AMMCL004AWP

PRELIMINARY AmMCL00XA 2 or 4 Megabyte 3.0 Volt-only Flash Miniature Card DISTINCTIVE CHARACTERISTICS s 2 or 4 Mbytes of addressable Flash memory s 2.7 V to 3.6 V, single power supply operation — Write and read voltage: 3.0 V –10/+20% — No additional supply current required for VPP s Fast access time — 150 ns maximum access time s CMOS low power consumption — Typical active read current: 35 mA (word mode) — Typical active erase/write current: 40 mA (word mode) — Typical standby current: 10 µA (4 Mbyte); 5 µA (2 Mbyte) s High write endurance — Guaranteed minimum 100,000 write/erase cycles per card — More than 1,000,000 cycles per card typical s Uniform sector architecture — 64K byte individually useable sectors — Erase Suspend/Resume increases system level performance — BUSY# and RESET# signals s Zero data retention power — No power required to retain data s Available in industrial temperature grade (–40°C to +85°C) s Miniature Card standard form factor — True interchangeability — 60-pad elastomeric connector — Supports multiple technologies — Sonic welded stainless steel case — PCMCIA Type II adapter available — Selectable byte- or word-wide configuration — Small Form Factor (38 mm x 33 mm x 3.5 mm) s 60 connection bus — 16-bit data bus — 25-bit address bus — Easy system integration — Low cost implementation — Low cost cards s Consumer-friendly mechanicals — User can easily insert and remove card, upgrade memory, and add applications s Voltage level keying — Does not allow a 3 V card to plug into a 5 V system and vice versa — Single power supply design — System does not need a separate program voltage supply; only one is necessary to read and write GENERAL DESCRIPTION The Miniature Card is an expansion card that provides a low cost, low power, high-performance, small form factor solution for data and file storage to the portable, handheld market, which includes audio, digital film, wireless, and PDA (Portable Digital Assistant) applications. Miniature cards can be easily “snapped” into the back of an electronic system and can be readily removed and replaced by end users. AMD’s 3 V Flash Miniature Cards are manufactured using AMD’s industry leading 3.0 volt-only, single-power-supply Am29LV081 Flash Memory device, ensuring high reliability and excellent performance. The Miniature Card is less than 30% of the size of a PCMCIA memory card. Applications include digital voice recorders, pocket PCs and intelligent organizers, smart cellular telephones, voice and data messaging pagers, digital still cameras and portable instrumentation equipment. The Miniature Card specification will be defined by PCMCIA as of October 1997. The participating association members include major Flash memory vendors and leading consumer electronics OEMs. The goal of the Miniature Card specification is to promote an open, Publication# 21138 Rev: E Amendment/0 Issue Date: September 1997 This document contains information on a product under development at Advanced Micro Devices. The information is intended to help you evaluate this product. AMD reserves the right to change or discontinue work on this proposed product without notice. PRELIMINARY interoperable small-form-factor memory card standard. For more information on the Miniature Card specification, visit the PCMCIA web site at http://www.pc-card.com. AMD Flash Miniature Cards can be read in either a byte-wide or word-wide mode, which allows for flexible integration into various system platforms. Compatibility is assured at the hardware interface and software interchange specification. The Miniature Card is also designed with low-cost and rugged handling in mind. The card contains virtually no control logic, which keeps cost and power consumption to a minimum. The Miniature Card is packaged in a sonic welded, stainless steel case that guarantees durability, provides good ESD protection and ease of handling. The Miniature Card has extensive third-party support, including socket and connector solutions, software support from the major FTL software vendors, and PCMCIA adapter solutions and programmer support. AMD's Miniature Flash cards can be used for both code and data storage. Since fast random access is possible, code can be directly executed from the card, reducing the amount of system RAM required. In addition. AMD’s Flash technology offers unsurpassed endurance, data retention and reliability, eliminating the need for complex error correction and defect management hardware and software. Each Flash sector provides a minimum of 100,000 cycles, and a typical card life of one million or more cycles. For more information, please contact your local AMD sales office or visit our Web site at http://www.amd.com/html/products/nvd/nvd.html. DEFINITIONS Table 1 lists the terms and definitions that may be used in conjunction with Miniature Card specifications. Table 1. Term AIS ESD FAT Flash Host Miniature Card Definitions Meaning Acronym for Attribute Information Structure. AIS is a Miniature Card specification for storing Miniature Card attribute information. Acronym for Electrostatic Discharge. ESD is part of the Miniature Card physical test. Acronym for File Allocation Table. Using an FAT is a common method for managing files in a DOS-based system. A type of non-volatile memory that is both readable and writeable, but requires the media to be erased before it is rewritten. Any system that incorporates a Miniature Card socket. User Perception: Insertion of the Miniature Card when the host is off. Insertion, Cold Host State: The host would be either off or in sleep mode, no bus activity is occurring, the host is non-operational by the user. The user inserts the Miniature Card and then presses a button to turn the host on before the system is operational. User Perception: Insertion of a Miniature Card when the host is running. Insertion, Hot Host State: The host would be in running mode, bus activity is occurring, the host is operational by the user. The user inserts the card, the host recognizes it, and the host continues to be operational. Note: Hot insertion may require buffering on the host system for proper operation. User Perception: Insertion of a Miniature Card when the host is running. Insertion, Pseudo Hot Host State: The host would be in running mode, bus activity is occurring, the host is operational by the user. The user inserts the card, the host immediately powers off before the Miniature Card makes contact with the host’s internal bus. The user would then need to press a button to turn the host on for it to become operational. Miniature Card signals that make connection through the 60-pad connector area. Acronym for Joint Electronic Device Engineering Council. The side of the Miniature Card that contains the latching mechanism. The backside is opposite the frontside. The side of the Miniature Card that contains the interface signals. The bottomside is opposite the topside. Interface Signals JEDEC Miniature Card Backside Miniature Card Bottomside 2 AmMCL00XA PRELIMINARY Table 1. Term Miniature Card Frontside Miniature Card Topside PC Card PC Card Adapter Miniature Card Definitions (Continued) Meaning The side of the Miniature Card that contains power, insertion, ground, voltage keys, and alignment notch. The frontside is opposite the backside. The side of the Miniature Card that contains the Miniature Card label. The topside is opposite the bottomside. A memory or I/O card compatible with the PC Card Standard. The hardware that connects the Miniature Card 60 contact bus to the PC Card 68 pin bus. This hardware can be mechanically implemented by following the PC Card Type II specification. The three signals on the frontside of the Miniature Card that provide ground, power and early detection of insertion. Resistors used to ensure that signals do not float when no device is driving them. Power/Insertion Signals Pull-Ups User Perception: Removal of a Miniature Card when the host is off. Removal, Cold Host State: The host would either be off or in sleep mode, no bus activity is occurring, the host is non-operational by the user. User would turn off the host, then remove the Miniature Card and then press a button to turn the host on for it to become operational again. User Perception: Removal of the Miniature Card when the host is running. Removal, Hot Host State: The host would be in running mode, bus activity is occurring, the host is operational by the user. User removes the card, the host recognizes the event, and the host continues to be operational. User Perception: Removal of the Miniature Card when the host is running. Removal, Pseudo Hot Host State: The host would be in running mode, bus activity is occurring, the host is operational by the user. User removes the card, the host recognizes the event, the host immediately powers off before the Miniature Card removes contact with the host’s internal bus. The user would then need to press a button to turn the host on for it to be operational again. Usually 64 KBytes. In word mode, a sector is 64 Kwords. An element of the PC Card Standard CIS that provides card attribute information, and a link to the next tuple in a string of tuples. All Miniature Cards should be inserted into the host by the user without the need for any special tools. This type of Miniature Card can be removed by the user without the need for any special tools. It contains programs and data that users may want to switch often. The use of this type of card is similar to a floppy disk. This type of Miniature Card must be removed by the user with a special tool. It contains memory upgrades or boot program that users switches only when they require an upgrade. The use of this type of card is similar to a SIMM memory expansion or boot hard disk. Acronym for eXecute-In-Place, which refers to code that executes directly from a Miniature Card. Sector Tuple User Insertable User Removable User Non-Removable XIP AmMCL00XA 3 PRELIMINARY Write Protect Switch (optional) Pad 60 Pad 31 Pad 30 VCC 3V/5V Key Alignment Notch Pad 1 CINS# GND 21138E-1 Figure 1. Miniature Card Connector (Card Bottom View) Note: Refer to the Physical Dimensions section for more information. Also refer to the MCIF specification for detailed mechanical information, available on the Web at http://www.mcif.org. Table 2. Family Part Number AmMCL002AWP AmMCL004AWP AMD Flash Miniature Cards and Flash Devices Density 2 Mbyte 4 Mbyte No. of Flash Devices 2 4 AMD Flash Memory Am29LV081 Am29LV081 4 AmMCL00XA PRELIMINARY BLOCK DIAGRAM VCC VCC 100K 100K VCC BUSY# RESET# RY/BY# 10K RESET# to all Flash devices WE# Write Protect Switch WE# to all Flash devices OE# D8-D15 D0-D7 A0-A20 OE# to all Flash devices VCC 100K VCC 100K A20 VSS VCC A0-A19 D0-D7 CE# WE# S0** OE# RESET# RY/BY# CEL0# CEH0# CEL1# CEH1# VSS VCC A0-A19 D8-D15 CE# WE# S1** OE# RESET# RY/BY# CEL# CEH# VSS VCC A0-A19 D0-D7 CE# WE# S2** OE# RESET# RY/BY# VSS VCC A0-A19 D8-D15 CE# WE# S3** OE# RESET# RY/BY# Decoder* 21138E-2 * 4 Mbyte card only. Not used on 2 Mbyte card. ** 2 Mbyte card: Two Am29LV081 devices, S0 and S1 4 Mbyte card: Four Am29LV081 devices, S0...S3 Note: On the 2 Mbyte card, A20–A24 are not connected. On the 4 Mbyte card, A21–A24 are not connected. Connections not shown in this diagram are not connected internally. AmMCL00XA 5 PRELIMINARY MINIATURE CARD PAD ASSIGNMENTS A0–A24 Address A0 to A24 are the address bus lines that can address up to 32 Mwords (64 Mbytes). The address lines are word addressed. The Miniature Card specification does not require the Miniature Card to decode the upper address lines. A 2 Mbyte Miniature Card that does not decode the upper address lines would repeat its address space every 2 Mbytes. Address 0h would access the same physical location as 200000h, 400000h, 600000h, etc. On the 2 Mbyte cards, A20– A24 are not connected. On the 4 Mbyte cards, A21–A24 are not connected. RESET# RESET# controls card initialization. When RESET# transitions from a low state to a high state, the Miniature Card resets to the Read state after a maximum delay of 20 µs. BUSY# BUSY# is a signal generated by the card to indicate the status of operations within the Miniature Card. When BUSY# is high, the Miniature Card is ready to accept the next command from the host. When BUSY# is low, the Miniature Card is busy and unable to accept most data operations from the host. In Flash Miniature Cards the BUSY# signal is tied to the components’ RY/BY# signal. D0–D15 Data lines D0 through D15 constitute the data bus. The data bus is composed of two bytes; the low byte is D0–D7 and the high byte is D8–D15. These lines are tristated when OE# is high. CD# CD# is a grounded interface signal. After a Miniature Card has been inserted, CD# will be forced low. The card detect signal is located in the center of the second row of interface signals, and should be one of the last interface signals to connect to the host. Do not confuse CD# with CINS#. OE# OE# indicates to the card that the current bus cycle is a read cycle. The output enable access time (tOE) is the delay from the falling edge of OE# to valid data at the output pins (assuming the addresses have been stable for at least tACC – tOE time). CINS# CINS# is a grounded signal on the front of the Miniature Card that is used for early detection of a card insertion. CINS# makes contact on the host when the front of the card is inserted into the socket, before the interface signals connect. WE# WE# indicates to the card that the current bus cycle is a write cycle. The falling edge of WE# (or CE#), whichever occurs later, latches address information and the rising edge of WE# (or CE#), whichever occurs first latches data/command information. BS8# The BS8# (Bus size 8) signal indicates to the Miniature Card that the host has an 8-bit bus. AMD Flash Miniature Cards ignore this signal (no internal connection). An 8-bit host must connect its D0–D7 data lines to D8–D15 on the Miniature Card to retrieve the upper (odd) byte. VS1# Voltage Sense 1 signal. This signal is grounded. VS2# Voltage Sense 2 signal. This signal is left open or not connected. GND Ground CEL# CEL# enables the low byte of the data bus (D0–D7) on the card. VCC Vcc is used to supply power to the card. CEH# CEH# enables the high byte of the data bus (D8–D15) on the card. NC No connect RFU Reserved for future use 6 AmMCL00XA PRELIMINARY 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 the following: AM MC L 004 A WP -150 I TEMPERATURE RANGE Blank = Commercial (0°C to +70°C) I = Industrial (–40°C to +85°C) SPEED OPTION WRITE PROTECT SWITCH OPTION WP = Switch installed REVISION LEVEL MEMORY CARD DENSITY 002 = 2 Megabyte Card 004 = 4 Megabyte Card 3 V, SINGLE SUPPLY OPERATION 2.7 V to 3.6 V, extended operating voltage MINIATURE CARD AMD AmMCL00XA 7 PRELIMINARY INTERFACE SIGNAL ASSIGNMENTS Pad Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Signal Name A18 A16 A14 NC CEH# A11 A9 A8 A6 A5 A3 A2 A0 NC A24 A23 A22 OE# D15 D13 Pad Number 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Signal Name D12 D10 D9 D0 D2 D4 RFU D7 NC NC A19 A17 A15 A13 A12 RESET# A10 VS1# A7 BS8# Pad Number 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Signal Name A4 CEL# A1 NC NC CD# A21 BUSY# WE# D14 RFU D11 VS2# D8 D1 D3 D5 D6 RFU A20 Note: NC = No Connect; RFU = Reserved for Future Use. FLASH MINIATURE CARD OPERATIONS Voltage Sensing AMD Miniature Cards provide two voltage sense signals for hosts that support multiple voltages. The multivoltage host can sense the voltage level of the Miniature Card and power up the card at that voltage. S ee Tab l e 3 f o r a de s c r i pt i on o f th e v ol t ag e sense signals. In addition to the voltage sense pins, there are also mechanical voltage keys on the Miniature Card that ensure the card can only be inserted into host systems that can supply the proper voltage levels to the card. Refer to Section 4.1.2 in the Miniature Card specification for more information on mechanical keying. Table 3. Voltage Sense Signals Miniature Card Power-Up Voltage 3 volt-only VS1# Gnd VS2# Open 8 AmMCL00XA PRELIMINARY Data Accesses The Miniature Card has a 16-bit data bus that can accommodate word or byte accesses. By individually asserting CEL# and CEH#, a host can access either byte. However, byte swapping (moving the high byte data to the low byte) is not supported. Figure 2 shows the connections between the host and Miniature Card. The host system address lines range from A0–A25, whereas the Miniature Card address lines range from A0–A24. On the host, A0 and the byte/word line are sent to a decoder and output to CEL# and CEH# on the Miniature Card. These two bits enable a single device for byte accesses and two devices for word accesses, as shown by the decoder truth table in Figure 2. Again, the Miniature Card address lines do not receive input from host address bit A0. In this document, all address references are card addresses, unless otherwise noted. Table 4 shows the read/write modes for Miniature Cards. A0 Byte/Word Decoder Decoder Truth Table Input Output CEL# CEH# A0 B/W 0 0 0 1 0 1 0 0 0 1 0 1 0 0 Host Bus A24 A25 60-Pad Connector A23 A22 A21 A2 A1 1 1 A24* A23* A22* A21* A20** A1 A0 CEL# CEH# Card Bus * ** Not connected Not connected on 2 Mbyte card 21138E-3 Figure 2. Host/Card Address Connections Word-Wide Operations The AMD Miniature Card provide the flexibility to operate on data in a byte-wide or word-wide format. In word-wide operations, the low bytes are controlled with CEL#. The high bytes are controlled with CEH#. Refer to the block diagram for more information. Card Detection Each CD# (output) pin should be detected by the host system to determine if the memory card is adequately seated in the socket. CD# and CINS# are internally tied to ground. If both bits are not detected, the system should indicate that the card must be re-inserted. Byte-Wide Operations Byte-wide data is available for read and write operations (CEL# = 0, CEH# = 1). Even and odd bytes are stored in separate memory devices (for example, S0 and S1) and are accessed by controlling CEL# and CEH#. The even byte is the low order byte and the odd byte is the high order byte of a 16-bit word. Each memory sector or device pair must be addressed separately for erase operations. Refer to the block diagram for more information. Data Protection An optional mechanical write protect switch provides user-initiated write protection. When this switch is activated, WE# is internally forced high. The Flash memory command register is disabled from accepting any write commands. This prevents the card from responding to any commands (for example, an Autoselect command). See Figure 3. AmMCL00XA 9 PRELIMINARY ensure that the control pins are in the correct logical state when VCC > VLKO to prevent unintentional writes. Write Pulse “Glitch” Protection Write Enabled Noise pulses of less than 5 ns (typical) on OE#, CE#, or WE# will neither initiate a write cycle nor change the command registers. 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. Power-Up Write Inhibit Power-up of the device with CE# = WE# = VIL and OE# = VIH will not accept commands on the rising edge of WE#. The internal state machine is automatically reset to the read mode on power-up. Write Disabled Figure 3. Write Protect Switch (Card Right Side View) 21138E-1 In addition to card-level data protection, AMD Flash Miniature Cards offer several device-level data protection features. Device-Level Data Protection AMD Flash memory devices offer protection against accidental erasure or programming caused by spurious system level signals that may exist during power transitions. During power up, each device automatically resets the internal state machine to the read mode. The control register architecture allows alteration of the memory contents only occurs after successful completion of specific multi-bus cycle command sequences. AMD Flash memory devices also incorporates the following features to prevent inadvertent write cycles resulting from VCC 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 AMD memory devices in the Miniature Card lock out write cycles for VCC < VLKO (see “DC Characteristics” on page 22 for voltages). When V CC < V LKO , the command register is disabled, all internal program/erase circuits are disabled, and the device resets to the read mode. The memory devices ignore all writes until V CC > V LKO . The user must Read Mode Two Card Enable (CE#) pins are available on the memory card. Both CE# pins must be active low for word-wide read accesses. Only one CE# is required for byte-wide accesses. The CE# pins select and determine when to apply power to the high-byte and low-byte memory devices. The Output Enable (OE#) controls gating accessed data from the memory device outputs. Refer to Table 4. The Miniature Card automatically powers up in the read/reset state. In this case, a command sequence is not required to read data. Standard microprocessor read cycles will retrieve array data. This default state ensures that no spurious alteration of the memory content occurs during the power transition. Refer to the AC Read Characteristics and Waveforms for the specific timing parameters. Output Disable Data outputs from the card are disabled when OE# is at a logic-high level. Under this condition, outputs are in the high-impedance state. 10 AmMCL00XA PRELIMINARY Table 4. Function Read Mode Word Access Low Byte Access High Byte Access Write Mode Word Access Low Byte Access High Byte Access Standby Mode Standby H H X X High-Z High-Z L H L L L H L L L H H H High Byte Data High-Z High Byte Data Low Byte Data Low Byte Data High-Z L H L L L H H H H L L L High Byte Data High-Z High Byte Data Low Byte Data Low Byte Data High-Z CEH# Miniature Card Read/Write Modes CEL# WE# OE# D8–D15 D0–D7 Notes: 1. Unlisted access combinations are invalid and may return unexpected results. 2. X indicates a don’t care value. Erase Operations The AMD Flash Miniature Card is organized as an array of individual devices. Each Am29LV081 device contains sixteen 64 KByte sectors, for a total of 1 Mbyte of memory space per device. Flash technology allows any logical “1” data bit to be programmed to a logical “0”. The only way to reset bits to a logical “1” is to erase that entire memory sector or memory device. Once a memory sector or memory device is erased, any address location may be programmed. Two or more devices may be erased concurrently when additional ICC current is supplied to the card. However, erasing more than two devices concurrently is not typical in battery-powered applications, but may take place during procedures such as card testing. Erase operations can be performed in several ways: s Erase a single sector or multiple sectors in a device s Erase a sector pair s Erase multiple device pairs* s Erase the entire card* * This operation is only feasible in solutions capable of supplying more than the specified miniature card supply current requirement (150mA) per system. Each AMD Flash memory d evice pair c an accept a maximum of 120mA supply current. The common memory space data contents are altered in a similar manner to writing to individual Flash memory devices. An on-card address decoder activates the appropriate Flash device in the memory array. Each device internally latches address and data during write cycles. Refer to Table 4. Standby Mode The AMD flash devices are designed to accommodate low standby power consumption. In order to achieve standby mode, the CE# line must be deselected. In addition, while in the standby mode, data I/O pins remain in the high impedance state independent of the voltage level applied to the OE# input. See the DC Characteristics section for more details on Standby Modes. Deselecting CE# (CE# and RESET# = VCC ± 0.3 V) puts the device into the I CC3 s tandby mode. If the device is deselected during an Embedded Algorithm operation, it continues to draw active power (ICC2) prior to entering the standby mode, until the operation is complete. When the device is again selected (CE# = VIL), active operations occur in accordance with the AC timing specifications. Automatic Sleep Mode Advanced power management features such as the automatic sleep mode minimize Flash device energy consumption. This is extremely important in battery-powered applications. The AMD memory devices automatically enable the low-power, automatic sleep mode when addresses remain stable for 300 ns. Automatic sleep mode is independent of the CE#, WE#, and OE# control signals. Typical sleep mode current draw from each device is < 1 µA. Standard address access timings provide new data when addresses are AmMCL00XA 11 PRELIMINARY changed. While in sleep mode, output data is latched and always available to the system. ing them in the improper sequence will reset the device to the read mode. The byte-wide commands are defined in Tables 6 and 7; word-wide commands are defined in Table 5. Note that the Erase Suspend (B0h) and Erase Resume (30h) commands are valid only while the Sector Erase operation is in progress. Command Definitions Each memory device contains a command register, which is a latch that saves address, commands, and data information used by the state machine and memory array. The state machine is active when VCC is greater than VLKO (2.3 - 2.5 V). This is required for valid program and erase operations. When Write Enable (WE#) and appropriate CE# signals are at a logic-low level, and Output Enable (OE#) is at a logic-high, the command register is enabled for write operations. The falling edge of WE# or CE#, whichever occurs later, latches address information and the rising edge of WE# or CE#, whichever occurs first, latches data/command information. Commands are accomplished by writing non-specific address and specific data sequences into the command register of accessed Flash memory devices. Writing incorrect address and data values or writ- Autoselect Operation A host system or external card reader/writer can determine the on-card manufacturer and device I.D. codes. Codes are available after writing the 90h command to the command register of a memory device, as shown in Tables 5 through 7. When the autoselect command is issued to card address 00000h, the Miniature Card returns the manufacturer I.D. If the autoselect command is issued to card address 00001h, the Miniature Card provides the device I.D. To terminate the autoselect operation, the Read/Reset command sequence must be written to the same device. The Autoselect command operates only if the card is not write protected. 12 AmMCL00XA PRELIMINARY Table 5. Cycles Word Command Definitions Bus Cycles (Notes 2–9) Embedded Command Sequence (Note 1) Read Reset Autoselect Manufacturer ID (Note 4) Autoselect Device ID (Note 4) Word Write Device Erase Sector Erase Sector Erase Suspend (Note 7) Sector Erase Resume (Note 8) First Addr RA XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX Second Addr Data Third Addr Data Fourth Addr Data Fifth Addr Data Sixth Addr Data Data RW F0F0 AAAA AAAA AAAA AAAA AAAA B0B0 3030 1 1 4 4 4 6 6 1 1 XXXX XXXX XXXX XXXX XXXX 5555 5555 5555 5555 5555 XXXX XXXX XXXX XXXX XXXX 9090 9090 A0A0 8080 8080 XX00 XX01 PA XXXX XXXX 0101 3838 PW AAAA AAAA XXXX XXXX 5555 5555 XXXX SA 1010 3030 Legend: X = Don’t care RA = Address of the memory location to be read. RW = Data read from location RA during read operation. PA = Address of the memory location to be programmed. Addresses are latched on the falling edge of the WE# pulse. Notes: 1. Write protect must not be enabled for proper operation of all commands. No command required for reading array data, and can thus be done with write protect enabled. 2. During word addressing, CEL# = 0, CEH# = 0, and address is applied to Memory Device Pair 0 (S0 and S1). On 4 Mbyte cards, address for Memory Device Pair 1 = (Addr) + 200000h, and address is applied to Memory Device Pair 1 (S2 and S3). For host-to-card address bit connections, see Figure 2. 3. All values are in hexadecimal. 4. The last bus cycle in an autoselect command sequence is a read operation. 5. Word = high byte + low byte. 6. Address bits = X = Don’t Care for all commands except for Read Address (RA), Program Address (PA), and Sector Address (SA). 7. The Erase Suspend command is valid only during a sector erase operation. Refer to “Sector Erase Suspend”. 8. The Erase Resume command is valid only during the Erase Suspend mode. 9. See Table 4 for bus operations. PW = Data to be programmed at location PA. Data is latched on the rising edge of WE#. SA = Address of the sector to be erased. Refer to Table 8 for sector addresses. AmMCL00XA 13 PRELIMINARY Table 6. Cycles Even Byte Command Definitions Bus Cycles (Notes 2–8) Embedded Command Sequence (Note 1) Read Reset Autoselect Manufacturer ID (Note 4) Device ID (Note 4) Byte Write Device Erase Sector Erase Sector Erase Suspend (Note 6) Sector Erase Resume (Note 7) First Addr RA Data RD Second Addr Data Third Addr Data Fourth Addr Data Fifth Addr Data Sixth Addr Data 1 1 4 4 4 6 6 1 1 XXXX XXF0 XXXX XXAA XXXX XXXX XXAA XXXX XXXX XXAA XXXX XXXX XXAA XXXX XXXX XXAA XXXX XX55 XX55 XX55 XX55 XX55 XXXX XXXX XXXX XXXX XXXX XX90 XX90 XXA0 XX80 XX80 XX00 XX01 PA XX01 XX38 PD XX55 XX55 XXXX SA XX10 XX30 XXXX XXAA XXXX XXXX XXAA XXXX XXXX XXB0 XXXX XX30 Note for Table 6: During even (low) byte accesses, CEL# = 0, CEH# = 1. Address is applied to Memory Device 0 (S0). On 4 Mbyte cards, address for Memory Device 2 (S2) = (Addr) + 200000h. Table 7. Cycles Odd Byte Command Definitions Bus Cycles (Notes 2–8) Embedded Command Sequence (Note 1) Read Reset Autoselect Manufacturer ID (Note 4) Autoselect Device ID (Note 4) Byte Write Device Erase Sector Erase Sector Erase Suspend (Note 6) Sector Erase Resume (Note 7) First Addr RA Data RD Second Addr Data Third Addr Data Fourth Addr Data Fifth Addr Data Sixth Addr Data 1 1 4 4 4 6 6 1 1 XXXX XXF0 XXXX AAXX XXXX XXXX AAXX XXXX XXXX AAXX XXXX XXXX AAXX XXXX XXXX AAXX XXXX 55XX 55XX 55XX 55XX 55XX XXXX XXXX XXXX XXXX XXXX 90XX 90XX A0XX 80XX 80XX XX00 XX01 PA 01XX 38XX PDXX 55XX 55XX XXXX SA 10XX 30XX XXXX AAXX XXXX XXXX AAXX XXXX XXXX XXB0 XXXX XX30 Note for Table 7: During odd (high) byte accesses, CEL#= 1, CEH# = 0, and address is applied to Memory Device 1 (S1). On 4 Mbyte cards, address for Memory Device 3 (S3) = (Addr) + 200000h + 100000h. Legend for Tables 6 and 7: X = Don’t care RA = Address of the memory location to be read. RW = Data read from location RA during read operation. PA = Address of the memory location to be programmed. Addresses are latched on the falling edge of the WE# pulse. Notes for Tables 6 and 7: 1. Write protect must not be enabled for proper operation of all commands. No command required for reading array data, and can thus be done with write protect enabled. 2. For host-to-card address bit connections, see Figure 2. 3. All values are in hexadecimal. 4. The last bus cycle in an autoselect command sequence is a read operation. 5. Address bits = X = Don’t Care for all commands except for Read Address (RA), Program Address (PA), and Sector Address (SA). 6. The Erase Suspend command is valid only during a sector erase operation. Refer to “Sector Erase Suspend”. 7. The Erase Resume command is valid only during the Erase Suspend mode. 8. See Table 4 for bus operations. PW = Data to be programmed at location PA. Data is latched on the rising edge of WE#. SA = Address of the sector to be erased. Refer to Table 8 for sector addresses. 14 AmMCL00XA PRELIMINARY Table 8. Card Address Bits Sector 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 A19 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 A18 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 A17 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 A16 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 Memory Sector Addresses Device 0 and/or 1 (Note 1) Card Address Range 00000h–0FFFFh 10000h–1FFFFh 20000h–2FFFFh 30000h–3FFFFh 40000h–4FFFFh 50000h–5FFFFh 60000h–6FFFFh 70000h–7FFFFh 80000h–8FFFFh 90000h–9FFFFh A0000h–AFFFFh B0000h–BFFFFh C0000h–CFFFFh D0000h–DFFFFh E0000h–EFFFFh F0000h–FFFFFh Device 2 and/or 3 (Note 1) Card Address Range 100000h–10FFFFh 110000h–11FFFFh 120000h–12FFFFh 130000h–13FFFFh 140000h–14FFFFh 150000h–15FFFFh 160000h–16FFFFh 170000h–17FFFFh 180000h–18FFFFh 190000h–19FFFFh 1A0000h–1AFFFFh 1B0000h–1BFFFFh 1C0000h–1CFFFFh 1D0000h–1DFFFFh 1E0000h–1EFFFFh 1F0000h–1FFFFFh Notes: 1. For word addressing, devices 0 and 1 (S0 and S1) together form Memory Device Pair 0; devices 2 and 3 (S2 and S3) form Memory Device Pair 1. Refer to the block diagram for device connections. 2. Card address bits range from A0 to A19. Host address bits range from A0 to A20. Host address bit A0 is used for controlling the CEL# and CEH# inputs to the card. Refer to Figure 2 for host-to-card address bit connections. AmMCL00XA 15 PRELIMINARY AMD FLASH MEMORY PROGRAM AND ERASE OPERATIONS To simplify program and erase operations, AMD Flash Memory devices include Embedded Algorithms (Embedded Erase Algorithm and Embedded Program Algorithm) that allow the host to simply issue a command, after which it is free to perform other tasks. The host then only needs to monitor appropriate status bits to determine when the operation is complete. requires that a valid address input to the device is supplied by the system at this particular instant of time. Otherwise, the system will never read a “1” on D7 (D15 on the odd byte). A system designer has the following choices to implement the Embedded Erase algorithm: 1. The host may keep the sector address (within any of the sectors being erased) valid during the entire Embedded Erase operation. 2. Once the system executes the Embedded Erase command sequence, the host may remove the address from the device and perform other tasks. The host is required to keep track of the valid sector address by loading it into a temporary register. When the host comes back to Data Poll the device, it must reassert the same address. 3. The host may monitor BUSY# (RY/BY#) to determine the status of the Embedded Algorithm in progress. A “0” indicates that the device is busy; a “1” indicates that the algorithm is complete. Sector Erase Sector erase is a six bus cycle operation. There are two “unlock” write cycles. These are followed by writing the “set-up” command. Two more “unlock” write cycles are then followed by the sector erase command. The sector address (any address location within the desired sector) is latched on the falling edge of WE# (or CE#), whichever occurs later, while the data is latched on the rising edge of WE# (or CE#) pulse, whichever occurs first. A time-out of 80 µs from the rising edge of the last sector erase command will initiate the sector erase command. Multiple sectors can be specified for erase by writing the six bus cycle operation as described above and then following it by additional writes of the Sector Erase command to addresses of other sectors to be erased. The time between Sector Erase command writes must be less than 80 µs, otherwise that command will not be accepted. It is recommended that processor interrupts be disabled during this time to guarantee this condition. The interrupts can be re-enabled after the last Sector Erase command is written. A time-out of 80 µs from the rising edge of the last WE# (or CE#) will initiate the execution of the Sector Erase command(s). If another falling edge of the WE# (or CE#) occurs within the 80 µs time-out window, the timer is reset. During the 80 µs window, any command other than Sector Erase or Erase Suspend written to the device will reset the device back to Read mode. Once the 80 µs window has timed out, only the Erase suspend command is recognized. Note that although the Reset command is not recognized in the Erase Suspend mode, the device is available for read or program operations in sectors that are not erase suspended. The Erase Suspended and Erase Resume commands may be written as often as required during a sector erase operation. Hence, once erase has begun, it must ultimately complete unless Embedded Erase Algorithm When erasing a sector or device, the Embedded Erase algorithm does not require the host to first entirely pr e- p r o g r am th e d ev i c e . Up o n e x ec u t in g t he Embedded Erase command sequence, the addressed memory sector or memory device automatically writes and verifies the entire memory device or memory sector for an all “0” data pattern. The system is not required to provide any controls or timing during these operations. When the memory sector or memory device is automatically verified to contain an all “0” pattern, a self-timed chip erase-and-verify begins. The erase and verify operations are complete when the data on D7 (D15 on the odd byte) of the memory sector or memory device is “1” (see Write Operation Status section), at which time the device returns to the read mode. The system is not required to provide any control or timing during these operations. If a Reset command is issued while the erase operation is in progress, the erase operation will stop, and the data in that device will be undefined. In that case, restart the erase on that sector and allow it to complete. 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 Embedded Erase command sequence is a command only operation that stages the memory sector or memory device for automatic electrical erasure of all bytes in the array. The automatic erase begins on the rising edge of the WE# and terminates when the data on D7 (D15 on the odd byte) of the memory sector or memory device is “1” (see Write Operation Status section) at which time the device returns to the Read mode. Please note that for the memory device or memory sector erase operation, Data Polling may be performed at any address in that device or sector. Figure 4 and Table 9 illustrate the Embedded Erase Algorithm, a typical command string and bus operations. As described earlier, once the memory sector in a device or memory device completes the Embedded Erase operation, it returns to the Read mode and addresses are no longer latched. Therefore, the device 16 AmMCL00XA PRELIMINARY Hardware Reset is initiated. Loading the sector erase registers may be done in any sequence and with any number of sectors (0 to 15). A Reset command issued after the device has begun execution stops the erase operation, but the data in the sector will be undefined. In that case, restart the erase on that sector and allow it to complete. The automatic sector erase begins after the 80 µs time out from the rising edge of the WE# (or CE#) pulse for the last sector erase command pulse and terminates when the data on D7 is “1” (see Write Operation Status section) at which time the device returns to read mode. Data Polling must be performed at an address within any of the sectors being erased. If DATA Polling or the Toggle Bit indicates the device has been written with a valid Sector Erase command, D3 may be used to determine if the sector erase timer window is still open. If D3 is high (‘1’), the internally controlled erase cycle has begun; attempts to write subsequent commands to the device will be ignored until the erase operation is completed as indicated by the DATA Polling or Toggle Bit. If D3 is low (‘0’), the device will accept additional sector erase commands. To be certain the command has been accepted, the software should check the status of D3 following each Sector Erase command. If D3 was high on the second status check, the command may not have been accepted. It is recommended that the user guarantee the time between sector erase command writes be less than 80 µs by disabling the processor interrupts just for the duration of the Sector Erase (30H) commands. This approach will ensure that sequential sector erase command writes will be written to the device while the sector erase timer window is still open. Figure 4 illustrates the Embedded Erase Algorithm using typical command strings and bus operations. Table 9. Bus Operation Standby Write Embedded Erase command sequence Start Write Embedded Erase Command Sequence (See Tables 5–7) Data Poll from Device or wait for BUSY# (RY/BY#) Erasure Complete 21138E-5 Figure 4. Embedded Erase Algorithm Note: The latest release of the software drivers for AMD Miniature Cards and devices may be downloaded from the AMD web site at http://www.amd.com. Embedded Program Algorithm The Embedded Program setup is a four bus cycle operation that stages the addressed memory location or memory device for automatic programming. Once the Embedded Program setup operation is performed, the next WE# pulse causes a transition to an active programming operation. Addresses are internally latched on the falling edge of the WE# (or CE#) pulse. Data is internally latched on the rising edge of the WE# pulse. The rising edge of WE# also begins the programming operation. The system is not required to provide further control or timing. The device will automatically provide an adequate internally generated write pulse and verify margin. The automatic programming operation is completed when the data on D7 of the addressed memory sector or memory device is equivalent to data written to this bit (see Write Operation Status section) at which time the device returns to the Read mode (no write verify command is required). Addresses are latched on the falling edge of WE# (or CE#) during the Embedded Program command execution and hence the system is not required to keep the addresses stable during the entire Programming operation. However, once the device completes the Embedded Program operation, it returns to the Read mode and addresses are no longer latched. Since a verify valid data must occur on D7, at this particular instant, the system is required to supply a valid address input to the device. A system designer has three choices to implement the Embedded Programming algorithm: Embedded Erase Algorithm Command Comments Wait for VCC ramp 6 bus cycle operation Data Poll or check BUSY# (RY/BY#) to verify erasure Read AmMCL00XA 17 PRELIMINARY 1. The system (CPU) keeps the address valid during the entire Embedded Programming operation, or 2. Once the system executes the Embedded Programming command sequence, the CPU takes away the address from the device and becomes free to do other tasks. In this case, the CPU is required to keep track of the valid address by loading it into a temporary register. When the CPU comes back for performing Data Polling, it should reassert the same address. 3. The host may monitor BUSY# (RY/BY#) to determine the status of the Embedded Algorithm in progress. A “0” indicates that the device is busy; a “1” indicates that the algorithm is complete. However, since the Embedded Programming operation takes only 9 µs typically, it may be easier for the CPU to keep the address stable during the entire Embedded Programming operation instead of reasserting the valid address during Data Polling. Any commands written to the device during this period will be ignored. Figure 5 and Table 10 illustrate the Embedded Program Algorithm, a typical command string, and bus operation. Table 10. Embedded Program Algorithm Bus Operation Standby Write Write Command Comments Wait for VCC ramp Embedded Program 3 bus cycle operation command sequence Program Address/Data 1 bus cycle operation Data Poll or check BUSY# (RY/BY#) to verify program Start Write Embedded Write Command Sequence per Tables 5–7 Data Poll Device or wait for BUSY# (RY/BY#) Verify Data Y N Last Address Y N Increment Address Completed 21138E-6 Figure 5. Embedded Program Algorithm Reset Command The device will automatically power up in the read/reset state. In this case, a command sequence is not required to read data. Standard microprocessor cycles will retrieve array data. This default value ensures that no spurious alteration of the memory content occurs during the power transition. Refer to the AC Characteristics section for the specific timing parameters. The reset operation is initiated by writing the read/reset command sequence into the command register. Microprocessor read cycles retrieve array data from the memory. The device remains enabled for reads until the command register contents are altered. Read Sector Erase Suspend The Erase Suspend command allows the user to interrupt a Sector Erase operation and then perform data read or programs in a sector not being erased. This command is applicable only during the Sector Erase operation, which includes the time-out period for Sector Erase. The Erase Suspend command will be ignored if written during the execution of the Chip Erase operation or Embedded Program Algorithm (but will reset the chip if written improperly during the command sequences.) Writing the Erase Suspend command during the Sector Erase time-out results in immediate termination of the time-out period and suspension of the erase operation. Once in Erase Suspend, the device is avail- 18 AmMCL00XA PRELIMINARY able for read (note that in the Erase Suspend mode, the Reset/Read command is not required for read operations and is ignored) or program operations in sectors not being erased. Any other command written during the Erase Suspend mode will be ignored, except for the Erase Resume command. Writing the Erase Resume command resumes the sector erase operation. The addresses are “don’t cares” when writing the Erase Suspend or Erase Resume command. When the Erase Suspend command is written during a Sector Erase operation, the chip will take between 0.1 µs and 20 µs to actually suspend the operation and go into erase suspended read mode (pseudo-read mode), at which time the user can read or program from a sector that is not erase suspended. Reading data in this mode is the same as reading from the standard read mode, except that the data must be read from sectors that have not been erase suspended. Successively reading from the erase-suspended sector while the device is in the erase-suspend-read mode will cause D2 to toggle. Polling D2 on successive reads from a given sector provides the system the ability to determine if a sector is in Erase Suspend. After entering the erase-suspend-read mode, the user can program the device by writing the appropriate command sequence for Byte Program. This program mode is known as the erase suspend-program mode. Again, programming in this mode is the same as programming in the regular Byte Program mode, except that the data must be programmed to sectors that are not erase suspended. Successively reading from the erase suspended sector while the device is in the erase suspend-program mode will cause D2 to toggle. Completion of the erase suspend operation can be determined two ways: s Checking the status of the toggle bit D2 s Checking the status of the RY/BY# pin To resume the operation of Sector Erase, the Resume command (30H) should be written. Any further writes of the Resume command at this point will be ignored. However, another Erase Suspend command can be written after the device has resumed sector erase operations. ded Program Algorithm, an attempt to read the device will produce the true data last written to D7. Note that just at the instant when D7 switches to true data, the other bits, D6–D0, may not yet be true data. However, they will all be true data on the next read from the device. Please note that Data Polling (D7) may give an inaccurate result when an attempt is made to write to a protected sector. During an Embedded Erase Algorithm, an attempt to read the device will produce a ‘0’ at the D7 output. Upon completion of the Embedded Erase Algorithm, an attempt to read the device will produce a ‘1’ at D7. START DQ7 = Data? Yes No No DQ5 = 1? Yes Yes DQ7 = Data? No FAIL PASS 21138E-7 Note: D7 is rechecked even if D5 = 1 because D7 may change simultaneously with D5. Write Operation Status Table 11 shows the status bit states for device program and erase operations. Data Polling—D7 (D15 on Odd Byte) The Miniature card features DATA Polling as a method to indicate to the host system that the embedded algorithms are in progress or completed. During the Embedded Program Algorithm, an attempt to read the device will produce the compliment of the data last written to D7. Upon completion of the Embed- Figure 6. Data Polling Algorithm AmMCL00XA 19 PRELIMINARY Table 11. Hardware Sequence Flags Status Byte Program in Embedded Program Algorithm Embedded Erase Algorithm Erase Suspend Read (Erase Suspended Sector) Erase Suspended Mode Erase Suspend Read (Non-Erase Suspended Sector) Erase Suspend Program (Non-Erase Suspended Sector) Byte Program in Embedded Program Algorithm Exceeded Time Limits Program/Erase in Embedded Erase Algorithm Erase Suspended Mode Erase Suspend Program (Non-Erase Suspended Sector) D7 D7 0 1 Data D7 D7 0 D7 D6 Toggle Toggle 1 Data Toggle (Note 2) Toggle Toggle Toggle D5 0 0 0 Data 0 1 1 1 D3 0 1 0 Data 1 0 1 1 D2 1 Toggle Toggle (Note 1) Data 1 (Note 3) 1 N/A N/A In Progress Notes: 1. Performing successive read operations from the erase-suspended sector will cause D2 to toggle. 2. Performing successive read operations from any address will cause D6 to toggle. 3. Reading the byte address being programmed while in the erase-suspend program mode will indicate logic “1” at the D2 bit. However, successive reads from the erase-suspended sector will cause D2 to toggle. BUSY# (RY/BY#—Ready/Busy) The BUSY# signal indicates to the host the status of operations within the Miniature Card. The BUSY# signal is tied to the components’ RY/BY# pins. The RY/BY# signal from AMD Flash devices in the Miniature Card indicate that the Embedded Algorithms are either in progress or have been completed. If the output is low, the device is busy with either a program or erase operation. If the output is high, the device is ready to accept any read/write or erase operation. When the RY/BY# pin is low, the device will not accept any additional program or erase commands with the exception of the Erase Suspend command. If a Flash device is placed in an Erase Suspend mode, the RY/BY# output will be high. Refer to the section “Sector Erase Suspend” for more information. WORD-WIDE PROGRAMMING T he Word-Wide Programming sequence will be as usual per Table 5. The Program word command is A0A0H. Each byte is independently programmed. For example, if the high byte of the word indicates the successful completion of programming via one of its write status bits such as D15, software polling should continue to monitor the low byte for write completion and data verification, or vice versa. During the Embedded Programming operations the device executes programming pulses in 9 µs increments. WORD-WIDE SECTOR ERASING The Word-Wide Sector Erasing of a memory device pair is similar to word-wide programming. The erase word command is a six-bus-cycle command sequence (see Table 5). Each sector is independently erased and verified. Word-wide erasure reduces total erase time when compared to byte erasure. Each Flash memory device in the card may erase at different rates. Therefore, each device (byte) must be verified separately. 20 AmMCL00XA PRELIMINARY ABSOLUTE MAXIMUM RATINGS Storage Temperature . . . . . . . . . . . . . –40°C to +90°C Ambient Temperature with Power Applied . . . . . . . . . . . . . . –40°C to +85°C Voltage at All Pins (Note 1) . . . . –0.5 V to VCC+0.5 V VCC (Note 1) . . . . . . . . . . . . . . . . . . . . –0.5 V to 3.6 V Output Short Circuit Current (Note 2) . . . . . . 200 mA 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 output and I/O pins is VCC + 0.5 V. During voltage transitions, outputs may overshoot to VCC + 2.0 V for periods up to 20ns. 2. No more than one output shorted at a time. Duration of the short circuit should not be greater than one second. Conditions equal VOUT = 0.5 V or 3.6 V, VCC = VCCmax. These values are chosen to avoid test problems caused by tester ground degradation. This parameter is sampled and not 100% tested, but guaranteed by characterization. 3. 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. OPERATING RANGES Commercial Devices Case Temperature (TC). . . . . . . . . . . . . .0°C to +70°C Industrial (I) Devices Case Temperature (TC). . . . . . . . . . . .–40°C to +85°C VCC Supply Voltages AmMCL00XAWP-150 . . . . . . . . . . . . +2.7 V to +3.6 V Operating ranges define those limits between which the functionality of the device is guaranteed. AmMCL00XA 21 PRELIMINARY DC CHARACTERISTICS Parameter Symbol ILI ILO ICCS Parameter Description Input Leakage Current Output Leakage Current VCC Standby Current Test Conditions VIN = VSS to VCC, VCC = VCC max VIN = VSS to VCC, VCC = VCC max CEL#, CEH#, RESET# = VCC ± 0.3 2 Mbyte V 4 Mbyte V = 3.6V; V = V or V CC IN SS CC Min Max ±5 ±5 30 40 40 60 Unit µA µA µA µA mA mA V V V V ICC VIL VIH VOL VOH VLKO VCC Supply Current, word mode (Note 2) Input Low Voltage Input High Voltage Output Low Voltage Output High Voltage Low VCC Lock-Out Voltage RESET# = VIH; CEL# and CEH# = VIL Read Write –0.5 0.7 VCC 0.8 VCC + 0.5 0.45 IOUT = 5.8 mA IOUT = –2.0 mA 0.85 VCC 2.3 2.5 V Notes: 1. VCC = 2.7 V to 3.6 V. 2. Supply current is a max RMS value. Read frequency = 5 MHz. CONNECTOR DC SPECIFICATIONS Parameter Interface Signal Resistance (Note 2) Interface Signal Current (Notes 1, 2) Power/Insertion Signal Resistance Power/Insertion Signal Current (Note 1) 500 125 0.060 Min Max 2.0 Units Ω mA Ω mA Notes: 1. This current is a minimum that the connector should withstand, and a maximum that the host should provide. 2. On the host, these specifications must be met for one conducting channel on elastomeric connectors. CARD AND PAD CAPACITANCE Parameter Symbol CCARD CHOST CI/O Parameter Description Card Input Capacitance System Load Capacitance I/O Capacitance D0-D15 Test Conditions Max 40 120 40 Unit pF pF pF Notes: 1. Sampled, not 100% tested. 2. Test conditions TA = 25°C, f = 1.0 MHz. 22 AmMCL00XA PRELIMINARY AC CHARACTERISTICS Read-only Operations Parameter Symbol JEDEC tAVAV tELQV tAVQV tGLQV tELQX tEHQZ tGLQX tGHQZ tAXQX Standard tRC tCE tACC tOE tLZ tDF tOLZ tDF tOH tReady Parameter Description Read Cycle Time Chip Enable Access Time Address Access Time Output Enable Access Time Chip Enable to Output in Low-Z Chip Disable to Output in High-Z Output Enable to Output in Low-Z Output Disable to Output in High-Z Output Hold from First of Address, CE#, or OE# Change RESET# Pin Low to Read Mode Min Max Max Max Min Max Min Max Min Max -150 150 Unit ns ns ns ns ns ns ns ns ns µs 150 150 50 5 30 5 30 5 20 AmMCL00XA 23 PRELIMINARY AC CHARACTERISTICS Write Operations (Erase/Program) Parameter Symbols JEDEC tAVAV tWLWH tELGL tELWL tAVGL tAVWL tDVWH tWHDX tWHAX tWHEH tRP tBUSY tWHWH1 Standard tWC Parameter Description Write Cycle Time WE# pulse width CE# setup time to WE# or OE# active Address setup time to WE# or OE# active Data setup time to WE# inactive Data hold time from WE# inactive Address hold time from WE# inactive CE# hold time from WE# inactive RESET# Pulse Width Program/Erase Valid to RY/BY# Delay Programming Operation Max Typ tWHWH2 Sector Erase Operation Max 15 s 300 1.5 Min Min Min Min Min Min Min Min Min Min Typ -150 150 50 0 0 50 0 0 0 500 90 9 µs Unit ns ns ns ns ns ns ns ns ns ns 24 AmMCL00XA PRELIMINARY KEY TO SWITCHING WAVEFORMS WAVEFORM INPUTS Must be Steady May Change from H to L May Change from L to H Don’t Care, Any Change Permitted Does Not Apply OUTPUTS Will be Steady Will be Changing from H to L Will be Changing from L to H Changing, State Unknown Center Line is HighImpedance “Off” State KS000010 SWITCHING WAVEFORMS tAVAV tAVQV tAVGL tAXQX A0–A25 tELGL tELQV tELQNZ CEL#/CEH# tGLQV tGLQNZ OE# tGHQZ tGHQX tEHQX D0–D15 Valid Data 21138E-8 Figure 7. AC Waveforms for Read Operations AmMCL00XA 25 PRELIMINARY SWITCHING WAVEFORMS tAVAV tAVWL A0–A25 tWHAX tELWL CEL#/CEH# tWLWH tDVWH tWHEH tWHDX WE# D0–D15 Valid Data 21138E-9 Figure 8. AC Waveforms for Write Operations CE# tCH tDF tOE OE# tOEH WE# tCE * D7 tWHWH1 or tWHWH2 D0–D6 D0–D6=Invalid D7# tOH D7= Valid Data High Z D0–D7 Valid Data *D7=Valid Data (The device has completed the Embedded operation). 21138E-10 Figure 9. AC Waveforms for Data Polling During Embedded Algorithm Operations 26 AmMCL00XA PRELIMINARY SWITCHING WAVEFORMS CE# The rising edge of the last WE# signal WE# Entire programming or erase operations RY/BY# tBUSY 21138E-11 Figure 10. RY/BY# Timing Diagram During Program/Erase Operations RESET# tRP tReady 21138E-12 Figure 11. RESET# Timing Diagram AmMCL00XA 27 PRELIMINARY AC CHARACTERISTICS-ALTERNATE CE# CONTROLLED WRITES Write/Erase/Program Operations Parameter Symbols JEDEC tAVAV tAVEL tELAX tDVEH tEHDX tGLDV tGHEL tWLEL tEHWH tELEH tEHEL tEHEH3 tWS tWH tCP tCPH Standard tWC tAS tAH tDS tDH tOEH Parameter Description Write Cycle Time Address Setup Time Address Hold Time Data Setup Time Data Hold Time Output Enable Hold Time for Embedded Algorithm Read Recovery Time before Write WE# Setup Time before CE# WE# Hold Time CE# Pulse Width CE# Pulse Width HIGH (Note 3) Embedded Programming Operation (Notes 3,4) Max Embedded Erase Operation for each 64K byte Memory Sector (Notes 1, 2) VCC Setup Time to Write Enable LOW Typ Max Min 300 1.5 15 50 s µs Min Min Min Min Min Min Min Min Min Min Min Typ -150 150 10 50 50 0 10 0 0 0 50 20 9 µs Unit ns ns ns ns ns ns µs ns ns ns ns tEHEH4 tVCS Notes: 1. Rise/fall time ≤10 ns. 2. Maximum specification not needed due to the internal stop timer that will stop any erase or write operation that exceed the device specification. 3. Card Enable Controlled Programming: Flash Programming is controlled by the valid combination of the Card Enable (CE1#, CE2#) and Write Enable (WE#) signals. For systems that use the Card Enable signal(s) to define the write pulse width, all setup, hold, and inactive write enable timing should be measured relative to the Card Enable signal(s). 4. Under worst case condition of 90° C, Vcc = 2.7 V, 100,000 cycles. Excludes system level overhead, the time required to execute the four bus cycle command necessary to program each byte. 28 AmMCL00XA PRELIMINARY tWC Addresses XXXXh tAS PA tAH Data# Polling PA WE# tWH OE# tGHEL tCP CE# tWS tCPH tDS tDH Data A0h PD DQ7# DOUT tEHEH3_or_4 VCC tVCS Notes: 1. PA is address of the memory location to be programmed. 2. PD is data to be programmed at byte address. 3. D7 is the complement of the data written to the device. 4. DOUT is the data written to the device. 5. Figure indicates last two bus cycles of four bus cycle sequence. 6. These waveforms are for the x16 mode. 21138E-13 Figure 12. Alternate CE# Controlled Write Operation Timings The AIS supports up to four different memory technologies on a card. Some of the information areas are repeated in the memory map in order to specify different technologies (see Table 12). The Technology Count field in the Identification Data section defines the number of different technologies on a card. The first memory technology is defined in the AIS memory map from address 40H through 7FH. The second memory technology is defined from 80H through BFH. The third memory technology is defined from C0H to DFH. The fourth memory technology is defined from E0H to FFH. The AIS is stored as bytes within the 16-bit Miniature Card data word. The even byte D0–D7 stores the AIS data, and the odd byte D8–D15 is reserved by the card manufacturer for manufacturing information. AIS MEMORY MAP The AIS (Attribute Information Structure) is an area of memory used for storing information about the configuration of the Miniature Card. The AIS is recommended to be stored in the first sector of the first device of the Flash array. As this area is not explicitly protected, the AIS information must be reloaded onto the card in the event that the information is erased. The AIS has five unique information areas: 1. Identification Data: This data includes Manufacturer information (Manufacturer and card name). 2. Compatibility Data: This data specifies basic information about the card (memory size, access time, memory type, power, etc.) 3. Burst Data (not applicable) 4. DRAM Data (not applicable) 5. Reserved Data: This data area is reserved for future use. AmMCL00XA 29 PRELIMINARY Table 12. Miniature Card AIS Memory Assignments Card Address 00h–0Fh 10h–1Fh 20h–2Fh 30h–3Fh 40h–4Fh 50h–5Fh 60h–6Fh 70h–7Fh 80h–8Fh 90h–9Fh A0h–AFh B0h–BFh C0h–CFh D0h–DFh E0h–EFh F0h–FFh Section PC Card Compatibility Area* Identification Data Identifies Card Type Identification Data Identifies Card Type Identification Data Identifies Card Type Compatibility Data (Area 1) Burst Data (not applicable) DRAM Data (not applicable) Reserved for future use Compatibility Data (not applicable) Burst Data (not applicable) DRAM Data (not applicable) Reserved for future use Compatibility Data (not applicable) Reserved for future use Compatibility Data (not applicable) Reserved for future use (Memory Technology #4) (Memory Technology #3) (Memory Technology #2) Memory Technology #1 Description Reserved for PC Card Tuples * For more information on PC Card Compatibility refer to table 13 or the Miniature Card PC Compatibility Guide. Note: “Not applicable” indicates the address space does not apply to AMD Flash Miniature Cards, but is defined by MCIF. 30 AmMCL00XA PRELIMINARY Table 13. PC Card Compatibility Memory Assignments Address 00h 01h 02h 03h 04h 05h 06h 07h 08h 09h 0Ah 0Bh 0Ch 0Dh 0Eh 0Fh Values 01h 03h 53 2 MB = 7C, 4 MB = FC FF 1Ch 03h 53h 2MB = 7C; 4MB = FC FFh 00h 00h 00h 00h 80h F0h Description TPL_CODE CISTPL_DEVICE TPL_LINK Device ID Device Size End of CISTPL_DEVICE TPL_CODE CISTPL_DEVICE_OC TPL_LINK Device ID Device Size End of CISTPL_DEVICE_OC CISTPL_NULL CISTPL_NULL CISTPL_NULL CISTPL_NULL TPL_CODE CISTPL_MINI TPL_LINK AmMCL00XA 31 PRELIMINARY Identification Data The identification data provides basic identification information about the card. This data section is required on all cards. Table 14 shows the Identification Data for AMD’s 3 volt-only Miniature cards. Compatibility Data The compatibility data provides basic compatibility across all cards. This data section is required on all cards. The addresses in parentheses are specified for cards with more than one memory technology on the card. Table 15 shows the compatibility data for AMD 3-volt only Miniature Cards Table 14. Card Address 10h 11h Value 99h 11h AMD Identification Data Description Miniature Card Identifier: Fixed value for a host to identify an inserted Miniature Card Level of Compliance: Defines the level of AIS supported. The Miniature Cards described in this document are rev 1.1 compliant. AIS Checksum: The modulo-256 sum of all even bytes from 10h–FFh. A valid checksum sums to 00H (2’s complement). 9 2 Mbyte card: 88h + 78h = 00h 4 Mbyte card: 8Ah + 76h = 00h Manufacturer Name: 13h–26h. String of ASCII characters at addresses 13H to 26H to identify the manufacturer of the Miniature Card. ASCII character “A” ASCII character “M” ASCII character “D” ASCII character - SPACE ASCII character - “I” ASCII character - “N” ASCII character - “C” ASCII character - NULL ASCII character - NULL Unused space in manufacturer name field Card Name: (addresses 27h–3Ah). String of ASCII characters to identify the card name. ASCII character “3” ASCII character “V” ASCII character “M” ASCII character “C” ASCII character - SPACE ASCII character “S” ASCII character “e” ASCII character “r” ASCII character “i” ASCII character “e” 12h 78h or 76h 13h 41h 14h 15h 16h 17h 18h 19h 1Ah 1Bh 1Ch–26h 27h 28h 29h 2Ah 2Bh 2Ch 2Dh 2Eh 2Fh 30h 4Dh 44h 20h 49h 4Eh 43h 00h 00h 00h 33h 56h 4Dh 43h 20h 53h 65h 72h 69h 65h 32 AmMCL00XA PRELIMINARY Table 14. Card Address 31h 32h 33h–3Ah 3Bh 3Ch–3Fh Value 73h 00h 00h 01h 00h ASCII character “s” ASCII character - NULL Unused space in card name field Technology Count: Defines the number of different memory technologies on the Miniature Card. Technology count set to 1 Reserved space set to 00h; for future use AMD Identification Data (Continued) Description . Table 15. AMD Compatibility Data Card Address 40h 41h 42h 43h 44h 45h 46h 47h 48h 49h 4Ah 4Bh–4Fh, 8Ch–8Fh, CCh–CFh, ECh–EFh 80h–8Bh, C0–CBh, E0h–EBh 100h 101h 102h 103h 104h 105h 106h 107h 108h 109h 10Ah 10Bh Value 00h 01h 38h 01h or 03h 00h 0Fh 00h 00h 24h 00h 00h 00h 00h 18h 02h 01h 38h 1Eh 06h 02h 01h 01h 01h 01h 01h Description Defines the type of memory technology; Flash = 000 Binary Device JEDEC Manufacturer ID Device JEDEC Component ID: Am29LV081 = 38h Memory array size: 02 = 2 Mbyte, 04 = 4 Mbyte N/A 3.3V access time: 150 ns N/A N/A Typical read/write current at 3.3V: 20 mA read, 40 mA write (word mode) N/A Typical card standby current: 10 µA for 2 Mbyte, 40 µA for 4 Mbyte Reserved for future use These addresses are designated for other memory technologies, which are not used in AMD Flash Miniature Cards. TPL_CODE CISTPL_JEDEC_C TPL_LINK Manufacturer ID Device ID TPL_CODE CISTPL_DEVICEGEO TPL_LINK DGTPL_BUS: Bus Width DGTPL_EBS:11h = 64K Byte Erase Block size DGTPL_RBS: Read Byte Size DGTPL_WBS: Write Byte Size DGTPL_PART: Number of partition FL DEVICE INTERLEAVE: No interleave. Note: All reserved bytes must be set to 00h. All reserved fields (bits) within bytes must be set to 0Bh. All unused fields must be set to 00h. AmMCL00XA 33 PRELIMINARY PHYSICAL DIMENSIONS Top View 33.00 mm 1.299 in. .118 in. 3.212 mm .118 in. 3.00 mm .217 in. 5.50 mm .118 in. 3.00 mm center line .284 in. 7.21 mm .161 in. .189 in. 4.09 mm 4.81 mm 38.00 mm 1.496 in. .217 in. 5.50 mm .118 in. 3.00 mm 34 AmMCL00XA PRELIMINARY PHYSICAL DIMENSIONS Bottom View 0.600 0.245 Write Protect Switch Location Right Side View 0.245 Write Protect Switch Location AmMCL00XA 35 PRELIMINARY REVISION HISTORY FOR AMMCL00XA Distinctive Characteristics Added industrial temperature bullet. Revised low power consumption specifications. Deleted “Small Form Factor” bullets. General Description Revised text to indicate that the Miniature Card specification will be defined by PCMCIA. Deleted references to the elastomeric connector. Table 2, AMD Flash Miniature Cards and Flash Devices Added WP as part of required base part number. Miniature Card Pad Assignments BUSY#: Revised to indicate that the Miniature Card cannot accept most operations when BUSY# is low. CD#: Deleted last sentence. Ordering Information Added Industrial temperature range. Deleted NP option from part number. Added WP as part of required base part number. Figure 2, Host/Card Address Assignments Labeled host bus in drawing. Deleted NC callouts in drawing. Tables 5–9, Command Definitions Revised for easier reference: removed “H” designators from table (now indicated in notes), removed 4-cycle Reset/Read command, separated Read and Reset commands, moved RA, RW, RD, PA, PW, PD, X, SA definitions to legend. Moved Erase Suspend and Erase Resume definitions from table to notes. Operating Ranges Added industrial temperature range. AC Characteristics, Write Operations Deleted tELQV, tAVQV, tGLQV, tELQX, tEHQZ, tGLQX, tGHQZ, tAXQX, tWHGL, tGLQNZ Embedded Erase Algorithm Removed last paragraph. Absolute Maximum Ratings Revised storage and ambient temperature ratings. Operating Ranges Added industrial temperature range. DC Characteristics Revised ICC specifications. Added frequency specification to Note 2. AC Characteristics, Write (Erase/Program) Operations Deleted tELQV, tAVQV, tGLQV, tELQX, tEHOZ, tGLOX, tGHQZ, tAXQX, tWHGL, tGLQNZ. Table 19, AMD Compatibility Data Added two tuples of data to list, covering addresses 100h–10Bh. Trademarks Copyright © 1997 Advanced Micro Devices, Inc. All rights reserved. AMD, the AMD logo, and combinations thereof are 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. 36 AmMCL00XA