EPROM, FLASH EPROM, EEPROM, SERIAL EEPROM, NVRAM, BIPOLAR PROM and MICROCONTROLLER Information

How is the EPROM organized inside?
The EPROM, like any memory chip, is designed to store individual bits of information. The bit (short for Binary Digit) is the smallest numeric element used in a computer and can have the value of 0 or 1. This is called binary because the bit can only have two states. Groups of bits are called by different terms. Sixteen (16) bits together is called a WORD. Eight (8) bits together is called a BYTE. Four (4) bits together is called a NIBBLE. Obviously a single bit can only represent two states, therefore it will take multiple bits used together to represent items such as letters and numbers. The most common grouping of bits is the byte. Since there are 8 bits and each can be a 1 or 0, there are 256 possible combinations of 1's and 0's which can be arranged from the 8 bit grouping.

The typical EPROM is organized as a sequence of bytes (8 bit groups). Each byte resides at a specific location in the EPROM called an ADDRESS. The first address in the EPROM is address 0. The last address in the EPROM is determined by the EPROM storage capacity. The EPROM in the photo is a 27C010. It has a storage capacity of 128K (1FFFF hex) bytes. When a CPU accesses the EPROM, it provides an ADDRESS to the chip and then reads the BYTE from that address. If the BYTE is an instruction, the CPU decodes the 1's and 0's in the byte and performs the operation the instruction specifies. This is how the EPROM stores a program. The CPU sequentially addresses each byte in the EPROM to get the instructions which allow the computer to do its job. The illustration above shows eight addresses and corresponding data. The data values are shown in BINARY (left) and HEX (right). HEX is a more compact form of presenting binary numbers. HEX numbers are described later in this page.

**EPROM ORGANIZATION**
ADDRESS 0	01001000 (48)
ADDRESS 1	00000010 (02)
ADDRESS 2	11000110 (C6)
ADDRESS 3	11100000 (E0)
ADDRESS 4	11110001 (F1)
ADDRESS 5	00011111 (1F)
ADDRESS 6	10011001 (99)
ADDRESS 7	00000000 (00)

What does the part number mean?
The EPROM part number normally indicates the total number of bits of storage in thousands. For example, a 27256 EPROM has 256,000 bits of storage. If you divide 256,000 by 8 (number of bits in a byte) you get 32,000 or 32K (K is an abbreviation for kilo or thousand). Please note that not all EPROM part numbers represent the number of bits. For example the 27C040 stores 4 million bits or 512K bytes of data.

What is EPROM speed?
The speed specification for an EPROM is actually a specification for access time. The access time is the time it takes the EPROM to deliver the data from the requested address to the CPU. The access time is specified in nanoseconds (billionths of a second). The access time is usually found following the part number. Example: M5M27C64A-25 is a 27C64 EPROM with a 250 nanosecond access time. The access time is normally found after the dash (-) which separates the part number from the access time. When you replace an EPROM or any memory device, you may substitute a part of equal or faster speed but not one of slower speed. Example: You can replace a 27C64A-25 with a 27C64A-20 but not with a 27C64A-30. The data can be available from the EPROM before it is needed however if the CPU attempts to take the data before it is there, the product the computer is in will not work.

Figuring out the part number on your chip
EPROMs and most other memory parts conform to a standard pin configuration and operate in an identical manner. This allows parts from any manufacturer to be used interchangably. The key is to figure out the part number on your device. Usually the part number is printed directly on the top of the part. What you are looking for is the core part number. This is the industry standard number for the part. The typical EPROM is a 27 series device. EPROMS compose about 80% of the programmable memory device market. Regardless of who manufactured the part, the part number should have 27XXX or 27CXXXX printed somewhere on the device. This number may have a prefix or suffix associated with it. An example would be the HITACHI 2716 EPROM. The part number on the chip is HN42716G. Note where the 2716 appears (HN42716G). An identical part made by National Semiconductor is numbered as NMC27C16Q-35 where a part manufactured by Intel is D2716-1. Don't get confused by a date code. A date code is printed on the chip to indicate when it was manufactured. A date code is the year of manufacture followed by the week. An example would be 8925. This indicates that the part was manufactured in the 25th week of 1989. Although a date code is not easily confused for a 27XXX chip, it can be for other types of memory parts which are not 27XXX devices.

Erasing an EPROM and the window
The information (bits) which compose the data in the EPROM are stored as an electrical charge on a transistor arary in the EPROM itself. There is one transistor for each bit of storage. The programming process forces the charge on the transistor where it remains trapped until the part is erased. A charged (programmed transistor) holds the bit value of 0 where an erased (unprogrammed transistor) holds the bit value of 1. Once a transistor in the EPROM is programmed, the only way to remove the trapped charge is to expose the entire chip to high intensity ultraviolet light using an EPROM ERASER. If you look at a standard EPROM you will see that it has a clear window in the center of the part. The purpose of this window is to allow the ultraviolet light to reach the transistors which compose the EPROM memory array and dissipate the electrons trapped on the transistors.

Usually an EPROM, once it has been programmed, will have a cover or sticker placed over the window. You cannot erase an EPROM or change its contents by removing this cover and exposing it to normal room light. Even in direct sunlight it would take two weeks of constant exposure to alter the chip. You cannot erase an EPROM with a black light or tanning bed, it is best to use an eraser manufactured to erase EPROMs and other programmable devices.

EPROMs in plastic packages
A very expensive part of the EPROM is the ceramic package with the quartz window. All manufacturers provide their EPROM products in plastic packages without windows. These are called OPT (One Time Programmable) devices simply because there is no window to erase the part. There is nothing to prevent an OTP device from being read and the contents placed into another EPROM. OTP devices are standard EPROMs which can only be programmed and used once. The part number is usually something such as P27CXXX where the "P" indicates a plastic case.

What is the difference between a 27 and a 27C part?
The 27XXX EPROMs use the first memory cell technology. The technology process is called NMOS. NMOS stands for "N-channel" "Metal" "Oxide" "Semiconductor". The next generation technology is called CMOS, thus the letter "C" in the part number. CMOS stands for "Complimentary" "Metal" "Oxide" "Semiconductor". CMOS parts use much less power than NMOS parts, however all other aspects of the devices are identical. You can substitute a CMOS for an NMOS part with no problems, however since an NMOS part consumes more power, it is not advisable to put an NMOS part in place of a CMOS. This is especially true in battery powered equipment.

Different package types
There are several package types into which memory parts or microcontrollers are found. The traditional package, which has been shown throughout this page, is called a DIP. DIP stands for "Dual" "Inline" "Package". It is so called because it has two rows of pins which insert into a socket or holes in the circuit board. The primary advantage of this package is that it is easily removed and installed by humans. In reality, the actual memory chip is much smaller than the package into which it is placed. DIP packages can be of any size between 8 and 42 pins. Typical EPROMs are normally found in 24, 28 or 32 pin packages depending on device storage capacity. There are 16 bit EPROMs which are found in 40 pin packages. The 40 pin package is used because the data is read 16 bits at a time and thus requires more physical pins to connect to the processor.

In order to conserve space there are several package styles which have become common. One is the PLCC. PLCC stands for "Plastic" "Leaded" "Chip" "Carrier". A PLCC package is physically smaller than a DIP and has leads on all four sides. The most common size for EPROMs is 32 pins, although smaller and larger PLCC packages are not uncommon. PLCC packages have the primary advantage of size. With electronic products becoming increasingly small, designers use PLCC parts for surface-mount or space limited designs. There are sockets available for PLCC parts, however they require a special tool to remove the device. If the PLCC part is soldered directly to the board, the part must be desoldered before it can be read. A special adapter is required to use a PLCC part with a device programmer. The adapter performs the conversion from the PLCC package to a traditional DIP base.

A second space conserving package style is the SOIC or PSOP. SOIC stands for "Small" "Outline" "Integrated" "Circuit". PSOP stands for "Plastic" "Small" "Outline" "Package". SOIC/PSOP devices have leads on only two sides and as such are similar to a DIP although much smaller. SOIC/PSOP packages may also be installed in a socket, although they are usually soldered directly to the circuit board. The device shown in the photo is a 29F100T flash memory part. A special clam-shell style adapter is required to support SOIC/PSOP devices which insures proper lead alignment and connection.

One of the smallest package styles available is the TSOP. TSOP stands for "Thin" "Small" "Outline" "Package". These devices are always soldered to the board and are also found in FLASH MEMORY CARDS and other space critical applications. A special adapter is required to support these devices. The adapter socket has tiny retractable contacts which mate with each fine-pitch device lead.

ADDITIONAL PROGRAMMABLE MEMORY PARTS

FLASH EPROMS
Flash EPROMs are the most recent type of EPROM memory part. These parts are always in plastic cases because the erasing of the part is done electrically without the need to expose the internal chip to ultraviolet light. The erase function is performed by the EPROM programmer and requires less time than a UV EPROM to be erased in an eraser. They erase in a "FLASH", and thus the name. Flash memory parts cannot be substituted for a standard UV erasable part because the minimum number of pins on a typical flash part is 32. The smallest flash device is a 28F256 which is equal in storage capacity to a 27256 UV EPROM. Note the part number 28FXXX is a FLASH device designated by the "F" in the part number.

EEPROMS
EEPROMs are memory parts which use the same technology as an EPROM, however each bit of storage has extra transistors to allow the bit to be individually erased. The term EEPROM means "Electrically" "Erasable" "Programmable" "Read Only Memory". Unlike Flash devices, these memory parts can have individual bytes changed without the need to erase the entire part. The way this works is that the memory chip itself performs an erase on a single byte before programming the byte with the new data. The erase and program of a single byte can take as long as 10 milliseconds. This delay prevents an EEPROM from replacing a normal RAM part. An example of a EEPROM is a 28C64. Note that the number 28 is used both for EEPROMs and FLASH devices. The "C" in the 28C64 part number means it is made with CMOS technology. EEPROMs are more costly than EPROMs because they have more transistors. EEPROMs are normally used for data storage in a product since they can be "written to" by the computer.

SERIAL EEPROMS
A serial EEPROM is created with the same technology used in larger parallel EEPROMs. The difference is that a serial EEPROM typically has only 8 pins on the package. This is because the address and data are sent to and from the chip one bit at a time using two or three wires. In order for the computer to read and write the serial EEPROM it must actually send commands, data and address information via a predefined communication protocol. This takes more time and is more complex than dealing with a larger EEPROM, however the cost and size savings make these parts very popular where the amount of information which must be stored is small. There are three serial EEPROM technology families: "Microwire" whose part numbers start with 93. The "I2C BUS" whose part numbers start with 24 and "SPI" (Serial Peripheral Interface). SPI part numbers start with 25 or 95. These parts are very common and used in many applications including automotive.

Serial EEPROMs are also found in surface mount packages called SOIC. SOIC stands for "Small" "Outline" "Integrated" "Circuit". These packages are very small and are always soldered directly to a circuit board. Under many circumstances it is possible to connect to these parts "in-circuit" and successfully read and program information.

NVRAMS
An NVRAM is a standard RAM (RAM stands for Random Access Memory) which is used for general operation in computers. The NV stands for non-volatile. The RAM would normally lose its contents when power is removed, however the NVRAM is manufactured with a built in battery which keeps power applied to the memory after power has been removed from the product. The primary advantage of an NVRAM is speed. Since the part operates as a normal high-speed memory part while power is applied, there is no delay when data is written into the device. There are also memory parts which are called NVRAMs however they do not have an internal battery. They maintain there contents by transferring the data from the RAM memory array into a duplicate EEPROM memory array. When power is removed, the EEPROM memory retains the data. When power is restored, the RAM memory is refreshed to its original state from the EEPROM array. Example parts of the battery based NVRAM is the Dallas Semiconductor DS1220(photo) or DS1225. An example of the EEPROM based NVRAM is the X22C10 and X22C12 from Xicor.

BIPOLAR PROMS
A BIPOLAR PROM is a memory chip which typically contains a small amount of data, however the access time is very fast. These memory chips store the bits of data by blowing small fuses inside the memory device. Once they are programmed they cannot be changed. The term BIPOLAR reflects the transistor technology used in the manufacturing process. PROM means "Programmable" "Read Only Memory". Programming a BIPOLAR PROM requires a high voltage, high current pulse to blow the fuse. These parts are found in many products and were a favorite in the United States Space Program because the data, once programmed, is immune to changes caused by radiation.

MICROCONTROLLERS
A microcontroller is a complete computer on a chip. This includes the central processing unit (CPU), the memory (EPROM or FLASH EPROM) and the I/O (input/output). Microcontrollers are usually found in products which perform a single function such as microwave ovens, computer printers, automatic sprinkler controllers, etc. A microcontroller, although much more complex than a standard memory part, still has internal EPROM/FLASH EPROM as its program memory. A microcontroller normally has 40 or more pins since the input and output operations occur directly from the chip. Some microcontrollers, like the PIC family from Microchip and others, have fewer I/O pins and are therefore fabricated in smaller packages. Since the microcontroller has internal memory, it is this memory area that you may read and program using a device programmer. In the past, most microcontrollers used internal EPROM for their program code and data, however now all new parts incorporate FLASH memory.

Microcontrollers with FLASH memory are excellent if you are interested in development since it takes much less time to erase and reprogram a flash based part than one with UV EPROM. Some examples of 40 pin FLASH microcontrollers are the ATMEL 89C51, 89S51 and 89C52. Examples of 18 and 20 pin FLASH microcontrollers are the Microchip PIC16F84 and ATMEL 89C4051.

WHAT IS HEX?
HEX is short for hexidecimal. Hexidecimal is base 16. Where humans are used to the decimal system (base 10) with numbers 0 through 9, hex has the digits 0,1,2,3,4,5,6,7,8,9,A,B,C D,E, and F. Although this may appear confusing, the reason HEX was chosen is that it provides a simple way to represent 4 binary bits as a single digit. EXAMPLE: Binary 0000 = HEX 0 or binary 0101 = HEX 5 or binary 1100 = HEX C. The table below shows binary numbers and their hexadecimal equivalents.

BINARY	HEX
0000	0
0001	1
0010	2
0011	3
0100	4
0101	5
0110	6
0111	7
1000	8
1001	9
1010	A
1011	B
1100	C
1101	D
1110	E
1111	F

From the above table you can see that if you wish to represent the binary number 1110010100110010 you can write it as E532 hex. This is why hex has become the defacto standard to represent and display binary data; it allows humans to represent large groups of binary data in a compact form.

WHAT IS ASCII? ASCII stands for American Standard Code for Information Interchange. This is a standard translation for HEX (binary) codes into printable or readable characters. The letters you are reading on the screen right now are represented in the computer as groups of bits. Without a standard, any computer manufacturer could decide which group of bits would represent which letter or number. In ASCII an upper case A is represented by the value 41H. the number 3 is represented by the value 33H. When you view the editor display of a device programmer you should see a display of HEX information and also (usually to the right) a display of the ASCII translation of the HEX data. If the data from an EPROM contains characters which are to be displayed to a user, these characters will appear in the ASCII display area.