This topic is dictated by the presence of many "excess" in the form of matrices (TFT displays) remaining from laptop and netbooks, tablets, monitors, and even from televisions. Therefore, developments on options for their use are discussed here. The topic is very difficult and requires a big attention and careful approach to the implementation of its projects. Those. "Nakhpom" at best, you can not achieve anything, and at worst - also to block firewood. Better I will do as a teacher - I will try "on the fingers" and professionally deselect various technical details arising in the process of incarnation of one or another idea.
Before we immediately proceed to specific schemes and examples, it is necessary to have an idea about everything that accompanies this, and only then draw conclusions about the possibility of implementing this or that project, so let’s consider the technical aspects.
So, the key concept is the display controller. It is called the "controller", and not the "scoleler", as is customary in common. Scaler (from the word scale - scale, scale) is one of the components of the controller functional scheme that is responsible for scaling the image. The controller is exactly the "piece of iron", which is engaged in the conversion of certain interfaces (HDMI, VGA, CVBS, RGBS) into signals, "understandable" matrix. In the overwhelming majority (almost 90% of the total market), controllers are focused on connecting matrices with the LVDS interface. In order to connect the matrix to the controller, you need to make sure whether the controller supports this matrix, as well as comply with the entire list of conditions that this controller must also satisfy and what else should be on it. The controller itself is actually a "blank", which is managing the firmware. And on how much it approaches a specific matrix, the entire success of the enterprise depends.
How everything should look as an assembly is easier to portray in the picture:
This is the most common scheme for the assembly of such a "venture." There is nothing scarce and super-expensive. According to this scheme, matrixes are connected from all monitors and matrixes with lamp illumination from laptops. If we are talking about the launch of matrices with LED backlighting, the scheme will be slightly modified:
The differences are not global, but otherwise both of the above designs, with the condition of serviceable parts, have a guaranteed 100% repeatability.
Of course, I am talking about the available gland, which is produced in China precisely for homemade structures. On the one hand, this is good - mostly all Chinese controllers are designed for the maximum coverage of available matrices. But it is necessary to note the fact that not all 100% of existing matrices can thus be translated into the project. It is meant that there is a series of matrices that cannot be called universal, interchangeable, etc. There are also very non-standard loners, but pleases that they are not so much. We are talking about matrices with rarely encountered interfaces, "raise" which is pretty difficult, and sometimes impossible. It can be safely attributed to the matrix with purely individual temporal characteristics that were laid in tablets or netbooks (where these matrices stood in the first life). Not every controller is able to give these characteristics because in the overwhelming majority of firmware, all the same on some "averaged" and unified parameters are calculated. Controllers who hypothetically could correctly earn with such matrices are those that with a "TV" on board (will be below the spoiler). Those. Those in the settings of which there is a MAP LVDS item in which there is at least some possibility of choosing from 16 presets settings.

BEFORE BEGINING THE PROJECT
1. It is necessary to stock documentation for your matrix - without it the success of the event is nearing zero. What I mean. The documentation on the matrix is ​​the file in the .pdf format, which is accompanied by the release of the matrix itself. Correctly called Datasheet. It is written in it: from the electrical and optical parameters of the panel itself and the illumination parameters to geometric and installation sizes. There is also necessarily there is data on its interface and its connection scheme. A person without experience without it to connect the matrix will be very problematic. Datasheets in the overwhelming majority in English, very rare, but come across Chinese or Korean. Therefore, do not hope that it will be in Russian - no one translates these things. Datasheets are searched on the Internet through a search engine by tag "[Matrix Name] .pdf. Cancel and look. I want to note that the site panelook.com (he is first found in the search for any matrix) - just the World Handbook and Datasheets just does not distribute so much, only for money. So do not hope for him in the search. You can try to look intolibraryBeyondinfinite site. There is certainly not a panel, but quite a lot of things and is heard free. However, it often happens that the documentation on the matrix is ​​not located - this is not the fault of the forum and not mine, but the manufacturer of the matrix. If faced with a similar situation, try to search for a matrix with the same name (or with a very similar name), but another revision or other suffixes - in 95% of cases it helps. Also in this case, the search for compatible models of matrices is helped - for example, see the sales sites and may be listed there will be compatible part-numbers. Show logic and smelting.
2. Which controller is the best - you decide for yourself, because this is primarily a matter of taste and technical requirements that you set.
3. Before buying something, you need to decide according to paragraph 2 which controller is best for you, and then it is MANDATORY to view the firmware archive available for it to determine if there is a firmware that matches your matrix. If there is no suitable one, no one will write it to you, I guarantee it to you. Whether the firmware is suitable or not, it will become clear to you after reading the entire material.
What matrices fall under the possible implementation:
- The vast majority of matrices from monitors, laptops, tablets and TVs. If we expand a little on what has been said - matrices with LVDS, TTL and RSDS interfaces.
- matrices with eDP interface (from laptops of the middle and upper price segment). However, such a project will be released 1.5-2 times more expensive than with "ordinary" matrices.
- matrices with MIPI interface (these are rather high-quality matrices from some tablets). It is quite possible to raise such a matrix, controllers with a MIPI interface exist, only their cost in China is around $ 180-200. Who wants to - buy and do ...
What matrices conditionally fall under the possible implementation:
- matrices from oldportable Chinese DVD players, televisions, as well as a large number of photo frames. These are the so-called "analog" (more correctly RGB) matrices of very low resolution, often with lamp lighting. More I stayed on them here . In this topic, it makes no sense to consider them because of the almost complete absence of universal controllers for them, the lack of firmware (if there is something else) and, as a rule, it is impossible to recognize the matrices themselves to find out their parameters, because the overwhelming majority of them are noname.
What matrices do not fall under the possible implementation:
- Matrixes with LVDS interface with "portrait" resolution (this is when in specifications it is indicated that the number of pixels horizontally less than vertical, for example, 600 * 1024, 800 * 1280 - Check on the page PANELOOK.COM by the name of the matrix - everything is all write "correctly"). Why I focused on the type of interface - because the matrixes with the MIPI interface, in contrast to the matrices with the LVDS interface, on the contrary, in the overwhelming majority portrait, and the controller is responsible for the "rotation" of 90 degrees, which, of course, has much more intelligence, than the LVDS controller. And in those tablets where LVDs portrait matrices stood, the processor is responsible for the correct orientation of the picture.
- All matrices from phones (smartphones).
- matrices from cameras
- Matrix of plasma TVs. This is so clear from the title of the topic, but since questions arise from time to time, I mark it with a separate item. No iron for homemade for them does not even exist in nature.

As for large television matrices, since the abundance of combinations of technical solutions in them is an order of magnitude higher than that of monitor or notebook computers, the project with them must be approached very carefully. Why - it will become clear after a careful reading of this material to the end.

1. What do tablets have?
- Matrices come in three interfaces: LVDS, TTL and MIPI.
- The LVDS interface in tablet matrices is 1-channel and can be 6- and 8-bit (depending on the matrix).
- Illumination in tablet matrices only LED. Depending on the type of the resulting matrix, it can be removed by a separate connection of the LED line, and it can work from the built-in frame of the matrix backlight. It is clear that in the first case, a separate driver will need, providing LED current according to data from the documentation on the matrix. Built-in driver eliminates the need to search for external. It usually eats from some fixed voltage (3.5V, 5V, 12V, often the supply voltage has extensive limits, for example from 5-7 to 15-20V). Such illumination in budget plates is not found.
- The current consumption of the matrix panel itself for the 7 "matrix in white light is about 100mA. That is, at 3.3V supply voltages (in tablet matrices always), the panel consumes 0.33W of power.
- Matrices come with a widescreen resolution of 800 * 480, 1024 * 600, 1280 * 800, 1280 * 720, 1366 * 768 and "square" resolution of 800 * 600, 1024 * 768.
- Matrices are with "portrait" and "landscape" resolution. Ie, for example, there are 1280 * 800, and there are 800 * 1280. Technically, they are completely different and not interchangeable matrices.

2. What is in monitors.
- Matrices in monitors are in the vast majority of two interfaces: LVDS and RSDS. It is extremely rare, but still there are matrices with a TTL interface and even matrices with a VGA and DVI interface are noticed.
- The LVDS interface in monitor matrices is 1- and 2-channel. Single-channel LVDS is usually found in square matrices with a diagonal of 15 ", rarely in 17". However, starting from 17 "and the 2-channels are usually applied above. In premium monitors 23" ... 27 "- there is a 4-channel. LVDs interface in monitors is always 8-bit. Exception is some very old monitors small diagonal (11 - 15 inches) - There are 6-bit matrices and modern premium monitors in which even a 10-bit interface is seen.
- Illumination in monitor matrices is CCFL (luminescent, 2- and 4-lamp - edge, less often, in large diagonals there can be 6-8 lamps - Direct) and LED (from 1 to 4 ... 8 LED groups). What CCFL is that LED backlight in monitors involves the application of the standard backlight driver. It is usually combined with a monitor power supply and is controlled from the Mineland. From the lane, Enable (or BLON) signals (allow) and PWM (or DIM_B) (brightness adjustment) arrive at it. Sometimes in old models there is a DIM_A - analog adjustment when the brightness depends on the voltage level.
- Power consumption by the matrix panel from 1.5W (15 "square) to 4.5 ... 5W (22-23" widescreen). Matrix panels are powered by 3.3 or 5V, depending on the matrix model.
- Matrices come with a widescreen resolution of 1366 * 768, 1440 * 900, 1600 * 900, 1680 * 1050, 1920 * 1080, 1920 * 1200 and "square" resolution of 1024 * 768, 1280 * 1024, 1600 * 1200.
- Matrices are ONLY with "landscape" resolution.

3. What is in laptops.
- Matrices are "square", i.e. with a 4: 3 ratio (800 * 600 and 1024 * 768) and widescreen with a 16: 9 ratio (1024 * 600, 1280 * 720, 1366 * 768, 1600 * 900 and 1920 * 1080) and 16:10 (1280 * 800 , 1440 * 900 and 1920 * 1200)
- The backlight may be luminescent (one linear or g-shaped lamp) and LED (one LED band, consisting of several groups of LEDs). For CCFL illumination, the driver is used, which stood in this laptop. For matrices with CCFL illumination, laptop drivers almost all interchangeable (electrically not structurally!). For LED backlighting in buoy matrices, a driver is already built into the matrix, and its control and power contacts are displayed on the matrix interface connector. LED matrix without an embedded backlight driver in laptops are extremely rare.
- The interface in notebook matrices can be either LVDS or eDP (for high resolution matrices, retina, yablochniki, etc.). The LVDS interface in low-resolution matrices is single-channel, it can be both 6-bit and 8-bit. In the high-resolution matrices can occur and dual. It is necessary to look at the documentation on the matrix.

What is meant by "bit depth" and "channel". Let's try a little bit to find out what it is for.
Bit rate
First of all, it should be noted that the LVDS interface is “intermediate” and not “final”. The final data supplied directly to the matrix panel are presented as so-called signals. TTL levels.
The depth (level) of each of the color components (RGB) in the TTL structure is presented either 6 or 8 lines for each color. And since the signal on each line can take a high or low level value, each of these lines is a kind of carrier of 1 bit of information. Those. For the 8-bit presentation of the "green" color in the TTL structure of the signal, the G0 ... G7 lines are used, for "red" - R0 ... R7 and for "blue" B0 ... B7. Accordingly, for the 6-bit presentation of the RGB line will be with indexes 0 ... 5. In addition to the color TTL levels, of course, there are also service lines with various sync signals. So all these TTL signals are encrypted in LVDS signals and on the side of the matrix with the help of special microcircuits are "mined" of them. It should be noted that there are two standards for the location of RGB signals in LVDS - VESA and Jeida. In small tablet, laptop and monitor matrices, in the overwhelming majority, there are no choice of system and these matrices are designed to work in the VESA system by default. It is even possible to argue with some confidence that this system is used mainly in the technique of European and Korean brands (regardless of the place of assembly of devices). The Jeida system is more oriented to Japanese equipment. For example, it employs matrices in Sony TVs, Sharp and oddly enough, also "Korean" Samsung. Moreover, if in television matrices produced by LG-Philips, Auo, Boe, Chimei-Innolux, at least there is the ability to select a color coding system in LVDS, then in some instances of Samsung, Sharp matrices and even AUO (produced by Sony) this choice Simply no - the matrix besides Jeida nothing else "understands."
Without going into deep physical processes, it is possible to consider the differences between the VESA and JEIDA even in the datasheet in the matrix. I'm just giving an example in which these differences are clearly visible. On the left in the picture order of alternation of the color components, and the right - their location in the timing diagram LVDS:

The controller, in turn, "prepares" LVDS signals in its depths. At the very beginning of the development of LCD technology, TTL signals were present in any controller explicitly and then encoded in LVDS using specialized microcircuits. At the current level of chip integration, LVDS signals come out of the controller chip directly without additional processing.
This begs the question: why was all this done? It's simple: to reduce the number of connecting lines. After all, the 8-bit TTL interface contains about 30 lines, and the 8-bit LVDS is only 10. Well, there is a possibility, at the very least, to call the LVDS interface universal.
As probably many have already noticed that, for example, the controller is still, which color of the color to transfer to the matrix - it is set to firmware, and its capabilities are enough to give the highest. Therefore, the concept of visibility relates rather to the possibilities of the matrix itself. It so much that small matrices work with a color depth, described by 6 or 8 bits on color, and the matrix 17 ... 22 inches in the overwhelming majority of 8 bits on color. Therefore, in firmware and mentioned, what it is, 6- or 8-bit. Sometimes the values ​​themselves are multiplied by the number of colors (3 - R, G, and B) and you can meet the concept of 18 or 24 bits - such a classification is used in motherboards equipped with the release of LVDS .. Well, most importantly, what exactly does it affect This is the maximum number of shades that the matrix can convey. 6-bit matrix - 262 thousand, 8-bit - 16,7 million. I have a hurry to calm down, and possibly someone to surprise: if you put two matrices 6 and 8 bit and apply the same picture on them, but prepared with the maximum content of shades for each personally, then you can hardly see the differences in the unarmed eye. How many shades will be able to recognize the human eye, there is no accurate data, and there can be no - people are all different. According to data from various sources, the human eye is able to recognize from 3 thousand to a million shades, and maximum indicators relate to people with relevant professional activities. The remaining on average is capable of no more than tens of thousands. But the main thing - all studies of ophthalmic organizations are not at all relate to LCD matrices, because the matrix is ​​a source of light. And as is known as the range of colors and shades, the human eye is distinguished by the human eye significantly, than the range of perception of "passive" colors. So let everyone accept this information, as far as he allows his "religion".
To find out what kind of bits the matrix, just look at its Datasheet - everything is written there. But besides this can be determined by the LVDS interface. The LVDS interface itself is represented by a certain set of informational and clocking differential pairs. Informational differential pairs are denoted as RXn + and RXn-, where n is a digit (pair number) from 0 to 2 or 3. Thus, the 6-bit matrix interface has such a set of differential pairs:
RX0 (+/-), RX1 (+/-), RX2 (+/-), RXCLK (+/-)
and the 8-bit interface has another pair:
RX0 (+/-), RX1 (+/-), RX2 (+/-), RXCLK (+/-), RX3 (+/-)
This begs the question: "is it possible ...?"
I answer:
- If there is a 6-bit signal on a 6-bit matrix (i.e., leave a pair of RX3 from the controller), then the matrix will show a brutal kaleidoscope of colors, but there will be no correct image. Externally, the picture looks like a color with highly low-lowered athletes (for example, if you put the color depth of 32, but for example, 4 bits). This is explained by the fact that according to the picture above, the part of the full picture does not reach the matrix (after all, the steam Rx3 is not connected) - respectively, part of the video information is lost.
- if you send a 6-bit signal to an 8-bit matrix, the picture will be correct, but too dark and no adjustments will be able to stretch the brightness and contrast to the desired level.
Some fine matrices can work with both 6- and 8-bit signals, and the bitness is selected by applying the appropriate level to a separate special output in the interface, which is designated as SELB or 6 / 8bit. However, as is well known eminent matrices have many clones. So often, even though the pin for selecting the bit rate is, but the matrix does not change its mode - the Chinese can save money on this ...
Often in the parameters of the color depth of the matrices can be found the concept of 6bit + Hi FRC. What is it - you can read on the wiki, but one thing is for sure - this is the optical parameter of the matrix panel, and not the possibility of its electronics. Such matrices are connected via an 8-bit interface.

Canal LVDS.
As mentioned above, the LVDS interface contains a certain set of information and one clocking differential pair. In the matrix of monitors and televisions to increase their bandwidth interface matrix make a two-channel. Those. the number of dif pairs is doubled and one more letter is added to their names RX:
O (Odd) - primary channel
E (Even) - the secondary channel
Then the pairs are called RXO0 +, RXOC +, RXO2-, etc. for primary and respectively RXE0 +, RXEC +, RXE2- (and so on) for the secondary. Those. it turns out that dual-channel 8-bit LVDS works in 10 pairs. Naturally, this is not about renaming pairs, but about a complete change of electronics and an algorithm for its operation. Therefore, one-channel and two-channel LVDS are not the same thing at all, and iron should produce exactly that signal on the matrix that it is designed hardwired for.
You can connect a single-channel matrix to a two-channel signal (and vice versa), nothing will burn. But nothing good will happen.

ATTENTION!!! Different matrix manufacturers often use their own internal labeling of LVDS channels, so do not panic, if you find in the matrix datasheet that LVDS pairs call differently. I give examples of such variations:
RXO = RXIN = RIN
RXOCLK = CLKIN = RCLK
sometimes "positive" (+) and "negative" (-) wires are labeled p (positive) and n (negative). Those. for example, RXIN0P means RXO_0 +.
it can be the replacement of numbers of pairs instead of numbers by letters, i.e. instead of "0", "1", "2" is found "A", "B", "C"
The most fanciful designations come across on the television matrices of large diagonals. For example, in the matrices LG-Philips there are notation of the type R1AN ... R1CN and R2AN ... R2CN, which means RXO_0- ... RXO_2- and RXE_0- ... RXE_2-. Synchro pairs are denoted R1CLKN / R1CLKP (RXCLK- / RXCLK +), etc.
I will separately dwell on television matrices with a diagonal of 32 inches and above.
All universal controllers are designed to use matrices with a refresh rate of 60Hz. This is usually 6 or 8-bit single or two-channel matrices. Many modern television Full HD matrices often go with a 10-bit LVDS interface, and the LVDS interface can be 2- and 4-channel. It is clear that it will not work directly to connect such a matrix to universal controllers. Separate instances of 2-channel matrices have an electrical ability to switch the LVDS interface between 8 and 10 bits - in this case, there will be no special problems with the connection - a sufficient matrix to switch to 8 bits mode. But the matrices having a 120Hz update frequency and a 4-channel 10-bit LVDS interface, without "special vehicles" with a universal controller, simply do not work because it is not designed for any 10 bits, nor on the 4th LVDS channel. What exactly I mean by "special needs" - this will be in the second part of the "caps".
If anyone is interested in digging deeper into this theory - you can start exploringhere. And then you will figure out what to look for (who cares, of course ...).

I repeat once again - we are talking about UNIVERSAL controllers for homemade products.
In China, a huge amount of them are produced and I cannot cover all the models. But among all the diversity there are several types with which success is more likely than with some rare and exotic (but still universal) controllers. First of all, I mean the availability of arrays of firmware and the presence of more or less intelligible manuals.
In any case, divide them into two groups:

CONTROLLERS FOR CONSTRUCTION OF MONITORS
This means that the controllers behave exactly like any computer monitor - "go to sleep" when no signal is input. sleep time - from 2 to 5 seconds.

- controller on chip RTD2660H (or RTD2662) (software designated as PCB800099):

- a simplified version of the above controller, in which there is only an HDMI input (software designated as PCB800661):

There is another version of such a controller, only it differs in that instead of the LVDS pin connector it contains an FFC 50pin 0.5mm connector and is oriented mainly under the matrix with a TTL interface (the software is designated as PCB800168):

It should be noted that in these controllers, the sound from the stream of HDMI is not extracted.
Not so long ago, a modified first version appeared on the market - it has audio extraction from HDMI on board, audio input "from the side" (automatically connected in VGA and AV modes) and a class D amplifier (most likely with a power of about 2 * 3W) (software The software is labeled PCB800196):
True, the introduction of sound did not pass without a “loss” - the FFC 50 pin connector disappeared from the board and the amplifier settled in its place. Thus, the controller "lost" some clients with a TTL interface and now it can only be lifted through a special adapter, which stands almost like a controller itself.
However, since the park of TTL matrices has not gone anywhere, the variant for these matrices was released as in the previous case (the software, if I am not mistaken, is designated PCB800809):

All of these controllers require firmware using the I2C bus and without the presence of a programmer will not work in any way. The first of these controllers has firmware practically under all imaginable and unthinkable solutions of matrices, the other two in the abundance of firmware are already very trimmed. The latter (the one that with sound) has simply the Spartan set of firmware around the area 20. Another point. Controllers at the RTD2660H chip (RTD2662) for some reason do not complain about the permission of FullHD. According to the description, the chip itself can, but by virtue of some reasons for such a purpose it is used quite rarely. And the Chinese are not recommended for this.

- The controller is only with one VGA entrance and a "jumper" selection of permission to the desired matrix. The cheapest of all options. In addition, he does not need firmware. When choosing (and there are several options, there are several). It is necessary to look at the list of the list on the reverse side of the board and the list of possible feed voltages of the matrix - there are options 3.3 / 5V and 3.3 / 5 / 12V. Another point - often on the charges of such controllers in the list of listed permissions, errors are made in the trimming of some modes, as well as in the signature of the LVDS cutting, the comb is confused "+" and "-" LVDS lines.
- controller with a full set of "computer" M.NT68676.2A inputs. Chip Novatek NVT68676UFG. for HDMI audio is decoded and output from the amplifier. When you turn to the work of VGA and DVI inputs audio from external input is also switched to the built-in amplifier. Automatic input selection. Programmable also on I2C bus. The first firmware release (in 2012) there were about 35, but in the second output (in 2015) they were released under 170 pieces. and they were virtually all existing permissions.
- new Chinese invention - controller with eDP output on RTD2556H chip:
VS-RTD2556H-V1.pdf(207.08 KB)
There is still little information on it, so I brought the documentation for a version with one HDMI entrance. There are also versions with one VGA input and with three (HDMI, VGA, AV) inputs. But one thing can be approved by reliably - the prices of them are so far very democratic and, I hope that the same remains. For example, a version with one HDMI entrance is in China about 75 yuan. And this means that with all the postal costs, this controller will increase in the amount of 1200 rubles (at the rate of 1 half of 2017). By the way, here is already on Aliexpress the ready version of the monitor controller (no sound):


Controller with digital and analog inputs, and TV tuners on board
How do they differ from the above:
- of course, the tuner. Those. Controllers data represent a complete replacement for any TV offal
- Firmware poured through a USB port, not using tricks with various programmers
- in the mode of operation of any of the external inputs data controllers behave not as MONITOR - if there is time to fall asleep when there is no input signal, it is estimated minutes. From 1 to 5 minutes.
They can be easily divided into two groups:



All of them have quite good multimedia content using USB input with the exception of controllers on TSUM V29 chips.

Since the creation of this topic in the assortment of the Chinese industry, something fresh is constantly appearing, sometimes even replacing the old one. I basically listed them all. But the first detailed report on the successful launch of the universal controller on Android with digital television (it is not quite clear which one) on board providedFDS UA here . Information deserves attention, since many have long been concerned about the availability of such a controller. Meanwhile, the parallel topic about it is gaining momentum. here .

ATTENTION!!! I want to warn you that in all "television" controllers the HDMI input is purely multimedia and is designed to work with sources either HD Ready (720p) or FullHD (1080i). Those. if you expect to use it with a matrix other than 1366 * 768, 1280 * 720 and 1920 * 1080 and connect a computer via HDMI, then you may have problems. What is meant:
If (for example) a 1280 * 1024 matrix is ​​connected to the controller, then when applied to the HDMI inputFullHD and HD Ready signals Everything must work and the horizontal signal should turn around to the matrix horizontal. Those. We apply 1920 pixels on the input horizontal, and the controller displays 1280 pixes to the matrix (from Example). What will remain on the vertical - everything is proportional, i.e. There will be top and bottom to black stripes. The scaler is responsible for this process (in the correct sense of the word) controller. But all this is true for video content. If you are talking about connecting to an HDMI computer, then everything will be not so simple. After all, we (from the example) matrix 1280 * 1024 and to get the most clear image, the signal from the computer should also be the same. In fact, the computer's graphics adapter such a resolution when this controller is connected will not be issued. That's what I meant, calling the HDMI input "Multimedia". To use the HDMI input of the controller in conjunction with the computer, then without the "Bubn" earn except the matrix 1920 * 1080 and 1366 * 768. For a constructor collected on matrices with another resolution, at best you will have to pack the desired resolution using the capabilities of the video card, at worst (if the video card does not know how - for example, the GMA graphics from Intel) - refuse to use as a computer. For more information on this issue, you can find in the second part of the hat under the EDID spoiler.

If you look closely at the photographs of the devices, it is easy to see that all of them are united by the presence of a 2-row 2x15 pin connector with a 2mm pitch for the LVDS interface.In all the above-mentioned controllers, they are completely identical in size and pinout (well, except that some controllers may not have one pin from the power supply (1-3 pins) or masses (4-6 pins), which by the way does not matter. This is how its description looks in manuals for any universal controller:
However, many have not been able to understand this table, so I give a picture of how it looks in kind:
If the situation does not clear up after this either - take good advice: drop this case, it is not up to you.
ATTENTION!!!
For those who are not quite familiar with the concepts of LVDS interface, foster attention: higher under the spoiler it was said that in the interface of the matrix there are pairs of RX signals, and in the controller interface we see the TX pair. Literally this means: Rx (Receive) - "Take", TX (TRANSMIT) - "Transmit". The matrices are "receivers" LVDS signals from controllers, which are essentially their "transmitters". Therefore, the TX outputs of the controller are connected directly to the corresponding inputs of the RX matrix. Those. TXO 0- with RXO 0-, TXO 0+ with RXO 0+, etc. It should be in mind.
One more thing. As it became clear from the previous spoiler, the matrices are of different channels and different bit depths. The controller connector is one and "stuffed" for maximum 8 bits 2 channels. But where and how to connect the matrix, in which the other bitness and one channel, will have to be shown "on the fingers", because ... There is no better to keep silent:
What bit size and which pins to display is the “work” of the firmware and this does not need to be clogged.
The same table will be mentioned in one more place - about the replacement of connectors. I recommend reading even to those who do not want to read ...
The backlight control connectors of all universal controllers are also completely similar, but on some (for example, Realtek RTD2662), there is no leg for adjusting the backlight brightness. Those. The algorithm for controlling the backlight brightness is simply not implemented. By the way, there are generally strange things - the NT68676-2A controller has a backlight brightness adjustment knob and all its circuits are unsoldered, but the adjustment itself is not implemented either in the interface or in the service menu.
Manuals on different TV controllers and firmware files are below the spoiler "Files for download." Therefore, do not ask each time in the topic about pinout connectors, firmware, etc. Take the trouble to find it yourself. You will succeed.
Another and quite serious point. I ask never to write off.
Matrices have orientation. Those. It matters its horizontal side, which is a "riding" of the image. Some tablet matrices (you can even say almost half of their number of models) have the ability to turn the image of the internal electronics by the matrix itself. To do this, there are pins in their interface responsible for reflection of the image horizontally and vertical. Absolutely all matrices from monitors and laptops (netbooks) have such an opportunity. Accordingly, if such matrices were put into some small TVs, they do not cease to be monitor. Large television matrices have such an opportunity, but not all. You can even say that a very small percentage. Why I said all this now. Often (and this is usually happening when the TV is rewritten) the matrix is ​​impossible to expand due to the fact that the body is so arranged or the fasteners are only in certain positions. Therefore, in any author's projects it is impossible to do without a coup of the image. And if this possibility is not provided in the matrix, it can save only a coup of the image by the controller itself. But this opportunity is only at universal controllers with television on board. All the above-mentioned monitor controllers are not a coup. No image. So, when planning your project, please pay attention to this factor.
Yes, I would also like to focus on such a moment - all, buying the above controller, for some reason they forget that it needs something else. And if for a person who is confidently holding a soldering iron, this is not a problem, then for a newcomer - what else. These are buttons and a remote eye. The Chinese sell about such kits costing order.2$
For those who will independently produce a board with buttons, you will find "useful tips" under the spoiler.

Well, perhaps, even finally, a small remark. For all matrices that apply (or applied) in tablets and laptops, there are no controllers that would have only one low-frequency analog video input. So if someone has a desire to make a monitor for an analog video camera (well, for example for the intercom), it will be necessary to consider controllers that have such a video input as "incursions" to VGA or to some other digital. For example, in the first picture, or such a plan:

Now about one of the most difficult moments - about the backlight. Without going into the types of matrices, I will tell you about what kind of lighting happens in general and what it imposes its imprint on. To begin with, it should be clarified that the LCD matrix is ​​a kind of "sandwich", which includes several components:
- matrix frame at the bottom of which the reflector is located
- backlight system
- light diffusers (diffusers)
- LCD panel
The LCD panel itself is the most "valuable" in the matrix - an image is formed on it. However, the LCD panel is not a source of light - the image is formed ONLY by passing light through. And to see this image - and there is a backlight, which is placed exactly behind the LCD panel (glass matrix). By design, the backlight is:
- Edge - this is when there is a light guide from plexiglas (7-10mm thick) and light sources shine in the end of this light guide. The light guide has embossing in the form of matte points. These points are involved in the reflection of light from perpendicularly running light rays. Their intensity and diameter are carefully calculated and the density of their location depends precisely from which the faces there is light to cover the entire area of ​​the matrix as uniform. At the same time, the light guide is a substrate under the glass of the matrix.
Light diffusers (or diffusers) are located between the LCD panel and the light guide. Their purpose is to provide the most uniform illumination and its intensity over the entire area behind the LCD panel. They are a set of plastic sheets with a specially treated surface - there are both optically distorting and matte sheets.
- Direct - in this case there is no fiber, but the light source itself is located exactly behind the LCD panel. The light source in this case cannot be one - there are several of them, but they are not so much so that they are evenly shone throughout the matrix area. For example, lamps are located at some distance from each other and if not to take measures, the dark and light stripes will be visible in the image. If these are LEDs, stains will be visible. In this regard, the diffusers in such illumination differ fundamentally from the one that is used in EDGE. The first is the sheet of plexiglass (depending on the diagonal of its thickness 2.5-4mm). It performs two functions - support for the glass of the matrix (after all, in fact, the matrix of the matrix is ​​deep and the glass matrix is ​​based on it) and in connection with a specially treated surface - the optical "folding" spots from light sources. Then also lies a layer of diffusers and then all this is covered with the glass of the matrix.
Diffusers usually have the number of plastic sheets 3-4, but it should be noted that only all together, COMPLEX IN CERTAIN ORDER (!!!), they provide a uniform illumination of the entire area of ​​the matrix. Therefore, in the case of opening the matrix (for any manipulations, for example, for repairing or replacing the backlight), measures should be taken to ensure that these sheets lie in the same sequence and the same side (each !!!).
By the type of light source used, the backlight is CCFL (fluorescent) and LED (LED).
CCFL .
In the matrices of monitors, laptops and old tablets, only Edge backlighting is used.
In matrices less than 9 inches, L-shaped lamps were used, which were located along one long and one short side of the matrix. In the matrix of more than 9 - 10 inches used straight lamps. The lamp is located along one long side from the bottom. Very rarely, in the notebook matrices of exclusive models, a two-lamp design was used, while in order to preserve the small matrix thickness, the lamps were positioned side by side, but one after the other (towards the fiber end).
Two or four-particle design was used in monitors matrices. The lamps one or pairs were located along two long sides of the matrix (from above and below). As usual, there are rare exceptions: for example, in some monitors, the diagonal 20 "encounters the CLAA201WA03 matrix with a 6-lamp Edge CCFL backlight - along each of the long sides there are already 3 lamps in it. In the diagonal monitors (above 24") and TVs used multilapampous Direct backlight from direct or U-shaped lamps. These matrices (understandable business) differed great thick.
LED
In the matrices of tablets, netbooks, laptops, monitors and televisions up to 55-58 inches, an edge backlight is used in the form of one band of the LEDs along the lower edge (there are large diagonally from television matrices). LEDs are located in one line tightly to each other, but their inclusion circuit can have the most diverse configuration. It all depends on how many LEDs are combined into a group (called string, the LEDs are included in series) and how many such groups (number of studges) in this strip. The total number of LEDs in such a strip happens from 20-25 in small diagonals to hundreds in large.
A number of television matrices of diagonals from 28 "and are more executed with the Direct backlight. Thongs (they are true are fully fulfilled, rather than in EDGE backlight - for several LEDs on separate bands) are located behind the matrix and several of them. The LEDs used there are fundamentally different from those which are used in EDGE backlight - they are much more powerful, with wide-angle diffusion of light and are not even installed, but at a considerable distance (relative to their linear dimensions) from each other. The total amount compared to EDGE backlit is completely small - in the area of ​​20-50 pieces (also depends on the diagonal).

For what I wrote all this.
The fact is that the output of the backlight control on the controller is represented by 4 contacts:
+ 12V
Enable (enable the backlight level + 3.3V in many controllers, but in television - + 5V)
Adjustment (backlight brightness control, analog output - controlled by some voltage level. In most controllers from 0 to + 3.3V, in television V29-V59 - from 0 to + 5V, in television controller z.vst3463 - made according to the scheme "open collector")
GND
Looking at this list, it is difficult to imagine how to turn on the lamps or LED strips into this connector. In matrices with LED backlit from tablets (not budget) or laptops, the backlight driver is already built into the Matrix electronics and there are contacts in the interface to control the backlight - they are connected without any imaging to this connector (above mentioned about the backlight design in these matrices - It was about the design of the lighting elements of the backlight, and not about the built-in driver). And what about lamps? Or with thongs that need 4-6 lines with a voltage of 30, 50 or 85 volts, and even with current stabilization, and even the necessary power? So in order for such a matrix to light up, you need to apply the backlight driver.
For matrices with CCFL backlighting, there are universal drivers for 1, 2 or 4 lamps.
Those. You can find out how many lamps are in the matrix (this can be seen even by the number of connectors leaving it), buy a suitable driver (they are quite inexpensive) and connect it directly to the backlight control connector of the controller. The cord that comes with it in the kit, just for this purpose and intended. Such a driver is worth about 4-5 dollars, look for ibei or aliexpress by tagCCFL backlight inverter.
However, sometimes the status of the lamp matrix leaves much to be desired. As a rule, in laptop matrices, the lamps have long been sitting. The output is - you can change the lamp lighting onto LED, universal kits for which the Chinese also sell. Just should determine what exactly to take. For laptop (thin) matrices, you need to take a set with a 2mm wide strip. For matrices from monitors, you need to take stripes with a width of 4mm. Depending on whether lamps have been located along how many faces, you should select a kit with one or two LED strips.
I want to note the fact that the design of the EDGE backlight involves the appropriate replacement. Those. If the lamp was only one hand, then the LED backlight must be selected with one strip and put it from the same side. If the backlight was with two faces (no matter how many lamps - one or two each line), then, accordingly, you need to put LED backlight on both these faces. It is extremely recommended to put on only one face (for example in order to save). Light guides used in single-vapor lights differ very different from those that are used in the backlight systems from two faces.
Another pretty important point: in "Luster" (let's call the P-shaped lamp holder) only one LED strip is put, regardless of how much lamps stood in it !!! Exactly. Do not attempt to replace each lamp of a separate strip - except for the structural hemorrhoids you do not get anything else. I explain why. First of all, the width of the "chandeliers" is about 7mm (equal to the thickness of the fiber), and by virtue of the structural and technological and optical features of the bottom of this "chandelier" has a smooth surface of the width of usually 5-6mm. The replacement band has a width of 3.5-4mm and the two bands are correct simply not installed (and light from the LEDs should fall on the end surface of the fiber at right angles). Two strips next to except at an angle will not be lying. For clarity, I give a drawing about what has been said:
Secondly, when the strip falls in the middle of the "chandeliers", the light drops exactly in the middle of the thickness of the fiber. If you shift the strip closer to the edge, which the glass of the matrix lies, then through the matrix you will clearly see bright stains from each LED. Third power consumption. I speak applied to matrices from monitors. On average, the fluorescent illumination of 4 lamps, depending on the diagonal consumes the order of 30-40W. By installing one lane, you will receive a 15-20 t consumption, and if you set two strips, then the gain in power will be zero. In addition, LEDs, unlike lamps, are still tangible. Two stripes in the chandelier will mercilessly warm the light guide and the matrix, because it will simply have nowhere to get warm. The same band confidently falls on a smooth bottom, which carries out heat removal. Fourth - an extra driver ... Why? Well, the last argument, which, I suppose, should forever close this question: the brightness of one LED strip is several times higher than the brightness of the pair of lamps, look at it is very difficult because of blinding. The lamps did not differ.
The length of the LED strips can be chosen for a specific size, and if not, you can take a little more. Due to the fact that these kits are universal and LEDsunitedin groups of 3pcs, then these groups can be cut off. The length of such a "piece" is about 16mm. Those. for a matrix such as 15.4 inches, you can safely take strips for 17 inches and cut off the excess multiple of 3 LEDs. Tag for search on ebay or aliexpress -LED backlight strip kit. The driver that comes with these kits is also designed to connect to the controller directly. A little more about the driver and this backlighthere.
I also want to stop at one moment. Many by ignorance (and maybe because of the economy) are trying to plan to put interior LED ribbons in the end illumination (on the adhesive basis 60 LEDs per meter). I want to warn that they are absolutely not suitable for this purpose !!! Do not try to deceive yourself - besides empty work, you will not get anything good from this venture. First, you will get clearly pronounced light spots and you cannot overcome any constructive triggers. The reason is simple - the distances between the LEDs on such a tape are equal to the size of two LEDs. And secondly, these tapes are not calculated to produce the desired shade (cold white with an admixture of pink / purple) to obtain the correct color scheme of the matrix.
Lamps in designs with Direct CCFL backlighting can not be so easily powered. Only one way is possible - to use the regular driver for the backlight, which came with this matrix in the apparatus where it stood. There is another cardinal way that is described.here. One of the ways of such an embodiment, I explained a littlehere, and an example of such an embodiment (from myself) -here.

For matrices with LED backlight (with the exception of matrices from laptops and expensive tablets) the situation is a little worse - without a regular driver "raise" the backlight will be pretty hard. Those. Or search for a backlight driver for this matrix, or also in China to buy universal. True, it is constructively made of course worse than the native - that's the "universal". The fact is that the "native" driver issues (for example, if in the backlight 4 thong matrix) - four stabilized channels for thongs, and Chinese usually provides them with parallel inclusion. It will work, but it's still not entirely right.
There is one more way - to replace the backlight with a universal LED, which I mentioned above, but this is still necessary only in extreme cases, for example, if the LED backlight in the matrix is ​​partially faulty (by the way, this is quite common).

A more detailed description of the control signals for the inverters and backlight drivers is located below the spoiler on the power of the controllers.

So, here we will talk about one of the most difficult moments - the power supply of universal controllers from the "native" power supplies of future designs. One clarification is about the power of universal television controllers, since things are much easier with monitor controllers. But even if it becomes necessary to resolve the issue of nutrition and those, then after reading this material, I think everything will be clear and all questions will disappear. Let's try to cover all occasions.
Before proceeding to specific examples, it is necessary to understand how the internal supply circuits of the controller are arranged and it will depend on whether this is a simple task or very heavy.
Here is a block diagram of the power supply circuits of any of the V29-V59 universal TV controllers:
As you can see, there are two voltage converters in the controller - the first on the DC-DC buck converter (+ 12 / + 5) and the second on the linear regulator (+ 5 / + 3.3). I have a DC-DC converter ZTP7193i on the V59 board. The linear stabilizer chip is some kind of noname, but initially - it’s still a brainchild of Linear Technology LT1084 (if you are interested, you can search for the datasheet just for it). Here is where these nodes are located on the board (example on V59, but also on V29, V56 is about the same):
The power supply circuits of the z.vst3463 controller are slightly different and look like this:
and here is where these nodes are on the board itself:
In this controller, the power supply of 3.3V is somewhat different: instead of one powerful linear LT1084 stabilizer, the low-power AMS1117 is used to power the controller chip, and the power supply of 3.3V to the matrix is ​​removed from the 5-volt source and reduced due to the voltage drop on the transitions of two diodes in direct power on. Also introduced node power off controller nodes + 5V. For example, now the USB connector, power amplifier and backlight control circuits in sleep mode are de-energized.
ATTENTION!!! Before you run your design on this controller, I STRONGLY recommend that you look under the spoiler of "good advice" and carefully read the power-up item highlighted in red. The same applies to the designs on the 3663 controller.

As can be seen from the diagrams, the controllers can be safely powered not only from + 12V, but also from + 5V. Of course, provided that we do not need + 12V (for example, to power the matrix and (or) the backlight inverter). But even if it is needed, both of these power supplies can be spread and below I will give an example of how.
There is an opinion that it is possible to apply + 5V to the + 12V connector and not to think much. Before doing this, I would advise to think, nevertheless, and here's why. I quote the DC-DC converter circuit (a piece of datasheet and do not pay attention to the indicated voltages):
Arrows, I showed the passage of current through the chip. It is the passage, not the conversion (which is generally the main task of this chip). But the internal circuit of this chip:
So this is how the internal key works when the microchip is not in the mode - this is the question. In a word, whoever decides on this is only at your own peril and risk. I am personally not sure about the long-term use of the chip in this state. But even when she takes off, it's okay too - you can drop it off and serve + 5V directly - you can decide that already.

Another, in my opinion, very important point: the power supply must be obviously serviceable. I think comments are unnecessary here, since it is not difficult to guess what the launch of a controller with a faulty power supply will result in. It can be checked offline (of course with the backlight connected). To do this, you need several automotive light bulbs that are not powerful, for example, from dimensions, and then you need to connect loads and simulate switching. How? Read below - and it will become clear. Well, for examplethistopic to help.

I think, before moving on to the realization of connection diagrams, one should rather chew one rather important point: the difference between the power supplies of monitors and TVs.
In the overwhelming majority of monitors, the power supply when the cord is turned on to the network is always in active mode. Those. "Sleep" monitor mode only applies to the controller and backlight, but not to the power supply - it is always in operation. The "Awakening" signal in the monitors only serves to turn on the backlight - i.e. Only one bus (BLON, BKLT, INV_ON, BKLT_EN, BL_EN and similar abbreviations), which runs the inverter (driver) of the backlight. Summarizing said: The monitor power supply is always in operating condition, and the on / off of the monitor is the result of manifesting the status of the backlight circuit (it works accordingly / does not work) and this scheme is only controlled by one bus. The exceptions make up large monitors with CCFL backlight - they are more like a television BP.
BP TVs differ significantly from the BP monitors. The fact is that in relation to monitors, the difference in power consumption between the duty regime and the working is small - for example, 3W and 40W, respectively. Those. The 40-watt BP in the duty state is calmly working at idle and it does not go beyond its normal functioning. In the TVs (for example with a lamp backlight), consumption in the standby mode is also 3-5W, and in the working - up to 180W (42 inches). It is clear that to make such a powerful idle source work is technically difficult - it simply will not give out normal stresses and currents for the duty regime. In such BP, a separate low-power power supply is installed for the duty mode (its output is called on boards as VSTB, VSB), and the main (powerful) power supply is switched on in a separate bus. On BP boards, this tire is denoted as on_off (sometimes n_f), S / B and something like that. I repeat - only a powerful power source is activated by this signal, from which the backlight inverter is powered, the power supply chain of the matrix, the power amplifier, etc. In this case, this signal is not responsible for turning on the backlight. It is completely another signal for its inclusion. It is indicated in the same way as I called the paragraph above. What is it done for what. If you look at the time diagram, the backlight turns on after all the transactions associated with the supply of power to all TV nodes will be held. Thus, contemplation is excluded on the screen of these most processes. So, we summarize according to the one: the television BP is "on duty" by one source, and the TV is provided to others. It turns on in contrast to the monitor's BP with two control signals - one includes a powerful power supply, the second - backlight. I hope explained it is intelligible.

We now turn directly to the discussion of wiring diagrams.

1. CONNECTION TO BP MONITORS

Excuse my friends, but without theory in any way. Otherwise, everything will turn into a stupid copying schemes and then unnecessary questions. So try to learn what I'm trying to convey.
To begin with, the familiar scheme from the first spoiler, but with minor amendments, what we should have in the end:
The question may arise: why do you need to do this? The fact is that if a monitor is subject to alteration, it is not difficult to guess that its native power supply is already designed for the operation of its matrix and for ensuring the work of its illumination. There is nothing to be wise with external adapters, there is no need to re-buy individual inverters, and you don’t even need to redo the monitor's design - everything is ready only to replace one controller with another, that's all.
The power supply components are usually built ("built", if it comes to monitors with CCFL illumination) according to the source scheme for two voltages - + 5V and also "some". Why "What"? Yes, because it can be from 13 to 22V - it all depends on what kind of supply voltage is designed on the same card inverter of the lighting of lamps or LEDs. In addition, it is not a fact that this is the second voltage enters the controller - most often there is nothing to do there. The controller is usually powered by + 5V, and the matrix - or from + 5V or from + 3.3V. In the latter case, the stabilizer + 3.3V may be either on the controller board or on the BP board, but it does not change the essence - the primary for it is all the same + 5V. But it's not the main thing. The source + 5V is pre-first, i.e. It can issue a current up to 2.5-3A, and secondly, it is the only power supply on which feedback is carried out to stabilize the voltage. And if it remains unloaded, the probability of issuing unstable voltage on the power supply of the inverter is large. Those. Without a load on a 5-volt source, the voltage at the output of the second source at the specified value of 13V (for example) can vary from 13 to 19V. In addition, the backlight power supply is calculated for maximum current 1-1,5A. And if so, then 1-1,2A "eats" the inverter and the power of the controller (if the controller is powered from it) will remain anything. As a result, we simply get a "squid" in the form of triggering protection of PBC from overload. To this not happen, you need to power the controller from the source + 5V. How can I do that? Very simple -apply + 5V from the power supply to the pin of the 5V matrix power selector jumper, which sits on the + 5V bus. What do I mean by this phrase: if the matrix is ​​powered by 3.3V, then it will be pumped straight to the pin or an electrolytic condenser in the output of the built-in DC-DC converter, and if the 5-volt matrix (and this pin will simply be Busy jumper) - then only to the condenser. Or see where the tire is in the controller + 5V and select a convenient place where you can fall to it. For V29-V59 controllers, this is a stabilizer 1084, an amplifier power resistor, etc. This tire is easy enough to bring up the location of the legs of the stabilizer 1084 (the picture for V59 !!! for others - you have to search):
For the Z.VST3463 board, the + 5V bus can be found according to the above mentioned section of the board - this is the throttle pad (where the + 5V bus arrow is stuck).

However, in this connection there is one caveat: we feed + 5V to pin 3 (via a throttle, see the diagram above) of a completely unpowered DC-DC converter chip. Judging by the above internal scheme, nothing terrible should happen. But I still strongly recommend the throttle at the output of the built-in converter to unsolder away from sin. And this is about the V29-V59. But in relation to 3463 or 3663, it is probably necessary to unsolder it. There are cases of "loss" of controllers among forumtsev ...
I think no one will have problems with this connection, because the + 12V source in the designers of the monitors is not needed in any part of the circuit.

Sometimes, it is very rare monitors with a single source. For example, Samsung Syncmaster 225 - he has only one source and he is + 13V. Then of course without options - connect directly to the + 12V connector of the controller. The converter withstands input voltage up to 20V.
And, of course, one cannot say that all of the above applies to absolutely all monitors - of course, there are also non-standard solutions in the circuitry of monitors. In this case, you will need to understand in more detail.
As for the control signals in the power supply unit of the monitors, there are usually only two of them - turning on the backlight and controlling the brightness. Below I will summarize this information in a separate paragraph.

ATTENTION!!! All of the above relates to the fact that I was holding. Those. to the boards, where the feeding portion is expressed as clearly shown in the above schemes. This V29-V59, zvst3463, D3663. As for more recent versions of controllers, it is necessary to carefully examine their scheme for such a possibility. For example, I can say with confidence that the controller Qt526 thus only powered from + 5V will not succeed because the supply chain made it quite differently. Here It is an example of how this has been energized by the controller monitor PSU, on which there are two supply voltage + 12V and + 5V. Moreover, high-current + 5V and + 12V - low current and the entire controller to run from one + 12V could not be articulated by me above reason. Other controllers, deserve attention, I did not - so I can not speak for all manufactured in China glands indiscriminately.
I must say that use native PSU of reworks display certainly much preferable to seek some kind of external adapter. But it is not always easy to do so. Who has the desire - it is the theme, go to the questions will be solved.

2. CONNECTION TO PSU TVS

Why PSU? Because, in addition to the power supply, there can be quite a lot of other nodes on the board, so the TV power units are called the Power Supply Unit. I must say that the second board (often called the Main Board) is correctly called SSB - the Small Signal Board, and this has a definite meaning. The lion's share of the electricity eaten comes from the backlight, and everything else is trifle, and relying on some significant energy consumption by the controller is not worth it ...
Again the theory - without it, alas, no way ...
It is rather difficult to cover all the constructions of television units, therefore I will give the most common ones. Structural Schemes Again. I’ll just make a reservation that I selected the most significant nodes and if anyone decides to let me know that I forgot to draw the circuits of rectifiers and the correctors of the power factor - believe me, I remember that very well. Just the vast majority will not need it.
Well, at the very beginning I will make a note about their difference from the power supply of the monitors: on TVs (diagonals over 24 ") the power supply has a real standby mode. That is," something "in them in sleep mode" sleeps "and only works active mode. What "sleeps" and how - read below.

but). PSU TV on CCFL lamps
As can be seen from the figure, these are the power units that are used in the TVs with a lamp light, the inverter of which is present as a separate board attached on the back of the matrix. A distinctive feature of them is a rather powerful transformer from which the inverter power supply voltages (+ 24V) are removed, in the overwhelming majority of 12V (to power almost all SSB nodes and in the presence of such a power of the logical part of the matrix). It is possible on it more sources for powering sound amplifiers. They are just not interesting to us and I intentionally did not even draw them, because they can not load them in work and it will not affect anything. The duty mode is implemented as a separate low-power source, which always works if the cord is inserted into the outlet. VSTB voltage I intentionally did not indicate because it can meet with a dimension of 3.3V, 3.5V, 4.7V, 5V and it seems even 6B (perhaps in the last I am mistaken, but it could be in very old models). The main power supply (+24, + 12V) works only in active mode, i.e. There are no these stresses in the television mode.
The aforementioned illumination inverter (so-called. Stadalone, the one that hangs on the back of the matrix) has another one, in my opinion a wonderful feature. Almost all 100% of such inverters feed on 24V voltage (in connection with which if the matrices are similar in size, they are interchangeable, I'm talking about the inverter with other characteristics), but this is not the main thing - almost all such inverters have two brightness adjustment entries - PWM (PDIM) and analog (ADIM), which switches between these Special Pin modes. As a rule, ADIM is not connected, but it can always be activated by rearning the wire in the inverter connector and taking measures to switch the inverter to the ADIM mode (in this case it turns on its built-in PWM, which is controlled by the adjustment voltage). So this inverter is easiest to be adapted to adjust the brightness from the universal controller.

b). PSU TV on EEFL lamps
In principle, the differences from the previous PSU there are significant, although this does not affect the project on the basis of a universal controller. The main difference is the matrix there is no separate inverter and it is on the PSU board. Why is that? Because EEFL lamps (External Electrode Fluorescent Lamp) have such a design that they can be connected in parallel and the entire node of the lamp lighting is removed from the matrix with two wires. Accordingly, there is no need in a multi-edged inverter (as in the case of CCFL lamps). All other nuances mentioned above for the previous PSU are fair and for this PSU.

c) .PSU TV with LED backlight
If in the two previous cases, the circuit solutions of the vast majority of TVs did not differ much, then in the case of LED backlighting, PSU circuitry often harbors unexpected surprises in the form of non-standard circuit solutions of developers of a particular brand. Therefore, I present a picture of a strongly averaged average PSU for LED backlighting. Further it will be clear (I hope) that the main thing is not the circuitry, but all the same backlight control signals and supply voltages.
I think that there is nothing special to comment on - almost everything that was said above.
For what I gave these three schemes. To make it clear how the backlight is controlled and how the power circuits are organized, which we will need later to connect our controller.

And there is nothing more to say ... Anyone who has carefully studied everything that was above under this spoiler can figure out how to connect what and where. But in one, I think, I have to help - to unite everything into one “universal” scheme and it looks like this:
Here I have implemented the most successful solution for powering the universal controller, which can also be called universal for almost all PSUs. On the left of the diagram, there is a plot of any television PSU, on the right - a controller (the diagram that was above). Now I explain what elements are installed and why, and how everything works.
So, the key point is washing the controller from the PSU duty voltage. This scheme is suitable for cases when the supply voltage of the duty regime is 5V and it was stated as mentioned above in the paragraph of power from the monitor BP (but only in contrast to the example through a diode). How it works. + 5V pass through the diode and it falls about 0.3-0.5V (it is better to apply with a smaller drop of 1N5817). Thus, on the 5-volt bus of the controller approximately 4.5-4.7V. This voltage is quite enough for the linear stabilizer to give out confident + 3. prov in its output to power the controller chip. The fact is that the voltage supplied to his input should be higher than 3.3V at least 1.1V (i.e. not less than 4.4V). In standby mode, the controller consumes no more than 100mA and this is the load of the PSU power supply unit. When we are "awakening" the controller, it gives the command to output ON / OFF the "Invertor" connector for switching on (...). I did not say to turn on what exactly. And all because if you carefully look at the PSU diagrams of the diagrams, you will notice that they have separate pins to start the main BP and to start the backlight. In our case, we simply connect them together and the controller will successfully launch both.
So, the controller submitted a signal to turn on the backlight and the main BP. At the same time, with the main BP, the + 12V voltage appears, which is connected to a complete power connector of the controller. Next, it is converted to + 5V and it in turn is applied to the diode cathode. In the "right" power blocks when submitting the POWER ON resolution voltage, the voltage at the outlet of the duty unit is slightly reduced (so conceived in the diagram), the voltage on the anode becomes below 5V and the diode is closed, the duty from duty BP is turned off and the controller continues to be powered by + 12V, Going with PSU. It must be said that not all duty duty blocks are absolutely "honest." It is likely that the scheme does not provide a decrease in the voltage of the duty unit in the active mode of the main one. There is no problem even if the diode is not closed, it will not arise. At least the controller will definitely be securely driven from the built-in DC / DC converter, and the BP on duty does not threaten this situation. When switching the controller in the waiting mode, everything occurs in the reverse order. This method is no longer the theory, but a proven and 100% workable (I applied it in my last project, you can find under the spoiler of finished designs).
For cases when the voltage of the on-demand power supply unit PSU is less than + 5V (meaning +3.3, + 3.5V), it will be necessary to apply a circuit withboosting DC-DC boost converter. Converter not at all scarce and completely capricious and stable. Then the connection scheme will look like this:
This scheme allows you to connect the controller to the "standby" voltage of any size from 3 to 6 volts. At the heart of the same theory with the diode, of which I spoke above - it will close when it appears from PSU + 12V and the boost converter remains unloaded - it is not scary for it. The magnitude of the voltage converter installed at the output (+ 10V) was chosen for reasons of a diode drop and a sure 12/5 V DC-DC converter on the controller board. It is not critical and can be from 8.5 to 11V.
Since the connection by the second scheme does not require interference with the diagram of the controller itself, then this option can be recommended to those who are afraid to climb the controller card with a soldering iron and in general to spoil the freight type of controller. Just the second scheme pulls the mandatory acquisition of the converter. By the way, it is not necessary to buy it in China at all - they are and we have the cost of it about 150 rubles (more expensive than in China, but not wait a month ...) ... ATTENTION !!! Configure the voltage of the boost converter to avoid troubles better in advance. This can be done even without load. And then connect to the circuit.
It should be noted that there are additional resistors on these two circuits. Their purpose will be described below in the paragraph about adjustments. However, this scheme applies only to controllers on the V29-V56. For the Z.VST3463 controller, both circuits need to be slightly changed, and the location for connecting additional resistors will look like this:
Those. There will be only one resistor and is connected between On / Off and Adjust. Why it is this way - you can also read about it in the same place in the paragraph about adjustments. For the DS3663 controller resistors are not needed at all.
And one moment. As it is not difficult to notice, all three different PSU variants mentioned above contain the duty pressure supply unit (for example, 5VSB). What it is: This is a low-power voltage source to maintain SSB nodes in sleep mode. Of course, the television circuitry varies greatly and the BP of the duty mode can be used for the washing of low-power nodes (for example, the controller Du and Supervisor), and can be powered by the entire logical part of the SSB and in sleep and in active mode - then it is designed to load up to 2-2, 5a. So in order not to guess and not guess which one you have in stock STANDBY, and the above-mentioned laying schemes of universal controllers have been developed. We are talking about diodes in the supply chains and parallel use of the source + 12V, which accurately has sufficient power.

However, it is worth noting another fact - not all 100% of television PSUs can be combed with the same brush - there are also “ambushes”.
1. It so happens that sometimes on TVs a small diagonal of the + 12V source may not be at all and the whole SSB is powered by voltage, for example + 5V. Powering the controller in this case will need to be done on the principle of powering the monitor PSU. How - I hope you figure it out, all the schemes are above.
2. Sometimes there are schemes in which the on-duty power supply unit has a very small capacity - i.e. sometimes it does not even pull the operating mode of the bare controller. There is a way out in this situation. for exampleheredetailed solution, which was successfully implemented by the questioning forum participant.
3. Another option: in the TV's power supply unit there is also a voltage of + 12V (which the controller does not pull) and a working + 12V (from the output of a powerful power supply unit, which comes to life only in the active mode). Then the power supply method is described.here.
The main thing - to decide on the principle. And in particular - do not hesitate to ask questions in the topic. We will understand and make the best decision.

And more ... I practically anywhere did not show wires of "weight" on schemes. This does not mean that they do not exist. They are and their full! They are everywhere. Do not save them.

3. ABOUT POWER SUPPLY OF CONSTRUCTIONS PRODUCED "FROM ZERO".

What is meant: you have a matrix, there is a controller and there is an inverter as separate blocks. In this case, the easiest way is to purchase a 12V power adapter with a power of 40-60W (current 3.5-5A) and launch it. However, 12-volt power supplies are not as widely distributed as, for example, laptop adapters. There is a solution in this case. It is enough to apply a buck-to-dc buck converter (such assuch a). It is put on the power of the controller itself and the voltage of 12V (i.e. it transforms 19 in 12V), and the backlight inverter is just better to put from 19V (of course if it allows it). In this case, it will heat up less than from 12V. Putting the controller itself from 19V I would not advise. Well, if the board uses a reduction converter chip with a permissible voltage of 20-24V. But it all depends on the manufacturer of the controller - it can set another microcircuit with a permissible supply voltage, for example, 18V and then from the ones attached 19V immediately "PCHIKNET" (and do not give God with consequences). In the area of ​​100 pages of the topic, there is a confirmation of this - in the controller, accidentally connected from 20-22V, the DC-DC chip 12-5V flew and dragged another and important DC-DC to 1.2V, which feeds the controller chip.
By the way, those who know how to hold a soldering iron in their hands can safely use an unnecessary power supply from a computer - it will easily be able to issue 12V 5A, and perhaps the fan will not be needed. It can even be launched on the principle of television PSU (after all, in many of them there is also a separate 5-watt power supply system).
You can also buy not necessarily an adapter (in the classical sense, such as in a black plastic case), but an open power supply. I gave a link to a very good BPby thispost
Everything said in the last paragraph, of course, is focused mainly on designs with monitor or notebook matrices. Because it is necessary to invent a “bicycle” on large television matrices with greater care and miscalculation of power.

And of course, the eternal question of almost everyone who decided on the project. Let's even highlight in bold:
POWER SUPPLY SELECTION
In fact, the question is quite simple. I bring him here only for the reason that not everyone here has gathered with radio engineering education or with amateur radio experience.
Sorry, guys, but again the truisms - without them in any way. For those who absolutely no boom boom with electronics, I will explain on fingers.
1. If the power supply unit is written 12V 5A, then it literally means the following: the power supply module outputs a voltage of 12V to the output and at the same time it CAN eject current up to 5A. Does not issue, namelycan give out !!! Those. the controller can be powered from a power supply designed for at least 1000 amperes, but if the controller eats only 1A, then this 1000 A power supply unit will give only 1A to the controller. Who does not understand - re-read again and again until you understand.
2. The values ​​of voltage and current indicated on the power supply determine itsmaximum power. The power (in watts) at the output of the power supply is equal to the voltage (in volts) to the current (in amperes). Since, as mentioned above, the value of the current current depends on the consumption of its load (and on the power blocks, the maximum current value is specified, which can "give" this power supply), then, accordingly, the power calculated by the above method is also the maximum possible value. Those. The 12V 5A power supply can be in the load 60W, but if you connect a controller that consumes only 1a, the power given to the power supply will be equal to 12W.
And now to the choice of power supply unit. To unequivocally answer this question, you need to be patient and find the strength to read the datasheet on your matrix. So, what serves as a calculation data:
a) Power consumption of the logical part of the matrix. In the datasheet, this (VDD Power) is in the Electrical characteristics section. There may be several values ​​(for different backgrounds of the image) - choose the maximum.
b) The power consumption of the backlight. The same section and sub item Backlight Unit. It is called Power consumtion. We look at the value and multiply by 1.1 for LED backlighting or 1.3 for CCFL. The fact is that this power characterizes the net consumption of the backlight, but we need to lay stock on the efficiency of the inverter.
c) Power consumption of the controller itself in the active mode. Accurate data is not available, but the orientation is approximately 5-6W for the V29-V59 and 8-9W for the Z.VST3463.
We summarize the obtained power values ​​and divide by .... If the planned power supply unit is 12V, it means by 12. If the power supply unit from the laptop is 18.5 (here I have laid the efficiency of the DC / DC down-converter, without which the controller cannot be connected). We get the amount of current in amperes.
It so happens that the data are obtained completely "in the grinding". Then you can either rely on the spare durability of the power supply unit, or after starting the whole structure, reduce the backlight brightness in the service menu (so that it does not work at full capacity). That's all science.

In many PSUs, as well as in separate inverters, the control voltage is regulated within 3.0 ... 3.5V. From the output of ON / OFF controller, the voltage control signal + 5V is released. In the controllers V29-V59 from the tire + 5V on the collectors of control transistors are 1kom resistors. So in order for the voltage of the inclusion signal and a maximum backlight brightness signal signal to the level of 3.3 ... 3.4V, the circuits above are additional resistors connected to a mass, resistance of 2.7k. Together with resistors from the controller, simple voltage divisors are obtained.
In the controller z.vst3463 and the inclusion and adjustment made severaldifferently. The inclusion signal of the inverter is implemented as a permanent voltage from the key collector on the P-N-P transistor. It is most likely done with the goal of "Causes" of this exit, since it is rare, but inverters are found, which are low input resistance to the backlight input. For example, PSU from Phillips PFL3606 - it has an input resistance of the control input approximately 1k. Therefore, to bring the inclusion signal to the level of 3.3V, you will need to hang on the output of the controller an external resistive divider (you can also construct it from 1 and 2.7k, 2nd and 5.6k, etc.). About the same looksschemeand the controller D3663LUA. The outputs of the brightness control in z.vst3463 and D3663LUA are made almost the same and this will be discussed below.
All the above is just recommendations and perhaps no dividers are needed at all. At least you can safely connect without them - nothing will burn. And only then draw conclusions - we need dividers or not.
The ON / OFF controller output (or BLON at z.vst3463) connects to the inclusion of the backlight of the monitor or television inverter. It can be called BL_ON, BKLT, ENA, B / L_ENA, ENABLE and similar abbreviations from English words backlight, enable. There will be this connection directly or via a divider (if necessary) - can be solved later. And you can simply connect directly directly. If everything works, then there is no divider. But I want to warn that sometimes the signal control is involved in the stabilization of other PSU stresses. Rarely, but it is found (an example - PSU LG 42LN540 - there is an inclusion voltage + 3.5V is part of the output voltage stabilizer + 24V). Then the divider will need.
The output of the ADJ controller is connected to the input of the inverter brightness regulator. It can be called Dim, P_Dim, Dimmer, B / L-ADJ, BL_DIM, BRI and similar abbreviations from the words dimmer, adjustment, BRIGHTNESS. If there are two A_DIM and P_DIM inputs (or b_dim) in the inverter, such a backlight is found in monitors, you can try to connect first to A_DIM and check whether the backlight brightness is adjusted. If not - then to p_dim. The fact is that A_Dim is a voltage adjustment (analog), and P_dim is an adjustment of PWM, and in some inverters analog adjustment (A_DIM) could not be aligned at all.
All the inscriptions that I called may have to search. If they are not on the power supply, follow the home wires of the monitor or television tripe - maybe they are on the mainboard.
Above (under the spoiler about the backlight), I talked about the ways to adjust the brightness of the backlight in TVs and monitors (!!!) there are several, and more precisely four. PWM Adjustment - When the backlight brightness is regulated by changing the width of the pulses, following with a certain frequency, for example, 200Hz, and analog - when the brightness depends on the values ​​of the applied voltage. These are two fundamentally different ways. And plus each of them can have "direct" and "inverse" adjustments. Direct is when the maximum brightness for analog adjustment at maximum voltage at the ADJ output, and for PWM - the maximum pulse width (actually permanent voltage level), for inverse (reverse) adjustment - the opposite. I would recommend before the first inclusion to experience the existing inverter on the autonomous mode (without connecting the controller) for the study of its adjustment and starting levels. In order not to write all possible options for a long time, I will give an inverter launch algorithm:
I hope that everything is clear here - you just need to carefully study it. In the "ovals" in the picture - the results that we need to understand how the inclusion and adjustment of the backlight.
And I repeat: this algorithm involves the launch of the backlightWITHOUT PARTICIPATION OF THE CONTROLLER !!! For fans of unnecessary initiative, I propose to remove the controller away at the stage of testing the backlight so that it is not tempting to connect the contacts on it.
And another very important point, which I did not say. At the time of testing the backlight, the power supply should of course be in good condition and MUST be in operation. Those. in other words, it should be loaded, for example, with bulbs and the voltage on them should be regular. Those. we in this case simulate the normal operating mode of the PSU.
Starting PSU from TVs (in which the power supply itself and the backlight turn on via separate wires) suggests that the power supply enable and light backlight contacts are simply interconnected - in 99% of cases this is enough.
In addition, in order to avoid difficulties at start-up, which can be adjusted to the backlight, I recommend on the first launch to submit to the Adjust regulator inverter the maximum brightness level (according to what you learned from the above algorithm, and if for your inverter ADJ has a maximum with a filed "+ "And ON / OFF also turns on with the" + "filed, it is enough to simply connect ON / OFF and ADJ on the test inverter together, on the controller, while leaving PIN ADJ in the air). The fact is that most universal controllers have direct analog adjustment at the output. If the inverter has adjustment on the PWM principle, the backlight may not be lit, and you do not determine the reason. That's the first inclusion to do as I said. When you start and set up and adjust - then you can separately address the alignment of the backlight - to connect the adjustment correctly and then already conduct experiments, already knowing that everything else works fine.
Most of the television and monitors inverters are adjustable on the PWM principle, and all universal controllers have an analog principle (because they are focused on working with universal inverters, which are equivalent to the brightness adjustment). Consequently, these are two incompatible factors. Cases are possible when at some short section of the adjusting characteristic, changing the brightness of the backlight on the PWM inverter from the applied voltage still occurs (I talked about it in one of my projects), but it is rather an exception than the rule. In this case, you can either even refuse to adjust the brightness of the backlight, connecting the terminals BLON and PWM of the inverter (setting the brightness of the backlight to the maximum), provided that the PWM adjustment from this inverter is straight (or put a PWM on the ground if the inverse adjustment), Either make a not very complex converter analogue in PWM. This will be said at the end of this spoiler.
We continue the "lesson". If, however, you are "lucky" and your inverter is designed for analog adjustment, then in this case you still need to be diligent in order to achieve a result. What is meant. It should be noted that the controller Z.VST3463 brightness control is not as predictable as that ofV29-v59. And even with a large number of drivers and inverters, it is not directly friendly - installed in the "maximum" it still does not ignite the backlight, which normally ignites if the ADJ inverter is connected to BLON. As shown to check, the brightness adjustment output is made in this way: with 90 Output of the chip via sequentially enabled resistors in 4.7k and 1kom leaves the INVERTOR ADJ PIN. In the entire range of brightness adjustment on it, there is no voltage. The reason is that there is an open collector and this chain is needed to "pull up" through the resistor, for example, 2.22 to the BLON output. You can try to connect according to the scheme that I showedhere. So try and experiment ...
ATTENTION!!!
LIGHT BRIGHTNESS ADJUSTMENT IS NOT OPERATIONAL AND HAPPENS IN ALL UNIVERSAL CONTROLLERS ONLY FROM THE SERVICE MENU.
Adjusting the "brightness" from the remote from the normal menu changes only the brightness of the IMAGE, and not the backlight. Who does not understand - re-read again, so you do not ask questions in the subject.
And, since this is a service adjustment, if the monitor inverter has an inverse dependence on the applied adjusting voltage from the controller, I think it should not be a "blow below the belt" - just in the service menu you will have the backlight level specified in the number 90 or 95 (%), and 5 or 10. It does not affect the speed. I do not know how on other controllers, but for example, on the V29, there is even a PWM Invert point in the service, which can change the adjustment characteristic to the inverse. In general, the soil for activity is ...

In conclusion, I would like to focus more on this "topic." Slightly higher, I said that if the matrix (this is about notebooks or tablets) or the inverter has only a pulsed backlight brightness control (PWM or foreign PWM), and the controller only has an analog adjustment, you can take steps to set the maximum backlight brightness. But if all the same adjustment is necessary, then there are two ways to implement it:
1. Production of a simple electronic unit - the duty ratio with a variable resistor (then its handle is simply located somewhere behind, since it is often not necessary to turn it). The method is quite convenient and easy to implement. For example, on the Internet you can easily find the pattern of adjustment of the duty ratio on the timer NE555. There is a minimum of elements there, their cost even in the most barges radio base will not exceed 100 rubles for everything.
2. Making a simple electronic unit, but a higher level - an ADC (analog-digital converter) on the microcontroller, which will convert the analog brightness control of the backlight from the controller to PWM adjustment for the matrix (or inverter) on-line.HereForumartrm.grinko laid out a great adc on attiny13, for which he thanks a lot.
By the way, as such a node works depending on the input voltage, the photo sensor of the environmental dependence on the ambient light can be used with the same success, a kind of "straight array" can be applied. My experience of making such a car monitor is describedhere. By the way, I did at a PIC controller, and not on AVR and the current owner of this monitor from an array in a wild delight.
Of course, it is possible to manufacture such an end to specialized chips (not on microcontrollers), for example, on the same universal TIMERS NE555 or TL494 (reference to the concept of building ADCs I can give), but it will still be more difficult than on the microcontroller. for instancehereversion of such a scheme fromGrishanenko On a dual comparator. Here embodiment of the same idea with a reworked board, there is even a video from areon . By the way, this scheme works "Inverse" - i.e. The greater the voltage at the analog input, the lower the brightness of the backlight. To make the adjustment "straight", it is enough to change the inputs (conclusions 5 and 6) of the second comparator. Here of life hacking vicdu Chinese module with ready to TL494.
If there are any other options for powering (perhaps at the request of members of the forum) - then I will add, if necessary.
3. There are situations where direct PWM controller output, and analog input from the driver, and even inverted. This is what happens, if the SSB which serves a branded card from the monitor or TV, and the backlight driver - the universal Chinese. In this case, everything is solved quite simply.HereI painted this in more detail.
4. And finally, the rarest case, which nevertheless occurred twice during the life of the subject. There is a need to achieve frequency multiplication PWM dimming.Herelink fromandreyvirus on the way to achieve this objective by using Arduino. More specifically: branded lanes outputs a fixed PWM frequency of 50 Hz (and this frequency change is not possible), and the matrix need 200Hz Sharp (and other frequency it simply ignores). That's the question and decided.

This section is dedicated to the universal controller with a TV and is still in the filling stage. As required, I will add.
Manuals .
Look carefully at your version of the controller.
ND-LA.MV9.P-2.pdf(357.13 KB)

ATTENTION!!! The document in paragraph 6 (IR & KEY schematic) draws the wrong pinout of the control connector. In clause 7.2 (IR and Key Board Connector) the pinout table is painted correctly.
If you collect according to the above scheme, then nothing will burn, but the controller will not work, according to external signs, as if "dead."

T.VST29.02B.pdf(588.18 KB)

T.VST29.03B.pdf(1.64 MB)

TSUx9V5.1Spec.pdf(926.64 KB)

Users guide of V59 TV controller board_V1.1.pdf(2.34 MB)

Manuals for different controllers on RTD2662 (RTD2660)here. thankpashkovv78 .

Firmware
ATTENTION!!! I wrote this not for myself, but for you. Who does not want to read - make an effort over yourself and still read.
- How to flash the controller (we are talking about universal controllers with a TV on board, since the rest of the controllers are sewn only by "special means"):
File Firmware (ATTENTION !!! Not a folder with a file, namely, the file itself with the .bin extension) copy to the root formatted in FAT32 flash drives, in the controller turned off from the network, insert a flash drive with a firmware file (in the appropriate USB connector) and serve me. If the previous firmware installs the controller when the power is applied to the "Enabled" state, the flash drive state appears (according to its light indicator it can be seen) and then the controller LEDs will begin to "distort". If, when powering the controller is installed in the "Off" state - just press the POWER button on the keyboard or on the remote control. The firmware process lasts approximately a minute, the screen does not shine. After the controller's LEDs "were embroiled", the controller can be installed in the "enabled" state (as it happens after the firmware V59) or to the "off" state (z.vst3463) - it all depends on how the firmware itself is made. You need to wait 10-20 seconds and turn off the controller (or rather the TV) from the network. Turn out the USB flash drive and turn on - the V29-V59 are ready for work, and Z.VST3463 also has to go through the initial installation and setting.
If you try to re-flash or the same firmware or different firmware under the same resolution, the controller firmware to the stage will not go. Thus, if in the case of an unsuccessful firmware (well, for example, do not work or crooked work which it functions, firmware under the Russian version should be changed to the English, or something like that) you need to alter again, you must first flash another firmware with another resolution, and then sew on the right.
And last. Different controllers in different ways are "referring to generations and to USB volumes of carriers. It is not worth the firmware of about 1 gigabyte to use a flash drive to 16-32-64, because the flash drives of this volume are usually fresh generations and may not be perceived by controllers. It happens like this: you turn on the controller, the flash drive blinks a couple of times and nothing happens - the firmware does not go. For firmware, slow flash drives are best suited for 2-4 GB. And then there are problems. Most likely the MCU controllers do not "see" USB's internal controllers of carriers of some brands. So for the accuracy of the firmware (as well as for confident launch of the process itself) it is better to have a small choice from different USB media.
- Firmware does not necessarily have to wear the name of your template. Matrices - a thousand, and it does not mean that they were written under each matrix. You need to look for the firmware that is appropriate to your matrix resolution, bit depth, and the number of LVDS channels. Bit depth, and channel encoded in the names of both firmware SI6L, SI8L, DO6L, DO8L. 6 and 8 - is the bit depth, and SI - one channel (single) and DO - two channels (double).
Look at this nonsense:
I could not even suggest that such a question could arise, but since it did appear, I answer Russian and white:FIRMWARE DOES NOT HAVE RELATIONS TO DIAGONAL MATRIX and are in no way connected with it. Once again: ONLY PERMISSION, KANALNOST and VITALITY. Everything!
- If your matrix comes up a lot of firmware, I recommend to use "smart" search. Those. of firmware with suitable parameters primarily try EEPROM array on the same manufacturer. If you have "exotic" and nothing close to resembling not, then why not try a firmware matrix at the same diagonal.
*** At first glance, these my words go against the twisted two lines above. However, here I put in my words a completely different subtext. For example, you have a Matrix with a diagonal of 18.5 inches with a resolution of 1366 * 768. And the firmware for this resolution is a dozen. There are those that are on the same 18.5 "matrices, and those that are on large television matrices with the same resolution. In this case, it is just better to navigate the diagonally, since many small matrices are interchangeable and have similar time characteristics. But television Just despite the resolution, there may be completely different temporary parameters. Hence the logical conclusion ...And before you immediately sew, "punch" on the Internet, is there a matrix, which is mentioned in the name of the firmware. Because in the archives a lot of firmware in general is unclear under that.
- If you do not have the appropriate firmware, choose the firmware to another bit depth, but with the same resolutionand with the same number of channels LVDS . Then in the service menu, the controller can change the setting Map LVDS, to your matrix properly earned. But with the firmware with the desired resolution and bit depth, but for another channel LVDS This trick will not work.
- Firmware has to do with the supply voltage MATRIX. From the word ALL. It has no hands to reach out and to rearrange a jumper - it should be doing yourself.
WHO DIDN'T UNDERSTAND THAT THIS IS WRITTEN IN RUSSIAN - READ IT BEFORE, UNTIL THE MEANLE WRITTEN REACHED !!!
If you see that in the name of the folder with the firmware specified yet and the power supply - it is nothing more than the fruit sorting themselves proshivkopisatelyami firmware.
- Be careful when choosing a firmware to your controller. In the names of folders with firmware indicated, under which chip controller, as far as the buttons are calculated and which chip tuner. Boards on TSUM V29 and V59 controllers can be equipped with Rafael Micro R840 and R842 tuners chips. You can read the name of the tuner on the shallow chip, which is installed near the antenna connector. Ask in the topic which chip tuner is installed on your board meaningless. It is equivalent to this if you ask forum users to guess which tuner on the board that you keep in your hands. For V56 firmware are not separated, as it goes only with R842.
- Firmware for boards for V56 and V59 DIFFERENT, do not try to cheat with yourself - get a "brick" !!! Verified by one of the forum users.
- IMPORTANT!!! On different boards, but based on the same MCU - different firmware !!! Those. For example, if you have a board based on the TSUMV29 microcontroller, but not La.mv9p, and for example t.vst29-03, then the firmware from La.mv9p is impossible to pour into it, despite the same chip. Any fee with the same chip, but even a slightly different topology (in fact, the scheme) must already have exactly its "native" firmware. The thing is that the microcontroller has many I / O ports through which its connection with the periphery is carried out. So on the same board, the specific leg of the chip can perform one task, and in another board - a completely different, not even related. The algorithm for the work of all ports is just described in the firmware and of course depends on the developed scheme and wiring of the entire board.

Now do the firmware:

- LA.MV9P controller on TSUMV59
LAMV59_R840_5keys_part1.7z(106.62 MB)

LAMV59_R840_5keys_part2.7z(103.33 MB)

LAMV59_R840_7keys_part1.7z(106.82 MB)

LAMV59_R840_7keys_part2.7z(103.31 MB)

LAMV59_R842_part1.7z(122.18 MB)

LAMV59_R842_part2.7z(121.36 MB)


- LA.MV9P Controller on TSUMV29
FW_LAMV29_universal_TV_board.7z(3.04 MB)


- T.VST29.03 controller onTSUMV29 The names mentioned firmware V59. It will work on both MCU (they are interchangeable and have the same architecture). The main thing - the firmware for the chip tuner R620.
VST29.03.7z(14.14 MB)


- LA.MV56U controller (maybe LA.MV56U.A maybe - but not sure) on TSUMV56
LAMV56_KEY5.7z(45.58 MB)

LAMV56_KEY7_part1.7z(73.54 MB)

LAMV56_KEY7_part2.7z(72.17 MB)

I'm not sure that the difference in firmware under 5 or 7 buttons is fundamental, because the service menu contains the choice of the number of buttons. However, KEY7 firmware is much larger than KEY5. Therefore, it is more likely that the firmware will be found under the required matrix in KEY7. And select the required number of keyboard buttons through the service.

- The LA.MV56R.A controller is also on the TSUMV56.
It differs from the previous one - and consequently also their firmware for it.
Hereappearance and firmware.

- The controller on the chip TSUMV56RUUL-Z1 is called either on the name of the chip (the same), or LLV56 (the latter is not sure at 100%, but it seems so)
manualon tv_vst56 - TSUMV56RUU-Z1
firmware on "a cloud" (thankWest @ ), A lot of them out there, but they are only for this controller. Here are DIM4ELA here link gave.

- controller z.vst.3463.a1
manualhere
firmwarehere.
Another file with the manual, the entrance to the service description and firmware 20172017-3-2 3463A Firmware update.zip(39.79 MB)

Linkeven on someone's archive with firmware. Perhaps the same ...
Archivefromvoron12005 Basically, all the firmware are the same, but there are 1024 * 600, 1280 * 768 and 1280 * 800 fresh (2017). 1280 * 800 - 6-bit.

- Controller DS3663 LUA.A8.2PA
Firmwarehere. Manual.
Herealso someone has saved up the firmware ... Andherestill...
By the way there is a version of the controller DS3663 LUA.A8.1-A. You can rename the firmware of the above (for detailshere). There you will find a link to the recovery dump.
Link to the cloud with firmware for DS3663LUA controller. A81.2.PAhere. I think they do not differ from the version of firmware for the above-mentioned controller. Yetlinks. I don’t know if the firmware is duplicated for different links or not - see for yourself by the date of writing the binaries.

- Controller DD3663 LUA.A82PA slightly modified version of the DS3663.
Link to Yandex Disktyk.
If you came to the controller DD3663 LUA.A81PA, the specified version of the controller will not be sewn. You need to correct the firmware file from my archive renamed LUA36A81.bin and then pick up the controller firmware.Here VPP147. She laid out like the original firmware LUA36A81, though only three permits: 1920 * 1080 DO8, 1366 * 768 Si8 and 1600 * 900 DO8. There is a difference exactly what is the firmware to sew, or not - I do not know. Some launch your controller renamed firmware, others swear that in addition to native firmware, no shamans help - I do not know, and it's nothing to check with me.

- T.RD8503.03 controller . Positioned as a new version of SKR.03. Here link to firmware.

- controller RR8503 . Link to Google Drive tyk .

- T.V56.81 controller
cloud with firmwarehere. Ibid, and manual on it. The controller is not quite universal - firmware or 1366 * 768 or 1920 * 1080 resolution, but the bit depth of the Chinese claim that there is a bit of 6/8/10. Is it all true - not tested neither I nor anyone else (I'm talking about 10-bit).
- ZL.VST.3463GSA controller
About the same thing, as I said two lines above - there are firmwares only for resolutions of 1366 * 768 and 1920 * 1080. By bit is not aware of. Firmwarehere.

- QT526C V1.1 controller DVB-T2, DVB-S2
link to Google Drivetyk.
- QT526C V1.3 Controller DVB-T2, DVB-S2
link to Google Drivetyk.
- QT526D V1.2 controller DVB-T2
link to Google-ROM firmware and a description ofhere.
Stilllinkon Google Drive from the Chinese store with firmvarami on QT526 V1.1 and V1.3.

- DS.V53RL.BK controller (TSUMV53RUUL-Z1 on chip)
Links to the cloud with firmware here(1)and(2).

- M53V5.1 controller (TSUMV53RUUL-Z1 on chip) . Flash and manual for the controller firmware here .

- RR52C.81A controller . Link on google-drive with firmware.
- controller RR52C_03A . Link on google-drive with firmware.
- controller RR52C_04A . Link on google-drive with firmware.

The archive firmware on z.vst3463 there is one very interesting text document. However, whether the archive it spoils, so spread apart.
How to make it work with 1366x768 and 1280x800 SI6.txt(1.21 KB)
This suggests that you can use the firmware with one bit rate on the matrix on the other (for example, if there is simply no suitable firmware). After the firmware, it will be used to fix the trouble in the service menu to the desired one. And this is not only for Z.VST3463, for other controllers it is just as fair. In all television universal controllers, LVDS MAP has 16 fixed "cards" presets. But this method acts and even in those controllers that have only 2 presets (I personally made sure that the example of the M.NT68676-2A controller).Here isanother confirmation of this.
By the way, if the text and everything that is written here is poorly absorbed, I recommend watching other videos from langeron.ua. I am not familiar with them and this is not an advertisement (I hope they will not be offended for the link to their video). Just in them you will find a lot of video tutorials throughout the material. BUTherethey can dig and search for firmware not only on the above-mentioned controllers, but also on some others.


Link to the controller firmware M.RT2281 / M.RT2281.E5here. It will not be necessary at all often, but if I show this link, it means that someone here had an "experience".

Restore the controller in case of unsuccessful firmware.
If you are as a result of any incorrect actions with firmware (poured the firmware not from that controller, the light blinked during the firmware, the computer failed, etc.) Okirpichili your controller, you will have to remove the SPI Flash memory chip (8-legs chip With the exemplary name 25Q32 or 25Q64) and flash the recovery dump into it. Dump (eng. DUMP) (for those who do not know) - this is the firmware working body, which is prescribed in SPI Flash when you flash the controller with a regular way (via USB). Dump is not a firmware from the above archives, but what is written in SPI Flash as a result of the firmware process. However,hereThe method of firmware firmware ZVST3463 is described in detail without removing the chip (by the way there is no difference, which controller will be restored in this way). So there a man tried to pour not only a dump, and also the firmware binary as it is. As far as it is right and what will be the further results of the work of the controller - unknown. It is possible that if the controller is restored in this way, then you can make a "control" - after restoring to flash once again with a regular way. I would personally have stitched the dump - the benefit here they are available for all controllers. Choose someone who is more convenient - someone uses the programmer for SPI Flash, and someone is more convenient to program through Debug. You can also do this with ARDUINO (herestart and a few posts below). In any case, if the controller is oskarpichen, then incorrect experiments are no longer scary for him - nothing will burn. The main thing in our business is to restore the controller, and in what way it will turn out - the course of events will show.

Dump firmware for controller on TSUMV56Dump_LAMV56.rar(2.68 MB)
Heremore recent and from a more reliable source.

Memory dump for controller on TSUMV59Dump_LAMV59.rar(2.45 MB)

Memory dump for controller on MST3463 (board with DVB-T2)Dump_ZVST3463.rar(3.94 MB)
Several dumps for the same fee (z.vst3463) for different resolutionshere(thanks to the forumASS17 - he found it, it may turn out that it’s installed not Russian, but this is easily fixable)

Memory dump for LA.MV9P.V29 controllerDumpLAMV9P.V29.rar(242.32 KB)


Memory dump for controller T.VST29.03 (this is on a V59 chip with an R620 tuner). Dump with permission 1366 * 768.T.VST29.03 TSUMV59XU-Z1 1366x768 dump 20 06 2013.rar(2.65 MB)


Memory dump for the jumper controller HX6856_MT651-Bhere.

Memory dump for the controller D3663 LUA.A8.2PAhere.
Memory dump for the controller DS.D3663LUA.A81.2.PAhere.

The memory dump of the QT526 controllers is the firmware itself. Thus, you can immediately take the necessary and flash it in SPI Flash (personally did this).

Memory dump for controller SKR.03s_1542830117_SKR_03.rar(3.31 MB)


All recovery dumps have been removed from work boards stitched under ANY matrix (!!!). After flashing and installing the microchip in place, your controller will simply “come to life” and will be stitched for some kind of extraneous resolution. Then, in a regular way, you can upload the firmware you need.

And further. There are several ways to solder a memory chip. I warn you: do not try to unsolder one foot at a time - you will damage both the chip and the tracks. If there are no special means, then at least useby thisorby thisby methods.

thankkenst2009 , he gave a link on a whole cloud of someone's "accumulated experience". There you can also find something. But we must act carefully!
IMPORTANT!!!
There are cases that after uploading the firmware and turning on the controller, the matrix works in the "Pattern" mode (palette) - i.e. alternately enumerates color fields and no shamanism gives a positive result. Not often, but still happens. Most often - this is the lot of matrices Samsung and (possibly) BOE. The reason is a mismatch of the clock frequency (probably !!!). An example of the "recovery" of the matrix LTM200KT10 with a DS controller. d3663lua. a8.2 described in detailhere. For this, a datasheet on the matrix is ​​required. But this is not the main thing - you need to understand what you are doing and what the consequences may be if you use this tool incorrectly. In the above example, the parameter was changed, which is possible and has no relation to the clock pulses and the fact that the matrix has earned is pure chance. But the precedent has been created and the attempt does not threaten with malfunctions. To burn does not burn, but if something goes wrong, you will have to reload the firmware again ...