Information display LCD is officially not serviceable.

[BrianG]

My information display LCD has the tell-tale horizontal streaks with temperature dependence that many have experienced. I took the entire module out the other day and attempted to diagnose the problem. After removing everything from the circuit board to get a good look at the display LCD, it was very clear that there was no hope of repairing the display. I hoped that the display was mounted in the standard way that most LCD displays of this type are mounted - sitting atop two conductive rubber strips that make connections from the conductive film circuits printed on the glass LCD substrate to traces on the PCB below (usually gold-plated to eliminate the chance of oxidation, and thus, poor electrical contact). The two smaller LCD displays in the instrument panel (the odomoter and clock/date displays) are mounted in this fashion, but the main center LCD display is not. They used a high-density flex-circuit which is bonded (not reliably) to the glass LCD substrate edge on one side and soldered to the main PCB on the other. There is also a flip-chip display controller IC on the flex circuit (see picture).



The long and short of it is that you can desolder the flexcircuit from the main PCB, but if you don't have a replacement, you're SOL. It think that this display is custom made for VDO (the company that makes this entire instument module for Audi), and so it is not a part that you can buy for a electronics distributor. I will check into that, however.

There is no good way to repair the connection between the flex-circuit and the LCD glass subtrate. I tried a number of things including re-assembling the metal retainer clip with a couple of 0.050" shims under the locking clips in an effort to apply a considerable amount more force to the display edges (including the edge that has the flex-circuit termination). I hoped that squeezing the flex-circuit harder against the glass would make a better contact.... Nope.

I also made sure that the solder connections from the flex-circuit to the main PCB were solid. It can be nothing other than a bad connection from flex-circuit to the glass or a bad controller IC on the flexcircuit. Neither situation is any friendlier than the other.

Well, at least I've proven to myself that I can do nothing other than plunk down the cash for a new module. I'm out of warranty, so I don't think they will replace it for free, even though it is a know defect.

Sorry I couldn't come up with a solution, guys.

[S4mig@PEB tuning]

[BiturboBoy_V2.0_S3--]

when you say bonded to the glas...how so?

[ModifiedA4]

what is the exact failure mode of the screen? bad electrical connection or an internal flaw in the screen?

good info by the way..

[BrianG]

that if you so much as sneeze on those conductive film circuits on glass LCD substrates, they will no longer conduct. That is why I didn't try to mess around with that connection. If I buy a new module, I may sacrifice the old one to science and give it a try.

And yes, I am an electrical engineer.

[Vegito@BostonAudi]

i am a computer engineer so i know how fun that would be you probably are just going to have to get a new unit.

[BrianG]

and I'll tell you why. On my display, the failure has a distinct pattern to it. Every pixel in each affected line has an erroneous brightness. All pixels in one line might be very dark while all the pixels in another line are brighter than normal, while still the majority of the lines function fine. The reason that this is a good clue is that LCD screens like this (and the ones in laptops) have one electical connection for each column of pixels and one connection for each row. By applying a voltage across a single row connection and a single column connection, you illuminate the pixel common to those two lines. By very quickly cycling through all the combinations of row/column pairs and applying the appropriate voltages to each pixel, the screen appears to have pixels illuminated constantly. In reality, though, they are all being switched on and off many times per second.

OK. Now that we've got that straight, a problem in an entire horizontal line indicates that all pixels in that line have the same problem and what do all those pixels have in common? That's right- they all share one row connection to the flex-circuit.

If there were defects in the LCD itself, I would expect to see individual pixels behaving badly, not entire lines.

This problem could also be the display controller chip on the flex-circuit as well. It is what they call a "flip-chip" mounting of the IC. Underneath the chip, there are little metallized solder "bumps" that make contact to the flex-circuit. They underfill the edge of the chip after it is attached with an adhesive to prevent moisture from getting under there and causing oxidation. There could be bad contacts underneath there, as well.

[BrianG]

[BiturboBoy_V2.0_S3--]

i'm going to bed.

[UberVier]