Block Diagram VDD M VDD P COIL1 D2 R PON Q1 LOGIC CP D4 Shunt CR D3 N Q2 GAP C TST VSS COIL2 OSC VSS D1 DG GAP VDD CG VDD RG VSS VSS VSS SI XCLK TMC Fig. 5 Copyright 2002, EM Microelectronic-Marin SA 4 www.emmicroelectronic.com EM4022 Functional description Resonance capacitor The resonance capacitor CR has a nominal value of 110 pF and is trimmed to achieving a high stability over the whole production. For resonance at 125 kHz an external 14.7 mH coil is required. At 13.65 MHz the required coil inductance drops to 1.2 µH. Rectifier bridge Diodes D1-D4 form a full wave rectifier bridge. They have relatively large forward resistances (100 -200 Ω). This is sufficient at 125 kHz, where the output impedance of the tuned circuit is high, but at 13.5 MHz the diode resistance becomes significant and external diodes have to be used to bypass the internal ones. The diode resistance affects the rate at which the power capacitor CP can be charged. It also affects the modulation depth that can be achieved. Shunt regulator The shunt regulator has two functions. It limits the voltage across the logic and in high frequency applications it limits the voltage across the external microwave Schottky diodes, which typically have reverse breakdown voltages of 5 V. Oscillator The on-chip RC oscillator has a center frequency of 128 kHz. It gives the main clock of the logic and defines the effective data/rate. Power-on reset (PON) The reset signal keeps the logic in reset when the supply voltage is lower than the threshold voltage. This prevents incorrect operation and spurious transmissions when the supply voltage is too low for the oscillator and logic to work properly. It also ensures that transistor Q2 is off and transistor Q1 is on during power-up to ensure that the chip starts up. Modulation transistor The N channel transistor Q2 is used to modulate the transponder coil or antenna. When it is turned on it loads the antenna or coil, thereby changing the load seen by the reader antenna or coil, and effectively changing the amount of energy that is reflected to the reader. Its low on resistance is especially designed for high frequency applications. Charge preservation transistor The P channel transistor Q1 is turned off whenever the modulation transistor Q2 is turned on to prevent Q2 from discharging the power storage capacitor. This is done in a non-overlapping manner, i.e. Q1 is first turned off before Q2 is turned on, and Q2 is turned off before Q1 is turned on. Copyright 2002, EM Microelectronic-Marin SA Gap detection Poly-silicon diode DG is used to detect a gap in the illuminating field. It is a minimum sized diode with forward resistance in the order of 2 kΩ.==The low pass filter shown diagrammatically as CG and RG actually consists of a pull-up transistor (approximately 100 kΩ) in conjunction with the parasitic capacitance of the GAP input pad (approximately 2.5 pF). Through the diode the GAP input will be pulled low during each negative going cycle of the carrier. When the carrier is switched off, the GAP input will be pulled high by the pull-up transistor. At very high carrier frequencies (> 100 MHz) the carrier will be filtered out, so that the GAP input will be low continuously when the carrier is present. When the carrier disappears, the GAP input will go high with the time constant of the low pass filter. At very low frequencies the GAP input will go high and low at each cycle of the carrier, and will stay high when the carrier disappears. To detect the gap, the logic must check for a high period longer than the maximum high period of the carrier. As the rise and fall times of the GAP can be slow, a Schmitt trigger is used to buffer the GAP input. LOGIC block Depending on the state of the SI input at power-up, the EM4022 either enters a test mode (SI = 1) or its normal operating mode (SI = 0). The SI pin is internally pulled down, so that it can be left open for normal operation. After the power-on reset has disappeared, the chip boots by reading the SEED and CTL ROMs. The chip then enters its normal operating mode, which basically consists of clocking a 16 bit timer counter with the bit rate clock until it compares with the number in the random number generator. At this point a code (which is stored in the ID ROM) is transmitted with the correct preamble at the correct data rate and encoded correctly. The random number generator is clocked to generate a new pseudo random number, and the 16 bit counter is reset to start a new delay. The width of the comparison between the 16 bit random number and the 16 bit delay count determines the maximum possible delay between transmissions (repetition rate). Any one of eight maximum delay settings can be pre-programmed. The basic free-running mode as described above can be modified by the reception of GAP (MUTE and ACK) signals, if these are enabled by the CTL bits. If an ACK signal is received after transmission of a code, the chip either turns itself off completely or reduces the rate at which the delay counter is clocked, thereby slowing down the rate at which codes are transmitted. If a MUTE signal is received while the chip is not transmitting, the current operation of the chip is interrupted for 128 clock periods, after which it continues normally. Reception of more MUTEs during the sleep state restarts the sleep state. The sleep state is also terminated by the reception of a WAKE-UP signal (an ACK signal to a chip which has just completed transmitting). 5 www.emmicroelectronic.com EM4022 ACK timing diagram ACK timing Bit n Clock Data HF ACK LF ACK T1 T1 Fig. 6 GAP Detection Algorithm The GAP detection logic contains two main controllers, one for detecting the ACK signal, and one for detecting the MUTE and WAKE-UP signals. The WAKE-UP signal is also called an asynchronous ACK, as it is really an ACK meant for another chip. It also contains a preprocessor for low frequency GAP signals. Refer to the timing diagrams in Figure 6 and 7 for the following detailed description of the GAP detection algorithms. ACK The controller checks for a LOW 1.75 bit periods after the last bit of code has been transmitted. It then checks for a HIGH 3 bits later, a LOW 3 bits later and finally a HIGH a further 3 bits later. The reader should synchronise itself to the frequency of the received code, check the CRC and then send two GAPs so that the above pattern is matched. Ideally to achieve the lowest error rate, the GAPS should be as narrow as possible and situated 4.75 and 7.75 bits after the last bit of code. In practice allowance must be made for the fact that the on-chip oscillator can drift in the time between when the last code bit is transmitted and when the GAPs are expected. One reason for the drift is that the oscillator is supply voltage dependent, and the supply voltage will typically be rising during this time, since the transponder will not be modulating its coil or antenna. The slope of the rising and falling edges of the GAPs can also be adjusted to reduce reader power bandwidth. In the case of high frequency GAPs the envelope is used directly. Low frequency GAPs have to be pre-processed. They are detected by checking for high periods lasting longer than one bit period. For this reason there is a setup time of 1 bit. The minimum GAP width is therefore 1 bit period (T1 in the timing diagram). Copyright 2002, EM Microelectronic-Marin SA MUTE The MUTE signal is received asynchronously by the transponder. The controller checks for a HIGH less than 7 bits wide after pre-processing (T2 in the timing diagram). As in the case of the ACK, low frequency MUTE GAPs must be at least one bit wide (T1 in the timing diagram), but high frequency GAPs can be arbitrarily narrow. When transmitting a MUTE, the reader must take into account that there could be a spread in the clock frequencies of all the receiving transponders. The reader should therefor limit the width of a MUTE to be less than 5 bits of the nominal bit rate (T4 in the timing diagram). A low frequency MUTE should also be wider than 1.5 bits of the nominal bit rate (T3 in the timing diagram). The MUTE should be sent as early as possible after a code transmission has been detected, while still making sure that it is a code transmission and not just noise. The earlier the MUTE is sent, the more time the reader has to recover before the SYNCH and code bits arrive, and the smaller the probability that another transponder has started a colliding transmission 6 www.emmicroelectronic.com EM4022 MUTE and WAKE-UP timing diagrams Fig. 7 WAKE-UP An ACK sent after correct reception of a code is interpreted by the other transponders in the field as a WAKE-UP. The ACK arrives synchronously at the transponder that has just transmitted, but asynchronously at all the other transponders. If necessary, a WAKE-UP can also be sent if the code is not received correctly, ensuring that it will not be interpreted as an ACK by the transmitting transponder. This could speed up the protocol, but runs the risk of turning transponders off by accident. To detect a WAKE-UP, the chip checks for two GAPs, less than 7 bits apart and each less than seven bits wide. As with the MUTE allowance must made for the spread in clock frequencies. To be safely interpreted as a WAKE-UP, the GAPs should be sent less than 5 bits apart, and each should be less than 5 bits wide. Copyright 2002, EM Microelectronic-Marin SA This has an implication in the case of the high frequency ACK, which could theoretically consist of two very narrow GAPs 6 bits apart. In practice though, the GAPs will be typically at least one bit wide, making the separation five bits. Like the MUTE, the low frequency ACK GAPs should be at least 1.5 bits wide to serve as a reliable WAKE-UP. It should be noted that failure to reliably recognise WAKE-UPs is not critical. The protocol might be slowed down marginally, but will still work, as the chips time-out of the sleep mode automatically after 128 bits. 7 www.emmicroelectronic.com EM4022 Data Encoder The transmitted code always consists of an 11 bit preamble followed by the 64 code bits. The preamble consists of 8 start bits (ZEROES), followed by a SYNCH. The SYNCH consists of a LOW for two bit periods followed by a ONE. The EM4022 can be programmed for one of two data encoding methods. The first method is a variation on Manchester II, i.e. a ONE is represented by a HIGH in the first half of a bit period, and a ZERO is represented by a LOW in the first half of a bit period. The second encoding method is called GLITCH encoding. A ONE is represented by a HIGH in the first quarter of the bit period, while a ZERO is represented by a HIGH in the third quarter of the bit period. In GLITCH encoding the longest modulation period is one quarter of a bit period, compared to the Manchester encoding, where the longest modulation period is one full bit period. GLITCH encoding therefore requires a much smaller power storage capacitor. Data Encoding "0" "0" "1" Fig. 8 ROM programming The EM4022 contains three laser fuse ROM blocks that are pre-programmed by the foundry. CONTROL ROM The operational modes of the EM4022 are preprogrammed into the CONTROL ROM. The contents of this one is not read out. CODE ID ROM This ROM contains the 64 bit ID code. The foundry will automatically program a unique 48 bit ID and 16 bit CRC. In this case the most significant bit of the ID is programmed into bit 0 of the ROM, which will be transmitted first. CRC Block Diagram 15 14 13 12 11 10 9 8 7 6 5 4 3 2 Data Input 1 0 MSB LSB X15 X2 X0 Feedback before shift Exclusive OR X Shift Register CRC-CCITT Generating polynomial = X15 + X2 + X0 Fig.9 Copyright 2002, EM Microelectronic-Marin SA 8 www.emmicroelectronic.com EM4022 Anti-collision Protocol Overview The protocols are a collection of simple but fast and reliable anti-collision protocols. They allow fast reading of large numbers of transponders simultaneously using a single reader. It is even possible to identify transponders with identical codes, thereby making it possible to count identical items. Control ROM Bit definition Parameter Value Fast / Normal 0 Mode 1 Free-running 0 1 ACK mode 0 1 Maximum 0 initial random 1 delay 2 3 4 5 6 7 Data rate 0 1 2 3 4 5 6 7 Encoding 0 method 1 GAP type 0 1 Control ROM Map 15 14 13 Mode Normal Fast GAP detection enabled GAP disabled (Free-running) Slow-down Switch-off 0 (Continuous) 16 bits 64 bits 256 bits 1 kbits 4 kbits 16 kbits 64 kbits 64 kbit/s 32 kbit/s 16 kbit/s 8 kbit/s 4 kbit/s 2 kbit/s 1 kbit/s 0.5 kbit/s Glitch encoding Manchester encoding Low frequency GAP detection High frequency GAP detection 12 11 Byte[1] 10 9 HF GAP Manchester Copyright 2002, EM Microelectronic-Marin SA Free-running protocol The basis of the BTG-Supertag series of protocols is that transponders transmit their own codes at random times to a reader. By just listening and recording unique codes when they are received, the reader can eventually detect every tag. The reader detects collisions by typically checking a CRC. This basic protocol is known as the “Free-running” protocol. It requires uniquely coded tags. Its main advantage is that the reader design is simple, and the spectrum requirement is much less – a very narrow band is required. Figure 10 shows a sequence of three transponders. The reader starts first to read transponder 1 but during his data transmission, transponder 3 starts to modulate. In this case, due to the CRC check no transponder is detected. A transponder is taken into account, if it transmits a complete data stream without any disturbance 8 7 Data rate 9 6 5 4 3 Byte[0] Random delay 2 1 Switch- Freeoff running 0 Fast www.emmicroelectronic.com EM4022 Free running example Transponder 1 Transponder 2 Transponder 3 Data stream in collision Data stream detected Fig. 10 Bi-directional protocols Allowing bi-directional communication between reader and transponders can speed up the basic free-running protocol. Communication from the reader to transponders is achieved by turning the illuminating energy field off for short periods. The transponders detect these gaps in the energy transmission and interpret them as required. Switch-off and Slow-down Modes Reducing the effective population of transmitting transponders in the reader field can speed up the freerunning protocol. One method to achieve this is by either switching transponders off or slowing them down once they have been detected. To achieve this, the reader sends an ACK signal to a transponder after its code has been successfully received. The transponder then either switches off completely or reduces its repeat rate until it is powered down. This reduces the number of collisions between transponder transmissions, thereby reducing the time required to read a group of tags. Figure 11 shows a typical situation where a collision occurs between transponder 1 and 3. Then, as soon as transponder 2 is read, the reader sends an ACK signal to this one switching it off as long hi is powered up from the field. This eliminates the collision between transponder 1 and 2 in the next step The Switch-off protocol’s main advantage is that identical transponders can be counted. Copyright 2002, EM Microelectronic-Marin SA In the EM4022 the ACK signal is implemented as two consecutive gaps with the appropriate timing and received at a specific time after a code has been transmitted. The slow-down mode is a compromise between the freerunning mode and the switch-off mode. Etch time a transponder is read, the reader send an ACK to double the random repetition rate. This reduces the collisions and in the time increases the saturation level. Figure 12 show a typical case of this mode of operation. Fast Mode A second method of speeding up the reading of tags, is to inhibit other transponders from transmitting while one transponder is transmitting. This is done by sending a MUTE signal to all of the transponders when the start of a transmission is detected. The transponders stay muted long enough (128 bit of its own clock) to allow the transmission of one code (see figure 13). This allows the transponder that has started transmitting to complete its transmission without any collisions. The other transponders continue with their own protocols automatically after a time out, or continue immediately upon detection of an ACK signal indicating that the transmission that caused the MUTE has been completed. In the EM4022 the MUTE signal is implemented as a single gap received while the transponder is not transmitting. 10 www.emmicroelectronic.com EM4022 Switch-off example Transponder 1 Transponder 2 Transponder 3 Reader field Data stream in collision Transponder switched off Transponder detected Fig. 11 Slow-down example Transponder 1 Transponder 2 Transponder 3 Reader field Data stream in collision Transponder detected Fig. 12 Fast mode example Transponder 1 Transponder 2 Transponder 3 Reader field 128 bit shift Transponder detected Fig. 13 Copyright 2002, EM Microelectronic-Marin SA 11 www.emmicroelectronic.com EM4022 Protocol combinations The free-running and the two basic bi-directional protocols, switch-off and slow-down, can all be combined with the Fast protocol to give six different protocols, i.e. Normal free-running, slow-down, Normal switch-off, Fast free-running, slow-down, and Fast switch-off. The following should be noted about the different protocols: 1) The switch-off protocols must be used for counting applications. 2) All the protocols except the switch-off protocols have built in redundancy because of the fact that they can transmit a code more than once. 3) Normal free-running is the only unidirectional protocol. It has the lowest power spectrum requirement because the reader transmits a CW wave. 4) Fast switch-off and Fast slow-down are the fastest protocols, and should be used where speed is important, or where the data rate limits the reading rate. Fast slowdown is slightly slower, but theoretically has a lower error rate. 5) For 125 kHz inductive applications using a 4 kbit/s data rate, Fast slow-down is probably the best overall protocol. 6) For RF applications using a 64 kbit/s data rate, normal free-running protocol is probably the best protocol. Reader determined protocols If the reader does not send MUTE signals to transponders that were programmed for one of the FAST protocols, the protocol merely reverts to the equivalent normal protocol. Similarly, if the reader does not send ACK signals to transponders that were programmed for SLOW-DOWN or SWITCH-OFF, the protocol reverts to a FREE-RUNNING protocol. In this manner, the reader can determine the protocol that is used. Protocol saturation As the number of transponders in a reader beam is increased, the number of collisions increase, and it takes longer to read all the tags. This process is not linear. To read twice as many transponders could take more than twice as long. This effect is called protocol saturation. The normal free-running protocol saturates the easiest of all the protocols, because it does not have any means of reducing the transmitting population. The Fast protocols, on the other hand, are virtually immune against saturation, as they prevent collisions by muting all transponders except the transmitting one. One way of delaying the onset of saturation, is to reduce the initial repeat rate (not data rate) at which transponders transmit their codes. This is done by increasing the maximum random delay between transmissions. Figure 14 and 15 below show's reading times for some possible versions Optimum repeat delay settings The following table lists the optimum repeat delay settings for each of the protocols vx number of transponders in a group. Protocol Free-running Slow-down Switch-off Fast Free-running Fast Slow-down Fast Switch-off Number of transponders 3 10 30 100 1k 4k 16k 64k 1k 1k 4k 16k 1k 1k 4k 16k 256 1k 1k 4k 256 256 1k 1k 256 256 1k 1k Note, however, that unless a transponder was specifically programmed for the free-running protocol, its GAP input must be pulled down. This happens automatically in low frequency inductive applications, where the GAP input is pulled down by the internal GAP detector diode. In RF applications, however, the GAP input will have to be pulled down explicitly. Copyright 2002, EM Microelectronic-Marin SA 12 www.emmicroelectronic.com EM4022 Reading time versus quantities of transponders and protocol (4Kbauds) 100 versions 10 Reading time [s] V11 V14 V12 V15 V10 V13 1 0.1 1 10 100 1000 Number of tags Fig. 14 Reading time versus quantities of transponders and protocol (64Kbauds) 10 versions 1 reading time [s] V17 V20 V18 V21 V16 V19 0.1 0.01 1 10 100 1000 Number of transponders Fig. 15 Copyright 2002, EM Microelectronic-Marin SA 13 www.emmicroelectronic.com EM4022 Chip and Packaging Information Pad Location Chip size is 57 x 69 mil 11 EM 4 0 2 2 1 10 2 9 3 4 5 6 7 8 Fig. 16 Pad location table with reference on pad 3 center : Pad N° X [µm] Y [µm] size [µm] Pad name Function 1 0 2 0 417 98/98 XCLK external test clock input 234 98/98 VDD 3 positive supply 0 0 98/98 M connection to antenna 4 125 0 98/98 MTST test output 5 228 0 175/98 COIL1 Coil terminal 1 6 1037 0 175/98 COIL2 Coil terminal 2 7 1200 -13 98/98 VSSTST negative test supply output 8 1324 1 98/98 VSS negative supply 9 1324 196 98/98 GAP GAP input 10 1324 374 98/98 SI Serial test data input (pull down) 11 1324 552 98/98 TMC Test mode control (pull down) Test inputs and outputs must be left open. Copyright 2002, EM Microelectronic-Marin SA 14 www.emmicroelectronic.com EM4022 Ordering Information for samples For other versions or other delivery form, please contact EM Microelectronic-Marin S.A. Please make sure to give complete part number when ordering (without spaces between letters). Part Number EM4022 V10 WS11 Control ROM [hex] 302 Data rate Random Value Protocol Die Form & Thickness Bumping EM internal use only: Old version 4 Continuous Free-Running Sawn wafer, 11 mils no bumps 010 EM4022 V11 WS11 328 4 4096 Slow down Sawn wafer, 11 mils no bumps 011 EM4022 V12 WS11 32C 4 4096 Switch off Sawn wafer, 11 mils no bumps 012 EM4022 V13 WS11 303 4 Continuous Fast Free-Running Sawn wafer, 11 mils no bumps 013 EM4022 V14 WS11 321 4 1024 Fast Slow down Sawn wafer, 11 mils no bumps 014 EM4022 V15 WS11 325 4 1024 Fast Switch off Sawn wafer, 11 mils no bumps 015 EM4022 V16 WS11 202 64 Continuous Free-Running Sawn wafer, 11 mils no bumps 016 EM4022 V17 WS11 228 64 4096 Slow down Sawn wafer, 11 mils no bumps 017 EM4022 V18 WS11 22C 64 4096 Switch off Sawn wafer, 11 mils no bumps 018 EM4022 V19 WS11 203 64 Continuous Fast Free-Running Sawn wafer, 11 mils no bumps 019 EM4022 V20 WS11 229 64 4096 Fast Slow down Sawn wafer, 11 mils no bumps 020 EM4022 V21 WS11 EM4022 V%% WS11 220 Fast Switch off custom Sawn wafer, 11 mils Sawn wafer, 11 mils no bumps no bumps 021 %%% 64 4096 custom custom For ICs to be used in mass production, the customer must define its options with the control ROM bit definition (page 9-10). Using this information, EM Microelectronic-Marin S.A. will define a complete new Part Number for ordering. WARNING: Use of this product is subject to license from British Technology Group (BTG, www.btg-et.com) Product Support Check our Web Site under Products/RF Identification section. Questions can be sent to [email protected] EM Microelectronic-Marin SA cannot assume responsibility for use of any circuitry described other than circuitry entirely embodied in an EM Microelectronic-Marin SA product. EM Microelectronic-Marin SA reserves the right to change the circuitry and specifications without notice at any time. You are strongly urged to ensure that the information given has not been superseded by a more up-to-date version. © EM Microelectronic-Marin SA, 01/02, Rev. D/414 Copyright 2002, EM Microelectronic-Marin SA 15 www.emmicroelectronic.com