In electronic devices, one of the most important elements that ensure the operation of the entire system is memory, which is divided into internal and external. Elements internal memory consider RAM, ROM and CPU cache. External- these are all kinds of drives that are connected to a computer from outside - hard drives, flash drives, memory cards, etc.

Read-only memory (ROM) is used to store data that cannot be changed during operation, random access memory (RAM) to place information from the processes currently occurring in the system in its cells, and cache memory is used for urgent processing of signals by the microprocessor .

What is ROM

ROM or ROM (Read only memory - Read only) - a typical storage device that does not change information, included in almost every component of a PC and phone and required to start and run all elements of the system. The content in the ROM is written by the hardware manufacturer and contains directives for pre-testing and starting the device.

ROM properties are independence from power, the impossibility of rewriting and the ability to store information for long periods of time. The information contained in the ROM is entered by the developers once, and the hardware does not allow it to be erased, it is stored until the end of the computer or phone's service, or its breakdown. Structurally ROM protected from damage during voltage drops, therefore, only mechanical damage can damage the information contained.

By architecture, they are divided into masked and programmable:

In masks devices, information is entered using a typical template at the final stage of production. The contained data cannot be overwritten by the user. The separating components are typical pnp elements of transistors or diodes.
In programmable ROMs (Programmable ROM), information is presented in the form of a two-dimensional matrix of conductive elements, between which there is a pn junction of a semiconductor element and a metal jumper. Programming such a memory occurs by removing or creating jumpers by means of a current of high amplitude and duration.

Main functions

ROM memory blocks contain information on managing the hardware of a given device. The ROM includes the following subroutines:

Directive start and control for the operation of the microprocessor.
A program that checks performance and integrity all hardware contained in a computer or phone.
A program that starts and ends a system.
subroutines that control peripheral equipment and I/O modules.
Information about the address of the operating system on the physical drive.

Architecture

Persistent storage devices are made in the form two-dimensional array. The elements of the array are sets of conductors, some of which are not affected, other cells are destroyed. Conductive elements are the simplest switches and form a matrix by connecting them in turn to rows and rows.

If the conductor is closed, it contains a logical zero, open - a logical unit. Thus, data in binary code is entered into a two-dimensional array of physical elements, which is read by the microprocessor.

Varieties

Depending on the method of manufacturing the device, ROM is divided into:

Ordinary factory-created. The data in such a device does not change.
Programmable ROMs that allow the program to be changed once.
Erasable firmware, which allows you to clear data from elements and overwrite them, for example, using ultraviolet light.
Electrically cleanable, rewritable elements that allow multiple change. This type is used in HDD, SSD, Flash and other drives. The BIOS on motherboards is written on the same microcircuit.
Magnetic, in which information was stored on magnetized areas, alternating with non-magnetized ones. They could be overwritten.

Difference between RAM and ROM

The differences between the two types of hardware are in its safety when the power is turned off, speed and the ability to access data.

In RAM (Random access memory or RAM), information is contained in sequentially arranged cells, each of which can be accessed through software interfaces. RAM contains data about currently running processes in the system, such as programs, games, contains the values of variables and lists of data in stacks and queues. When you turn off your computer or phone, RAM memory completely cleared. Compared to ROM memory, it has faster access speed and power consumption.

ROM memory is slower and uses less power to run. The main difference lies in the inability to change the incoming data in the ROM, while the information in the RAM is constantly changing.

Read Only Memory (ROM)- A memory designed to store immutable information (programs, constants, table functions). In the process of solving problems, the ROM allows only reading information. As a typical example of the use of ROM, one can point out the LSI ROM used in the PC to store the BIOS (Basic Input Output System - basic input-output system).

In the general case, a ROM drive (an array of its memory cells) with a capacity of EPROM words, a length of r+ 1 bits each, usually a system of horizontal (address) and r+ 1 vertical (discharge) conductors, which at the intersection points can be connected by connection elements (Fig. 1.46). Communication elements (EC) are fusible links or p-n-transitions. The presence of an element of communication between j-m horizontal and i m vertical conductors means that in i-th digit of memory cell number j one is written, the absence of ES means that zero is written here. Writing a word to cell number j ROM is produced by the proper arrangement of the connection elements between the bit conductors and the address wire number j. Read word from cell number j ROM goes like this.

Rice. 1.46. A ROM drive with a capacity of EPROM words, a length of r+ 1 digits each

Address code A = j decrypted, and on the horizontal conductor number j drive is powered by a power source. Those of the discharge conductors that are connected to the selected address conductor by coupling elements are energized U 1 unit level, other discharge conductors remain energized U 0 level zero. Set of signals U 0 and U 1 on the discharge conductors and forms the content of the JP number j, namely the word at the address BUT.

Currently, ROMs are built from LSI ROMs that use semiconductor ES. LIS ROM is usually divided into three classes:

- mask (MPZU);

– programmable (PROM);

- reprogrammable (RPZU).

Mask ROMs(ROM - from Read Only Memory) - ROM, information in which is recorded from a photomask during the crystal growth process. For example, the BIS ROM 555RE4 with a capacity of 2 kbytes is a character generator according to the KOI-8 code. The advantage of mask ROMs is their high reliability, and the disadvantage is their low manufacturability.

Programmable ROMs(PROM - Programmable ROM) - ROM, information in which is written by the user using special devices - programmers. These LSIs are manufactured with a full set of ES at all intersection points of address and discharge conductors. This increases the manufacturability of such LSI, and hence the mass character in production and application. Recording (programming) of information in the PROM is made by the user at the place of their application. This is done by burning out the connection elements at the points where zeros should be written. We indicate, for example, on the TTLSH-BIS PROM 556RT5 with a capacity of 0.5 kbytes. The reliability of LSI PROMs is lower than that of masked LSIs. Before programming, they must be tested for the presence of ES.

In EPROM and EPROM, it is impossible to change the contents of their PL. Reprogrammable ROMs(RPZU) allow multiple changes of the information stored in them. In fact, EPROM is a RAM that has t RFP>> t Thu. Replacing the contents of the EPROM begins with the erasure of the information stored in it. RPZU with electric (EЕPROM) and ultraviolet (UVEPROM) deletion of information are issued. For example, a KM1609RR2A LSI with electrical erasure with a capacity of 8 kB can be reprogrammed at least 104 times, stores information for at least 15,000 hours (about two years) in the on state and at least 10 years in the off state. LSI RROM with ultraviolet erasure K573RF4A with a capacity of 8 kB allows at least 25 rewriting cycles, stores information in the on state for at least 25,000 hours, and in the off state for at least 100,000 hours.

The main purpose of EPROMs is to use them instead of ROMs in software development and debugging systems, microprocessor systems and others, when you have to make changes to programs from time to time.

The operation of a ROM can be viewed as a one-to-one transformation N-bit address code BUT in n-bit code of the word read from it, i.e. ROM is a code converter (digital machine without memory).

On fig. 1.47 shows a conditional image of the ROM in the diagrams.

Rice. 1.47. Symbolic ROM image

The functional diagram of the ROM is shown in fig. 1.48.

Rice. 1.48. Functional diagram of the ROM

According to the terminology adopted in the environment of specialists in storage devices, the input code is called an address, 2 n vertical tires - numerical rulers, m outputs - bits of the stored word. When any binary code enters the ROM, one of the number lines is always selected. At the same time, 1 appears at the output of those OR elements whose connection with a given number line is not destroyed. This means that 1 is written in this bit of the selected word (or number line). At the outputs of those bits whose connection with the selected number line is burned out, zeros will remain. The law of programming can also be inverse.

Thus, ROM is a functional unit with n entrances and m outputs, storing 2 n m bit words that do not change during operation of a digital device. When the address is applied to the input of the ROM, the corresponding word appears at the output. In logical design, permanent memory is considered either as a memory with a fixed set of words, or as a code converter.

On the diagrams (see Fig. 1.47), ROM is referred to as ROM. Read-only memory devices usually have an enable input E. With the active level at input E, the ROM performs its functions. In the absence of permission, the outputs of the microcircuit are inactive. There can be several enabling inputs, then the microcircuit is unlocked by the coincidence of the signals at these inputs. In ROM, the signal E is often called reading CHT (read), the choice of the VM chip, the choice of the VC crystal (chip select - CS).

ROM chips are designed to be scaled up. To increase the number of digits of stored words, all microcircuit inputs are connected in parallel (Fig. 1.49, but), and from the increased total number of outputs, the output word is taken correspondingly to the increased word length.

To increase the number of stored words themselves (Fig. 1.49, b) the address inputs of the microcircuits are connected in parallel and are considered as the least significant bits of the new, extended address. The added upper bits of the new address are sent to the decoder, which selects one of the microcircuits via the E inputs. With a small number of microcircuits, the decoding of the highest bits can be done on the conjunction of the enable inputs of the ROM itself. The outputs of the same bits with an increase in the number of stored words must be combined using the OR functions. Special OR elements are not required if the outputs of the ROM chips are made either according to the open collector circuit for combining by the wired OR method, or according to the three-state buffer circuit, which allows direct physical combination of outputs.

The outputs of ROM microcircuits are usually inverse, and input E is often also inverse. ROM expansion may require the introduction of buffer amplifiers to increase the load capacity of some signal sources, taking into account the additional delays introduced by these amplifiers, but in general with relatively small amounts of memory, which is typical for many CUs ( for example, automation devices), increasing the ROM usually does not give rise to fundamental problems.

Rice. 1.49. An increase in the number of digits of stored words when the microcircuit inputs are connected in parallel and an increase in the number of stored words when the address inputs of the microcircuits are connected in parallel

Very often, various applications require the storage of information that does not change during the operation of the device. This is information such as programs in microcontrollers, bootloaders (BIOS) in computers, tables of digital filter coefficients in signal processors, DDC and DUC, sine and cosine tables in NCO and DDS. Almost always, this information is not required at the same time, so the simplest devices for storing permanent information (ROM) can be built on multiplexers. Sometimes read-only memory devices are referred to in translation literature as ROM (read only memory). A diagram of such a read only memory (ROM) is shown in Figure 3.1.

Figure 3.1. A read-only memory (ROM) circuit based on a multiplexer.

In this scheme, a permanent storage device is built for eight single-bit cells. Storing a specific bit in a single-bit cell is done by soldering the wire to the power source (writing one) or soldering the wire to the body (writing zero). On schematic diagrams, such a device is designated as shown in Figure 3.2.

Figure 3.2. The designation of a read-only memory device on circuit diagrams.

In order to increase the capacity of a ROM memory cell, these microcircuits can be connected in parallel (the outputs and the recorded information naturally remain independent). The scheme of parallel connection of single-bit ROMs is shown in Figure 3.3.

Figure 3.3 Scheme of a multi-bit ROM (ROM).

In real ROMs, information is recorded using the last operation of the microcircuit production - metallization. Metallization is performed using a mask, so such ROMs are called mask ROMs. Another difference between real microcircuits and the simplified model given above is the use of a demultiplexer in addition to the multiplexer. This solution makes it possible to convert a one-dimensional memory structure into a two-dimensional one and, thereby, significantly reduce the volume of the decoder circuit required for the operation of the ROM circuit. This situation is illustrated by the following figure:

Figure 3.4. Schematic of the mask read-only memory (ROM).

Masked ROMs are depicted in circuit diagrams as shown in Figure 3.5. The addresses of the memory cells in this chip are fed to pins A0 ... A9. The chip is selected by the CS signal. Using this signal, you can increase the amount of ROM (an example of using the CS signal is given when discussing RAM). The chip is read by the RD signal.

Figure 3.5. Conventional graphic designation of mask ROM (ROM) on circuit diagrams.

The mask ROM is programmed at the factory, which is very inconvenient for small and medium production runs, not to mention the device development stage. Naturally, for large-scale production, mask ROMs are the cheapest type of ROM, and therefore are widely used at present. For small and medium production series of radio equipment, microcircuits have been developed that can be programmed in special devices - programmers. In these ROMs, the permanent connection of conductors in the memory matrix is replaced by fusible links made of polycrystalline silicon. During the production of the ROM, all jumpers are made, which is equivalent to writing logical units to all ROM memory cells. In the process of programming the ROM, increased power is supplied to the power leads and outputs of the microcircuit. In this case, if the supply voltage (logical unit) is applied to the output of the ROM, then no current will flow through the jumper and the jumper will remain intact. If a low voltage level is applied to the ROM output (connected to the case), then a current will flow through the memory matrix jumper, which will evaporate it, and when information is subsequently read from this ROM cell, a logical zero will be read.

Such chips are called programmable ROM (PROM) or PROM and are depicted on circuit diagrams as shown in Figure 3.6. As an example of a PROM, microcircuits 155PE3, 556RT4, 556RT8 and others can be mentioned.

Figure 3.6. Conventional graphic designation of programmable read-only memory (PROM) on circuit diagrams.

Programmable ROMs have proved to be very convenient for small- and medium-scale production. However, when developing electronic devices, it is often necessary to change the program written to ROM. In this case, the ROM cannot be reused, therefore, once written ROM, with an erroneous or intermediate program, you have to throw it away, which naturally increases the cost of developing equipment. To eliminate this shortcoming, another type of ROM was developed that could be erased and reprogrammed.

ROM with UV erasure is built on the basis of a memory matrix built on memory cells, the internal structure of which is shown in the following figure:

Figure 3.7. Memory cell ROM with ultraviolet and electrical erasure.

The cell is a MOS transistor with a polycrystalline silicon gate. Then, during the manufacturing process of the microcircuit, this gate is oxidized and as a result it will be surrounded by silicon oxide - a dielectric with excellent insulating properties. In the described cell, with the ROM completely erased, there is no charge in the floating gate, and therefore the transistor does not conduct current. When programming the ROM, a high voltage is applied to the second gate located above the floating gate, and charges are induced in the floating gate due to the tunnel effect. After the programming voltage is removed, the induced charge remains on the floating gate, and hence the transistor remains in the conductive state. The charge on the floating gate of such a cell can be stored for decades.

The block diagram of the described read-only memory does not differ from the previously described mask ROM. The only difference is that the cell described above is used instead of a fusible link. This type of ROM is called reprogrammable read-only memory (EPROM) or EPROM. In the EPROM, the erasure of previously recorded information is carried out by ultraviolet radiation. In order for this light to pass unhindered to the semiconductor crystal, a quartz glass window is built into the housing of the ROM chip.

When the EPROM chip is irradiated, the insulating properties of silicon oxide are lost, the accumulated charge from the floating gate flows into the semiconductor volume, and the storage cell transistor goes into the closed state. The erasing time of the RPZU chip ranges from 10 to 30 minutes.

The number of write-erase cycles of EPROM chips is in the range from 10 to 100 times, after which the RPZU chip fails. This is due to the destructive effect of ultraviolet radiation on silicon oxide. As an example of EPROM microcircuits, we can name microcircuits of the 573 series of Russian production, microcircuits of the 27сXXX series of foreign production. The EPROM most often stores the BIOS programs of mainframe computers. RPZU are depicted on circuit diagrams as shown in Figure 3.8.

Figure 3.8. Conventional graphic designation of RPZU (EPROM) on circuit diagrams.

Since cases with a quartz window are very expensive, as well as a small number of write-erase cycles, they have led to the search for ways to erase information from an EPROM electrically. Many difficulties were encountered along the way, which have now been practically resolved. Now, microcircuits with electrical erasure of information are quite widespread. As a memory cell, they use the same cells as in the EPROM, but they are erased by electric potential, so the number of write-erase cycles for these microcircuits reaches 1,000,000 times. The time for erasing a memory cell in such ROMs is reduced to 10 ms. The control circuit for electrically erasable programmable ROM turned out to be complex, so two directions for the development of these microcircuits have been outlined:

1. EEPROM (EEPROM) - electrically erasable programmable read-only memory

Electrically erasable EPROMs (EEPROMs) are more expensive and smaller in size, but they allow you to overwrite each memory cell separately. As a result, these microcircuits have the maximum number of write-erase cycles. The scope of electrically erasable ROM is the storage of data that should not be erased when the power is turned off. These microcircuits include domestic microcircuits 573РР3, 558РР3 and foreign EEPROM microcircuits of the 28cXX series. EEPROMs are designated on circuit diagrams as shown in Figure 3.9.

Figure 9. Electrically erasable read-only memory (EEPROM) symbol on circuit diagrams.

Recently, there has been a tendency to reduce the size of the EEPROM by reducing the number of external microcircuit pins. To do this, the address and data are transferred to and from the chip via a serial port. In this case, two types of serial ports are used - SPI port and I2C port (microcircuits 93cXX and 24cXX series, respectively). The foreign series 24cXX corresponds to the domestic series of microcircuits 558РРX.

FLASH - ROMs differ from EEPROMs in that the erasure is not performed for each cell separately, but for the entire microcircuit as a whole or the memory matrix block of this microcircuit, as was done in the EPROM.

Figure 3.10. Conditional-graphic designation of FLASH memory on circuit diagrams.

When accessing a permanent storage device, you must first set the address of the memory cell on the address bus, and then perform a read operation from the microcircuit. This timing diagram is shown in Figure 3.11.

Figure 3.11. Timing diagrams of signals for reading information from ROM.

In Figure 3.11, the arrows show the sequence in which control signals should be generated. In this figure, RD is the read signal, A is the cell address select signals (since individual bits in the address bus can take on different values, transition paths are shown both to the one and to the zero state), D is the output information read from selected ROM location.

4. Perform the addition operation in an additional code, presenting the given terms in binary form:

1) + 45 2) - 45

- 20 + 20

Solution:

1) x 1 \u003d 45 \u003d 0.101101 pr

x 2 \u003d - 20 \u003d 1.010100 pr \u003d 1.101011 arr \u003d 1.101100 additional

+ 1,101100

Answer: 0.011001 pr \u003d 25 10

2) x 1 \u003d - 45 \u003d 1.101101 pr

x 2 \u003d 20 \u003d 0.010100 pr

+ 0,010100

Answer: 1.100111 add \u003d 1.011000 arr \u003d 1.011001 pr \u003d - 25 10

Question number 5.

Complete the following tasks:

1) write a logical function in SNDF;

2) minimize the logical function using Karnaugh maps;

The main classification parameters of the memory

Parameter	Designation	Definition
Information capacity	N	The number of bits of memory in the memory drive
Number of words and memory	P	Number of word addresses in storage drive
Bit depth	T	The number of bits in the memory drive
Output fanout	Kp	The number of unit loads (inputs of other ICs) that can be simultaneously connected to the output of the memory
Number of reprogramming cycles	Ncy	The number of write-erase cycles at which the operability of the memory is maintained
Power consumption	p CC	Power consumption of the charger in the set mode of operation
Power consumption in storage mode	p CCS	The power consumed by the memory when storing information in the non-selection mode
Information storage time	tSG	The time interval during which the memory in a given mode stores information

3U static parameters

An important advantage of ROM over RAM is that information is retained when the power is turned off. The cost of a bit of information stored in ROM can be almost an order of magnitude lower than in RAM. Permanent memory can be implemented based on various physical principles.

The following types of ROM are currently in use:

MASK ROMS are programmed by their manufacturer, who, according to the information prepared by the user, makes photo templates, with the help of which he enters this information into the ROM chip during the production process. This method is the cheapest and is intended for large-scale production of ROMs.

Masked ROMs are built on the basis of diodes, bipolar and MIS transistors. In diode ROMs, the diodes are on at those intersections of the matrix that correspond to the entry "1", and are absent in the places where "0" should be written. The external control circuits for diode ROMs are very simple. Since the diode matrix is a galvanically coupled element, the output signals have the same shape as the input ones. Permanent memory on MOS transistors is somewhat easier to manufacture than bipolar ones.

Masked ROMs are highly reliable, but it is impossible to change the information in the ROM without making a new IC, which is especially inconvenient at the stage of developing the system programs.

USER PROGRAMMABLE ROM are more versatile and therefore more expensive. They are matrices of bipolar devices with fusible jumpers (their simplified diagram is shown in Fig. 17.7), the connections of which with the address and bit buses are destroyed when the code is entered on special PROGRAMMERS. These devices generate voltages necessary and sufficient to burn out the fuses in the selected ROM memory elements.

On fig. . PP fuses are shown in the form of fuses included in the emitters of multi-emitter transistors VТo...VТp. The programmable elements are connected between the emitters of the matrix transistors and the bit lines. The presence of a jumper corresponds to a logical 0 at the output of the readout amplifier, and the absence of a jumper corresponds to a logical one. The process of writing information to the circuit is the selective destruction of fusible jumpers by the current provided by the programming device.

ONE TIME PROGRAMMABLE ROM (PROM) the drive is made on the basis of cells. Permanent memories of this type allow only a single entry of information into a cell. When programming, these fusible jumpers made of nichrome or other refractory material are burned out using a special programming device. The jumpers are burned in the programming mode by a series of pulses according to a special program.

To increase the reliability of the ROM operation, the programming technique provides for the supply of a series of 40,.. 100 pulses after fixing the moment of jumper burnout, as well as the obligatory thermal training of the programmed ROM at a temperature (about 100 ° C).

More reliable are microcircuits with jumpers made of polycrystalline silicon, in which the process of irreversible transition of polysilicon from a conductive state to a non-conductive one occurs under the action of heating caused by the flow of current.

Programming mode support circuits are usually located on the chip itself, and the programming process proceeds as follows.

1) The address of the selected cell is applied to the address inputs.

2) The supply voltage of the +U microcircuit is increased to the programming voltage +10 V necessary to create a current, I ³ 400 mA, sufficient to melt the jumper.

3) A voltage of +15 V with a current of not more than 100 mA is applied to the programming input V through a resistor

FIRMWARE ROM (EPROM) The most widespread among them are ROMs with ultraviolet erasure and with electrical erasure and recording of information.

Microcircuits in which information is erased using ultraviolet radiation (UFPROM) have: the possibility of multiple programming, a fairly short access time and power consumption, and a large capacity.

The storage element in UV-erased ROM is a MOSFET. Information about the contents of this cell is stored as a charge on the second gate of the MOSFET. If it is necessary to reprogram the microcircuit, the previously recorded information is erased with ultraviolet light c l £ 400 μm (the source can be a DRT220 or DRT375 lamp) through a transparent quartz window on the surface of the microcircuit housing. UV radiation discharges the floating gate of the MOSFET. The storage time of information in ROM microcircuits of this type is determined by the quality of the gate dielectric and for modern microcircuits it is ten years or more.

ROM chips with electrical erasure of information are popular with developers of microprocessor technology due to the ability to quickly erase and write, a large allowable number of information rewriting cycles (10,000 times or more). However, they are quite expensive and complex compared to ROM chips with UV erasure, and therefore they are inferior to the latter in terms of their use in microprocessor equipment.

The basis of the memory cell in EEPROM is a floating gate MOSFET, the same as in UVE ROM. But in microcircuits of this type, technological methods provide the possibility of reverse tunneling, i.e. selection of electrons from the floating gate, which allows you to selectively erase the entered information.

FERROELECTRICITY, the electrical analogue of ferromagnetism. Just as in ferromagnetic substances, when placed in a magnetic field, residual magnetic polarization (moment) appears, in ferroelectric dielectrics placed in an electric field, residual electric polarization occurs.

The microscopic cause of ferroelectricity is the presence of atomic (or molecular) dipoles inside the substance. These dipoles are oriented by an external electric field and remain oriented after the field is removed; switching the direction of the field to the opposite leads to the reverse orientation of the dipoles. The fundamental difference between ferroelectricity and ferromagnetism is that free electric charges can screen electric fields created by electric dipoles, and this makes direct observation of static polarization difficult. Polarization is usually measured by the so-called hysteresis loop. The sample is placed between the plates of a capacitor, to which an alternating voltage E is applied. On the oscilloscope screen, a curve of dependence of the charge arising on the plates, and thereby the electric polarization, is recorded (since the charge per unit surface area of the plates is a measure of the electric polarization vector P), on voltage (field) E. The hysteresis loop shown in fig. 1 is characterized by two quantities: the remanent polarization P (of any sign), which is present even at zero field E, and the coercive field Ec, at which the polarization vector reverses direction. The area of the hysteresis loop is equal to the work of electric forces expended within one cycle of ferroelectric transition between two equivalent polarization states of opposite sign.

At the moment, there is a huge number of all possible combinations of the main elements from which a memory cell is built - a ferromagnetic ferroelectric transistor and the same capacitor. But when considering these combinations, it is possible to identify 4 main types that are basic, all other types of FeRAM cells are just their combinations. This is a 1T FeRAM single-transistor cell, a 1C FeRAM single-capacitor cell, also called SFRAM (statically read, non-volatile, ferroelectric random access memory - a complete analog of SRAM), the most common 1T-1C FeRAM transistor-capacitor cell and the most stable double cell of all of the above. 2T-2C FeRAM. And now in more detail.

In addition to these basic structures, there are a huge number of their combinations. Almost any university that is more or less self-respecting is now sorting through cell layout options and studying the properties of these hybrids. Diplomas on this topic are being defended, more and more patents are being obtained. To consider at least the most promising combinations within the framework of one article is unrealistic. There is material for at least one more article, but for now it is worth moving on to the further prospects of FeRAM.

This cell structure was used in one of the first working FeRAM models, but its performance was not up to par - the cell lost its charge too quickly and went into an unpredictable state, that is, it was not non-volatile, so work in the 1T area was curtailed. But the idea itself turned out to be tenacious - after all, having only one transistor as a cell, you can achieve its minimum size and, accordingly, a gigantic information capacity per unit chip surface. That is why in 2002 work on the creation of 1T FeRAM was continued by the two largest Japanese institutes - NERI (Nanoelectronics Research Institute) and AIST (National Institute of Advanced Industrial Science and Technology). Using the latest generation of ferromagnetic ferroelectrics - composite oxide SBT (SrBi2Ta2O9) with the addition of hafnium Hf and slightly modifying the structure of the field ferroelectric transistor (ferroelectric gate field-effect transistor), they managed to obtain a 1T structure with a significantly longer charge storage time, an order of magnitude more than previous developments.

The 1T FeRAM circuit itself looks like this:

On the left is a diagram of a traditional 1T-1C cell, on the right is only 1T. Even from the circuit diagram, it is clear that the 1T cell is smaller and easier to implement compared to the 1T-1C, which should have a positive effect on the cost and on the information capacity of the memory based on it.

The transistor itself looks like this:

Writing to a 1T FeRAM cell is carried out when a positive or negative charge is applied to the electrodes of the circuit. When the +6V voltage is applied to the drain electrode, a pulsating adequate current occurs in the conductor channel corresponding to the value "1". And vice versa - after applying a negative voltage - the pulsating current is extremely small - the cell goes to the "0" position.

On a chart, it looks like this:

As follows from this graph, the difference between the state "0" and the state "1" is sufficient to unambiguously determine the value of the cell, and the drop in the leakage current is insignificant - in 106 seconds (which corresponds to 11.6 days), the drop did not exceed 2%.

Summing up, we can say that this technology is quite viable - extremely small cell size, charge stability and high cell access speed (what could be simpler than a transistor?) - these are the key positions of 1T FeRAM. The main problem is the reliability of charge storage - memory based on 1T FeRAM loses data after 50-60 days. However, this is not relevant for the mobile computer market - it is unlikely that any of the PDA owners will have their favorite toy turned off for more than two months, and when turned on, the charge on the transistors is updated. Therefore, it remains for the creators of 1T to improve reliability and, most importantly, to put all this into practice - and this seems to be the main problem, none of the major FeRAM manufacturers have yet become interested in this new reincarnation of the old idea, preferring to deal with more traditional 1T-1C and 2T-2C . To date, there has been no news about the licensing of 1T technology by any major manufacturer. Apparently, stereotypes are tenacious - having once rejected the 1T structure, the giants of the computer industry have completely forgotten about it. I would like to believe that this, as the developers called it, ultra-Gbit FeRAM, will be lucky with publishers, and we will see cheap capacious non-volatile storage media on the shelves.

16 kbit non-volatile ferroelectric RAM (FRAM) with serial interface and 3V power supply

Distinctive features:

Ferroelectric non-volatile RAM with a capacity of 16 kbit
- Organization of memory cells 2048 x 8
- Unlimited read/write cycles
- 10 year information storage period
- Recording without delay (NoDelay™)
- Advanced highly reliable ferroelectric technology

Fast 2-wire serial interface
- Maximum serial bus clock frequency up to 1 MHz
- Direct hardware replacement of EEPROM

Low power consumption
- Work with 2.7-3.6V power supply (new feature)
- Active current - 75 μA (100 kHz, 3V)
- Quiescent current - 1 uA

Industry Compliance
- Operating temperature: -40° C … +85° C
- 8-pin. SOIC package
- Availability of environmentally friendly 8-pin. SOIC packages (new feature)

Structural diagram of FM24CL16:

Pin assignment FM24CL16:

General description:

FM24CL16 is a non-volatile memory with a capacity of 16 kbit, made using ferroelectric technology. Ferroelectric Random Access Memory or FRAM is non-volatile and performs read and write operations similar to RAM. It provides reliable storage of information for 10 years, while eliminating problems associated with the complexity, limited write speed and level of system reliability of EEPROM and other non-volatile memory.

Unlike the EEPROM, the FM24CL16 performs the write operation at bus speed. In this case, there are no delays in recording.

The next bus cycle can be started immediately without the need for data polling. In addition, the device has an unlimited number of write cycles, which is many orders of magnitude greater than that of EEPROM. Also, FRAM draws much less current when writing than EEPROM, which requires an additional internal power supply for the programming circuitry.

These features make the FM24CL16 ideal for non-volatile storage applications where frequent and fast data logging is required. Examples of such applications range from data storage, where write time can be critical, to industrial control, where EEPROM write delays can cause information loss. Together, these advantages allow you to write data more frequently without causing programming inconvenience.

FM24CL16 is available in industry standard 8-pin. SOIC package and uses a two-wire communication protocol. Performance is guaranteed over the entire industrial temperature range -40°C to +85°C. FM24CL16 requires 3V power supply and provides bus speed up to 1 MHz, while being functionally compatible with the 5V version of FM24C16.

Pin Description:

Order Information:

Thin films of lead lanthanum zirconate titanate (PLZT) are being actively studied with the aim of creating energy-dependent microelectronic memory devices using silicon technology. (Bistable polarization is the ideal basis for binary memory cells.)

As a result of the transition of technologies for the production of semiconductor products to a process of less than 1 micron, it became necessary to reduce the supply voltage accordingly. There is currently an increasing trend in the market to move from 5V systems to 3V systems. However, not all component base follows this trend, and system designers face the difficulty of using components with a single power supply. This problem is even more true for systems maintenance companies that save money by redesigning obsolete 5V parts.

Atmel has taken this into account when designing the new AT45DBXXXX series of the DataFlash family with 3V power only. However, the 3V DataFlash family can also be used in 5V systems. This how-to guide is intended to provide recommendations for using 3V DataFlashes in 5V or mixed-supply systems.

HEXADIMAL NUMBERS

The memory cell of a typical microcomputer may contain the binary number 1001 1110. Such a long chain of zeros and ones is difficult to remember and inconvenient to type from the keyboard. The number 1001 1110 could be converted to decimal, which would give 158 10 , but the conversion process would take a long time. Most microinformatics systems use hexadecimal notation to make binary numbers like 1001 1110 easier to remember and use.

The hexadecimal number system (hexadecimal), or base 16 system, uses 16 characters from 0 to 9 and A, B, C, D, E, F. In Table. 2.5 shows the equivalents of decimal, binary and hexadecimal numbers.

Note from Table. 1 that each hexadecimal character can be represented by a single combination of four bits. Thus, the representation of the binary number 1001 1110 in hexadecimal code is the number 9E. This means that the 1001 part of the binary number is 9 and the 1110 part is equal to E (of course, in hexadecimal code). Therefore, 1001 1110 2 = 9E 16 . (We should not forget that the indices mean the base of the number system.)

How to convert binary number 111010 to hexadecimal? We must start with MB and divide the binary number into groups of 4 bits. Then you need to replace each group of 4 bits with the equivalent hexadecimal digit: 1010 2 =A, 0011 2 =3, therefore 111010 2 =3A 16.

How to convert hexadecimal number 7F to binary? In this case, each hexadecimal digit must be replaced by its 4-bit binary equivalent. In the example, the binary number 0111 is replaced

Table 1. Decimal, hexadecimal and binary equivalents

Decimals	Hexadecimal	Binary











	A
	IN
	FROM
	D
	E
	F

hexadecimal digit 7, and 1111 2 replaces F 16 , whence 7F 16 = 11110111 2 .

Hexadecimal notation is widely used to represent binary numbers.

Table 2. Hexadecimal to decimal conversion

Power of sixteen	16 3	16 2	16 1	16 0
Position value
Hexadecimal		FROM		E
Decimal	4096 x 2 =	256 x 12 =	16 x 6 =	1 x 14 =
	8192+	3072+	96+	14 = 11374

Let's convert the hexadecimal number 2C6E to decimal. The procedure of actions corresponds to the table. 2. The positions of the first four hexadecimal digits are respectively from left to right 4096, 256, 16 and 1. The decimal number contains 14 (E 16) ones, 6 numbers 16, 12 (C 16) numbers 256 and 2 numbers 4096. Each digit is multiplied by the weight corresponding to it, the sum is obtained, which gives us the decimal number 11374.

Let's convert the decimal number 15797 to hexadecimal. On fig. 5 shows the operating procedure. In the first line, 1579710 is divided by 16, which is

15797 10:16 = 987 remainder 5 10 = 5 16 MP

978 10: 16 = 61 remainder 11 10 = B 16

61 10:16 = 3 remainder 13 10 = D 16

3 10: 16 = 0 remainder 3 10 =3 16 SR

15797 10 = 3 D B 5

Rice. 5. Decimal-Hexadecimal Conversion

gives the quotient 987 10 and the remainder 5 10 , which is then converted to its hexadecimal equivalent (5 10 = 5 16) and becomes the least significant digit (MP) of the hexadecimal number. The first quotient (987) becomes divisible in the second line and is again divided by 16, giving the quotient 61 and the remainder 11 10 or hexadecimal B. In the third line 61 is divided by 16, giving the quotient 3 and the remainder 13 10 or D 16 , and in the fourth line dividend 3 is divided by 16, gives the quotient 0 and the remainder 3 10 or 3 16. When the quotient is 0, as in the fourth line, the conversion ends. 3 16 becomes the most significant digit (SR) of the result, i.e. 3DB5 16 .

Send your good work in the knowledge base is simple. Use the form below

Students, graduate students, young scientists who use the knowledge base in their studies and work will be very grateful to you.

Posted on http://www.allbest.ru/

Novgorod State University I. Wise

abstract

Presentation on theme: "Permanent Storage Devices. Main characteristics, scope"

Completed: 1st year student gr. 5261

Bronina Ksenia

Checked by: Arkhipova Gelirya Askhatovna

Veliky Novgorod, 2016

1. The concept of permanent storage

1.1 Key Features of ROM

1.2 ROM classification

1.2.1 By type of execution

1.2.2 By types of ROM chips

1.2.3 By the method of programming microcircuits (writing firmware in them)

2. Application

3. Historical ROM types

Literature

1. The concept of permanent storage

A read-only memory (ROM, or ROM - Read Only Memory, read-only memory) is also built on the basis of modules (cassettes) installed on the motherboard and is used to store immutable information: operating system boot programs, computer device test programs and some drivers basic input/output system (BIOS), etc.

Permanent memory includes read-only memory, ROM (in English literature - Read Only Memory, ROM, which literally translates as "read-only memory"), reprogrammable ROM, PROM (in English literature - Programmable Read Only Memory, PROM), and flash memory. The name of the ROM speaks for itself. The information in the ROM is written at the factory of the memory chips, and its value cannot be changed later. The ROM stores critical information for the computer, which does not depend on the choice of operating system. Programmable ROM differs from the usual one in that the information on this chip can be erased by special methods (for example, ultraviolet rays), after which the user can re-write information on it. This information will not be deleted until the next erasing operation.

It is customary to refer to ROM as non-volatile permanent and "semi-permanent" storage devices, from which information can only be read quickly, information is written to ROM outside the PC in the laboratory or with a special programmer and in the computer. According to the technology of recording information, the following types of ROM can be distinguished:

§ microchips programmed only during manufacture - classic or masked ROM or ROM;

§ microcircuits programmed once in the laboratory - programmable ROM (PROM), or programmable ROM (PROM);

§ Reprogrammable microcircuits - reprogrammable ROM or erasable PROM (EPROM). Among them, electrically reprogrammable EEPROM (Electrical Erasable PROM) chips, including flash memory, should be noted.

1.1 Key Features of ROM

Read-only memory (ROM) data is permanently stored. Data stored permanently is called non-volatile, which means that it remains in ROM even when the power is turned off. Once data is written to ROM, it can be read by other devices, but new data cannot be written to ROM.

ROM is most commonly used to store the so-called "monitor program". A monitor program is a machine program that allows the user of a microcomputer system to view and modify all system functions, including memory. Another wide use of ROM is to store fixed tables of data, such as mathematical functions, that never change.

Four types of ROM are widely used by digital computer systems: mask-programmed ROM, programmable ROM (PROM), erasable programmable ROM (EPROM), and electrically programmable ROM (EPROM).

1.2 ROM classification

1.2.1 By type of execution

The data array is combined with the sampling device(reader), in this case, the data array is often called “firmware” in conversation:

§ ROM chip;

§ One of the internal resources of a single-chip microcomputer (microcontroller), usually FlashROM.

The data array exists on its own:

§ CD;

§ punched card;

§ perforated tape;

§ barcodes;

§ mounting "1" and mounting "0".

1.2.2 By types of ROM chips

According to the crystal manufacturing technology:

§ RO M English read-only memory - read-only memory, masked ROM, manufactured by the factory method. There is no possibility to change the recorded data later.

Figure 1. Mask ROM

§ PRO M English programmable read-only memory - programmable ROM, "flashed" by the user once.

Figure 2. Programmable ROM

§ EPROM erasable programmable read-only memory - reprogrammable / reprogrammable ROM (EPROM / EPROM)). For example, the contents of the K573RF1 chip were erased using an ultraviolet lamp. For the passage of ultraviolet rays to the crystal, a window with quartz glass was provided in the microcircuit case.

Figure 3. Flash ROM

§ EEPROM electrically erasable programmable read-only memory - electrically erasable reprogrammable ROM). This type of memory can be erased and filled with data several tens of thousands of times. Used in solid state drives. One of the varieties of EEPROM is flash memory (English flash memory).

Figure 4 Erasable ROM

§ ROM on magnetic domains, for example, K1602RTs5, had a complex sampling device and stored a fairly large amount of data in the form of magnetized areas of the crystal, while not having moving parts (see Computer memory). An unlimited number of rewriting cycles was provided.

§ NVRAM, non-volatile memory - "non-volatile" memory, strictly speaking, is not ROM. This is a small amount of RAM, structurally combined with a battery. In the USSR, such devices were often called "Dallas" after the name of the company that launched them on the market. In the NVRAM of modern computers, the battery is no longer structurally connected with the RAM and can be replaced.

By type of access:

§ With parallel access (parallel mode or random access): such a ROM can be accessed in the system in the RAM address space. For example, K573RF5;

§ With serial access: such ROMs are often used for one-time loading of constants or firmware into a processor or FPGA, are used to store TV channel settings, etc. For example, 93С46, AT17LV512A.

1.2.3 By the method of programming microcircuits (writing firmware in them)

§ Non-programmable ROM;

§ ROM, programmable only with the help of a special device - a ROM programmer (both once and repeatedly flashed). The use of a programmer is necessary, in particular, for applying non-standard and relatively high voltages (up to +/- 27 V) to special outputs.

§ In-circuit (re)programmable ROMs (ISP, in-system programming) - such microcircuits have a generator of all the necessary high voltages inside, and can be flashed without a programmer and even without desoldering from a printed circuit board, programmatically.

memory chip programming monoscope

2. Application

Read-only memory is often written to control firmware for a technical device: a TV, a cell phone, various controllers, or a computer (BIOS or OpenBoot on SPARC machines).

BootROM is firmware, such that if it is written to a suitable ROM chip installed in a network card, it becomes possible to boot the operating system to a computer from a remote local network node. For built-in network cards, BootROM can be activated through the BIOS.

ROM in IBM PC-compatible computers is located in the address space from F600:0000 to FD00:0FFF

3. Historical ROM types

Read-only memory devices began to find application in technology long before the advent of computers and electronic devices. In particular, one of the first types of ROM was a cam roller, used in hurdy-gurdies, music boxes, and striking clocks.

With the development of electronic technology and computers, the need arose for high-speed ROM. In the era of vacuum electronics, ROMs based on potentialoscopes, monoscopes, and ray lamps were used. In computers based on transistors, plug-in matrices were widely used as small-capacity ROMs. If it was necessary to store large amounts of data (several tens of kilobytes for computers of the first generations), ROMs based on ferrite rings were used (they should not be confused with similar types of RAM). It is from these types of ROM that the term “firmware” originates - the logical state of the cell was set by the direction of the winding of the wire that encircles the ring. Since a thin wire had to be pulled through a chain of ferrite rings, metal needles similar to sewing needles were used to perform this operation. And the very operation of filling the ROM with information resembled the process of sewing.