[Solved] T61 14.1" UXGA: Hitachi TX36D58VC1CAA woes

[leromarinvit]

Hi all,

I'm trying to treat my T61 to a nice UXGA panel. Originally, I wanted to buy a Sharp LQ141U1LH02 since this is known to work without any problems. I actually thought that's what I would get when I bought a Dell Inspiron 4100 with a UXGA screen, since I saw somewhere that Dell used these panels. But I got a Hitachi TX36D58VC1CAA instead.

I still proceeded with my attempt to exchange the display. With the combination of the T61 and at least this Hitachi panel, the pinmod described in these threads isn't needed. I examined both (panels') PCBs and (laptops') cables carefully, and both pins that came up in the discussions (1 and 2) didn't make any sense to cut: Pin 1 and GND are connected on both PCBs and cables, and the same goes for pins 2 and 3 (VCC).

Sure enough, no magic smoke emanated from my T61 when I turned it on. That is, however, where the sccess story ends: The image it shows is a mixture of a black/gray chess board, and irregular gray "clouds" that become brighter the longer it stays on.

The cable pinout seems to be compatible though: When I connect my old SXGA+ panel to the Inspiron 4100, it works, save for the fact that it still assumes a UXGA screen size, and therefore everything is cut off. Similarly, nothing is fried: If I connect the panels to the laptops they belong to, both work fine.

My T61's old panel is a Samsung LTN141P4-L02. I found a datasheet for the hopefully similar LTN141P4-L03, which describes the pinout. Based on that datasheet, I could also verify that both panels have the EDID EEPROM connected to pins 6 and 7. However, when I try to read the Hitachi's EDID on the T61 (with i2c-tools in Linux), it seems like all I2C reads from the internal panel fail (it still works over VGA). Even booting with the original panel and hot-swapping them produces the same effect. Using the Inspiron to change the EDID won't work either - the only EDIDs i2c-tools reads on that are those of external monitors, no matter which panel is connected.

Is it possible that the T61 detects panel hotswapping and disables I2C because it doesn't like the EDID? That seems to be the only explanation that fits the behaviour, other than the EEPROM being broken, but at the same time it seems quite unlikely. What do you think, would removing the EDID chip and flashing it externally help? Or would I waste my time?

[Raceboy]

It is not possible to read EDID with T61, neither with Intel nor nVidia card version.

T60 board works, but only Ati equipped one, Intel equipped T60 board cannot read/write EDID.

Found it out myself when I flashed SXGA+ EDID to the LCD (I used Powerstrip).

You will need to flash UXGA EDID info to the Inspiron panel, then I think it will work.

[leromarinvit]

I can read the original panel's EDID with i2cdump on Linux. "i2cdetect 2" shows an EEPROM at address 0x50, and "i2cdump 2 0x50" dumps the contents. It contains the string LTN141P4, so I'm pretty sure it's working correctly. I have an NVIDIA card.

If I couldn't access it at all, I'd have desoldered it already. But since I can still read the EDID after hot-swapping the panel twice (original-new-original), I assume I should be able to read (and, possibly, write) the new panel's EDID this way. Yet it doesn't work. That's why I'm worried that it might be something else.

[RealBlackStuff]

Like Raceboy said, you need an ATI-T60 to write into the EDID, at least with Powerstrip.
If I were you, I wouldn't waste my time or money on a 14.1" UXGA.
Text is way too small to read, and the quality of that screen is only mediocre at best.

[leromarinvit]

Well, I don't have a T60. But I just desoldered the chip, and now I have to find out three things:

1. If my makeshift heat isolation constisting of tinfoil around the PCB and a stack of paper between that and the panel was good enough (had to use a heat gun)
2. If my desktop's parallel port is good enough for a bitbanging I2C flasher
3. If the problem was actually caused by the incompatible EDID

Regarding the rest, the money is gone anyway since I already bought the panel. And there's no way I can return the laptop now that I've opened it, even if that were feasible in the first place (which it wasn't since I ordered it from overseas - I'm not about to pay half of what I paid for it just to ship it back).
Text size - let's see. I played around with the Inspiron a few minutes and it looked good to me. If something is too small, I can usually zoom in a little. And I think with enough resolution, even smaller text should be well-readable.
Screen quality: It's a shame noone ever produced 14.1" IPS panels. Regarding viewing angles, there doesn't seem to be much of a difference. It's a little darker, but I plan to replace the CCFL with a LED kit as soon as I can get my hands on one anyway.

[leromarinvit]

[censored]. I wrote the EDID of a BOE Hydis UXGA screen (form here) into the EEPROM and soldered it back on, but it still doesn't work. Exactly the same as before. The screen still works in the Inspiron. Any ideas what else could be wrong?

I'm at a complete loss what's going on here. I tried writing the EDID of the working SXGA+ panel into the new one and even exchanged the EEPROM (sacrificed an old RAM module), but it didn't help.

Edit: Got it!
The SDA and SCL lines didn't have pull-ups. I added them and now it's working nicely. But wtf, who designed this crap? Either leave out the EDID chip altogether, or add the pull-ups as well. But like this? This could never have worked!

[Raceboy]

Congrats on getting the screen to work!