This project is about building a custom flat-panel screen for operating a computer in my kitchen. The idea is to use the panel from an old laptop, mounting in a custom case, making it very flat (hopefully only about 15mm thick). Once I have the panel working with DVI input, I will look into adding touch input - preferably projected capacitive touch, which will allow sensing through the front glass, thus allowing easier cleaning.
The following tasks and ideas are considered:
A friend of mine had two different laptops laying around, which he generously allowed me to strip the panels from.
First panel is from an IBM ThinkPad T41. The panel is a Samsung LTN141XA-L01. This panel is 14.1" with a resolution of 1024×768 and a single CCFL for backlight.
The second panel comes from a Toshiba Integra M4 tablet PC. This panel is a Toshiba LTD141KN5K. It is also 14.1", but with a resolution of 1400×1050. The panel has a built-in digitizer for the tablet PC, but in this application the digitizer will not be connected. This panel also has a single CCFL for backlight.
Because of the higher resolution, I would prefer using the Toshiba LTD141KN5K panel for this project, but it might be more difficult to get it working, because it is not easy to find the timing specifications etc.
The panels used in laptops use LVDS-interface, so in order to use it with a computer, either a graphics card with LVDS-output is needed (found on some embedded motherboards), or a converter from VGA or DVI to LVDS.
I will go for the VGA/DVI to LVDS solution, and I have found the RM5451 controller board with audio from njytouch.com suitable. Njytouch also have an the controllers for sale on eBay, where they are bundled with the needed cables and inverter to suit the panel you have.
Since these controllers also drive the inverter and provide on-screen display for adjusting brightness etc, the kit you get from Njytouch contains everything you need to make a bare LCD panel into a complete flat-panel monitor that can be connected to your PC. Finally you can get a small screen with a high resolution.
I contacted Njytouch on eBay and asked for compatibility with the two panels I have. The Samsung LTN141XA-L01 is reported to be working, but as stated earlier, I prefer the Toshiba LTD141KN5K, due to the higher resolution. The initial comment from Njytouch, however, was that they did not know the specifications of the panel, so they could not guarantee that it would work. However, they are very friendly and supportive, and I believe it will be possible to get this working.
At this point, I will be ordering an R.RM5451 LCD controller board with audio, programmed for the LTN141XA-L01, and with the needed cables and inverter. That will allow me to test this panel, and I will have something to fall back to, if I cannot get the Toshiba LTD141KN5K working.
I will also order a programmer, allowing me to change the configuration of the controller, and a suitable LVDS-cable for the LTD141KN5K. Then I can experiment with the programming of the controller, and if everything works out, I will order a second controller for the other panel.
The Toshiba LTD141KN5K seems to be manufactured for use in the Integra series tablet PCs, and since both the panel and computer is made by Toshiba, I have not been able to find any datasheet, so there is some reverse-engineering to be done. I will try to document my findings here.
This is what the panel looks like:
I started out by looking at the cable connected to the LTD141KN5K (I have removed the cable from the computer). Njytouch suggested that the cable was most likely the FIX-S6 30pin cable needed for the panel. It seems correct in terms of the connector used. I also found a cross-reference for compatible connectors from different manufacturers and verified with a few data sheets that they seem correct.
Then I looked at the pin-out used in the cable, noticing what pins are wired as twisted pairs, and their color.
I found a datasheet for Toshiba LTD141EM4V, which is the same size and resolution, and uses the same 30-pin connector (the connector is JAE FI-X30). Comparing the connector pin-out with my findings from the cable gave me a feeling the connector pinout is as follows:
| Terminal | cable | symbol | function |
|---|---|---|---|
| 1 | blue single | VSS | GND |
| 2 | blue single | VDD | Power supply: +3.3V |
| 3 | blue single | VDD | Power supply: +3.3V |
| 4 | N/C | VEDID | DDC 3.3V POWER SUPPLY : +3.3V |
| 5 | N/C | NC | Non-connectionz |
| 6 | N/C | CLKEDID | DDC Clock (SCL) |
| 7 | N/C | DATAEDID | DDC Data (SDA) |
| 8 | blue pair 1 | RxOIN0- | Negative LVDS differential data input (Odd), [R0-R5, G0] |
| 9 | black pair 1 | RxOIN0+ | Positive LVDS differential data input (Odd), [R0-R5, G0] |
| 10 | blue single | VSS | GND |
| 11 | blue pair 2 | RxOIN1- | Negative LVDS differential data input (Odd), [G1-G5, B0-B1] |
| 12 | black pair 2 | RxOIN1+ | Positive LVDS differential data input (Odd), [G1-G5, B0-B1] |
| 13 | blue single | VSS | GND |
| 14 | blue pair 3 | RxOIN2- | Negative LVDS differential data input (Odd), [B2-B5, HS, VS, DE] |
| 15 | black pair 3 | RxOIN2+ | Positive LVDS differential data input (Odd), [B2-B5, HS, VS, DE] |
| 16 | blue single | VSS | GND |
| 17 | blue pair 4 | RxOCLKIN- | Negative LVDS differential clock input (Odd) |
| 18 | black pair 4 | RxOCLKIN+ | Positive LVDS differential clock input (Odd) |
| 19 | blue single | VSS | GND |
| 20 | black pair 5 | RxEIN0- | Negative LVDS differential data input (Even), [R0-R5, G0] |
| 21 | blue pair 5 | RxEIN0+ | Positive LVDS differential data input (Even), [R0-R5, G0] |
| 22 | blue single | VSS | GND |
| 23 | black pair 6 | RxEIN1- | Negative LVDS differential data input (Even), [G1-G5, B0-B1] |
| 24 | blue pair 6 | RXEIN1+ | Positive LVDS differential data input (Even), [G1-G5, B0-B1] |
| 25 | blue single | VSS | GND |
| 26 | black pair 7 | RxEIN2- | Negative LVDS differential data input (Even), [B2-B5, HS, VS, DE] |
| 27 | blue pair 7 | RxEIN2+ | Positive LVDS differential data input (Even), [B2-B5, HS, VS, DE] |
| 28 | blue single | VSS | GND |
| 29 | blue pair 8 | RxECLKIN- | Negative LVDS differential clock input (Even) |
| 30 | black pair 8 | RxECLKIN+ | Positive LVDS differential clock input (Even) |
The connections on terminal 6 and 7 caught my attention, and I managed to mock up an I2C-adapter and read out the EDID-data (read 128 bytes from I2C-address 0xA0).
This is what my quick I2C-hack looks like:
Parsing the data with EDID Viewer gives the following report: ltd141kn5k.rtf. A few interesting snippets:
| X Resolution | 1400 |
| Y Resolution | 1050 |
| Vertical Frequency | 60 |
| Maximum Vertical Frequency | 63 Hz |
| Minimum Vertical Frequency | 40 Hz |
| Maximum Horizontal Frequency | 67 KHz |
| Minimum Horizontal Frequency | 43 KHz |
| Maximum Pixel Clock | 110 MHz |
| Preferred Pixel Clock | 108 Mhz |
| Horizontal Active | 1400 pixels |
| Horizontal Blanking | 288 pixels |
| Horizontal Sync Offset | 8 pixels |
| Horizontal Sync Pulse Width | 8 pixels |
| Vertical Active | 1050 lines |
| Vertical Blanking | 16 lines |
| Vertical Sync Offset | 4 lines |
| Vertical Sync Pulse Width | 4 lines |
It seems I have found the needed timing specifications of the Toshiba LTD141KN5K, without needing the datasheet.
The backlight CCFL connector on the Toshiba LTD141KN5K seems to be somewhat non-standard. Here are a few images of the connector:
The inverter i gutted from the computer looks like this:
The high-voltage connector on the inverter looks like this:
I think the inverter that comes from Njytouch has a different connector, but I guess I can just replace one of the connectors (i.e. the one on the panel or the one on the inverter). Getting the backlight working will likely be the least of my problems.
Linux
Electronics
Nostalgia
Code
Projects
Spiff
Links