The principle of operation and device flash

In this article we will talk about what is the basis of creation and the principle of the flash memory device (not to be confused with USB flash drives and memory cards). In addition, you will learn about its advantages and disadvantages over other types of ROM (permanently storage device) and get acquainted with the range of the most common drives that contain flash memory.

The main advantage of this device is that it is nonvolatile and does not need electricity to store data. All stored information in the flash memory can be considered an infinite number of times, but the number of full write cycles are unfortunately limited.

Flash memory (flash memory) — refers to the semiconductors electrically reprogrammable memory (EEPROM). The technical solutions without the high cost, large volume, low power consumption, high speed operation, compactness and mechanical strength, a flash memory embedded in portable digital devices and storage media.

The flash memory to other devices (hard drives and optical drives) type ROM has its advantages and disadvantages, which you can get from the table below.

Hard driveA large amount of stored information.


High-speed operation.

Low cost storage data (per 1 MB).

Large size.


Sensitivity to vibration.



The optical discEase of transport.


low cost storage of information.

The possibility of replication.

a Small amount.


Should the reader.

Constraints during operations (read, write).

Low speed operation.

Sensitivity to vibration.


Flash memoryHigh-speed data access.


Low battery consumption.

Vibration resistance.

Ease of connection to the computer.

Compact size.

a Limited number of write cycles.

Today no one doubts that flash memory will continue to strengthen its position in information technology, especially in the range of mobile devices (PDAs, tablets, smartphones, iPods). Based on flash memory are the most popular USB flash drives and a removable memory card for electronic devices (SD, MMC, miniSD…).

Memory cards and USB drives don't stand apart and attract the attention of potential buyers for its diversity. From such abundance of memory devices only wins the manufacturer and the consumer suffers a number of inconveniences. After all, we all know those situations, when the phone need one map, PDA other, the camera third. This range drives into the hands of the producers, because they derive from a wide exclusive sale of great benefit. Here is a small list of common flash memory:

  • Compact Flash Type I (CF I)/Type II (CF II);
  • Memory Styck (MS Pro, MS Duo);
  • Secure Digital (SD);
  • and
  • xD-Picture Card (xD);
  • MultiMedia Card (MMC).
  • USB Flash Drive.{: Type}ROM

In one of the publications I wrote how to choose a USB flash drive, and on how to choose a card format SD card (microSD, miniSD) see here.

Principle of operation the flash memory.

Elementary data storage cell, flash memory is a transistor with a floating gate. The peculiarity of this transistor is that it is able to retain electrons (charge). That's based on it and developed the basic types of flash memory and и and. Competition between them, because each type has its advantage and disadvantage. By the way, on their basis build a hybrid version such as and и Noise..

In the flash memory manufacturers use two types of memory cells MLC and SLC.

  • Flash memory with MLC (Multi-level cell — multi-level memory cells)of the cell is more capacious and cheap, but they had great access time and fewer cycles of write/erase (about 10000).
  • Flash memory, which contains SLC (Single-level cell — single-level memory cells) of the cell has a maximum number of cycles of write/erase(100000) and it has a smaller access time.

The change in the charge (writing/erasing) is performed by the application between the gate and source of great potential to the electric field in the thin dielectric between the transistor channel and the pocket was sufficient for the occurrence of the tunnel effect. To enhance the effect of tunneling of electrons in a pocket while recording, there is a small acceleration of electrons by passing a current through the channel of a field effect transistor.

principle of operation the flash memory based on the change and registration of electric charge in an isolated area («pocket») of the semiconductor structure.

A read is performed the field-effect transistor for which the pocket acts as a shutter. The potential of the floating gate alters the threshold characteristics of the transistor that is registered by the reading circuits. This design is supplied with elements that allow it to work in a large array of the same cells.

Now consider in more detail the memory cells with one or two transistors are…

 A memory cell with one transistor.

If the control shutter to apply a positive voltage (initializing the memory) then it will be in an open state, which would correspond to a logical zero.

And if the floating gate of a to put the excess negative charge (electron) and apply positive voltage to the control shutter ,then it compensates for the generated control shutter electric field and will not be allowed to form the channel conductivity, and hence the transistor will be in the closed state.

So, the presence or absence of charge on the floating gate determines the exact open or closed state of the transistor when supplied the same positive voltage to the control gate. If we consider the supply voltage to the control gate, as the initialization memory cell, according to what the voltage between source and drain it is possible to judge the presence or absence of charge on the floating gate.

Thus obtained a sort of elementary memory cell is able to store one information bit. To all this it is very important that the charge on the floating gate (if available) could be preserved there for a long time, as in the initialization memory cell, and in the absence of voltage on the control gate. Only in this case, the memory cell is non-volatile.

So how, if necessary, on the floating gate to place charge (save memory contents) and delete it from there (erasing memory contents) when necessary.

To place charge on the floating gate (writing process) is possible by injection of hot electrons (CHE-Channel Hot Electrons) or by tunneling the Fowler-Nordheim of.

If you use the method of injection of hot electrons, the drain and the control shutter high voltage, giving the electrons in the channel enough energy to overcome the potential barrier, which creates a thin dielectric layer, and to send (tunnel) in the area of the floating gate (while reading on the control shutter lower voltage and the effect of tunneling does not occur).

To remove charge from the floating gate (execute erasing of the memory cell) on the control shutter high negative voltage (about 9 V) and on the area of the source is supplied a positive voltage. This leads to the fact that the electrons are tunneling from the floating gate region to the area of the source. So is quantum tunneling of Fowler — of Nordheim (Fowler — Nordheim).

you've Probably realized that a transistor with a floating gate is a unit cell of the flash memory. But cells with one transistor have some disadvantages, the main of which is the poor scalability.

creating the memory array, each memory cell (i.e., transistor) is connected to two perpendicular to the tire. The control gates are connected to the bus, which is called the line of words (Word Line), and drains connected to the bus, its called the bit line (Bit Line). In consequence of which in the circuit is high voltage when data is written by injection of hot electrons all line — words, bits, and origins must be placed at a great distance from each other. This will give the desired level of isolation, but will affect the limit of the flash memory.

Another disadvantage of such memory cells is the presence of an excess removal of the charge from the floating gate, and he is not compensated for by the recording process. In consequence, on the floating gate is formed a positive charge, which makes the steady-state condition of the transistor and it always remains open.

A memory cell with two transistors.

Dvuhreaktornaya the memory cell is a modified anotherstory the cell, which is a conventional CMOS transistor and a transistor with a floating gate. In this structure, a conventional transistor acts as a insulator transistor with a floating gate from the bit line.

Whether the benefits dvuhreaktornaya memory? Yes, because it can help to create a more compact and scalable memory chips, because here the transistor with a floating gate is isolated from the bit line. Everything else, unlike onetransistor memory cells where information is written by injection of hot electrons in dvukhchastotnoi memory cells for writing and erasing of information is used the method of quantum tunneling of Fowler — of Nordheim. This approach makes it possible to reduce the voltage necessary for write operation. Looking ahead I will say that dvukhchastotnyi cell used in the memory structure of the and.

The flash memory device with a and architecture.

The type of this memory is the source and kind of push in the development of the entire EEPROM. Its architecture was designed by Intel in 1988. As was written previously to access the contents of the memory (initialize the cell), you need to supply voltage to the control gate.

Therefore, the developers of all of the control gates connected to the control line called a line of words (Word Line). Analysis of information memory cells is performed by the signal level at the drain of the transistor. So the developers all the drains of the transistors connected to the line, called the line bit (Line Bit).

Architecture and got the name due to logical OR — NOT (in English and). The principle of logical operation and is that it on multiple operands (the data argument of the operation…) gives a value of unity when all the operands equal to zero, and a value of zero in all other operations.

In our case, under the implied operand is the value of the memory cells, and therefore in this architecture, a single value of the bit line is observed only in the case where the value of all cells that are connected to the bit lines will be equal to zero (all transistors are closed).

In this architecture, well organized random access memory, but the process of recording and erasing of data is performed relatively slowly. In the process of recording and erasing method of injection of hot electrons. In addition, the chip flash memory architecture and and the size of her cell turns out big, so this memory scales poorly.

The structure of the six cells in a and Flash

Flash memory with and architecture is typically used in devices for storage of program code. It may be phones, PDAs, BIOS, motherboard…

The flash memory device with and architecture.

This type of memory was developed by Toshiba. These chips thanks to its architecture used in small drives , which received the name and (logical AND-NOT). When performing a and operation yields the value zero only when all the operands equal to zero, and a single value in all other cases.

As described earlier, a value of zero is the open state of the transistor. In consequence, in the and architecture assumes that the bit line has a value of zero in the case where all the connected transistors are open and a value of one when at least one of the transistors is closed. This architecture can be build if you connect the transistors with a bit line is not one (as architecture and) , and successive series (a column of series-connected cells).

This architecture compared to and scales well because allows compactly to place the transistors on the diagram. In addition, the architecture of and writes by tunneling Fowler — of Nordheim, and this permits to realize fast recording than in the structure of and. To increase the speed of reading, in the chip embedded and flash internal cache.

As the clusters of the hard disk and and cells are grouped in small blocks. For this reason, sequential read or write speed advantage will have and. But on the other hand and loses much in the operation of the random access and does not have the ability to work directly with bytes of information. In a situation when you need to change a few bits, the system is forced to rewrite the entire block and is considering a limited number of write cycles, will lead to greater wear of the memory cells.

the Structure of one and Flash column

Recently, there are rumors that the company Unity Semiconductor is developing flash memory of the new generation, which will be built on technology, CMOx. It is expected that the new memory will replace flash-memory and flash and overcome its limitations, which in the and memory due to the architecture of the transistor structures. The advantages of CMOx include higher density and write speed, as well as more attractive. Among the applications of the new memory are SSD and mobile devices. Well, true or not time will tell.

For more details to bring to You all the necessary information I posted a video on the topic.

PS to Explain in simple >

If you found an error, please highlight the text and press Ctrl+Enter.