VOGONS


First post, by Tommaso74

User metadata
Rank Newbie
Rank
Newbie

Hi, it's my first post here but I'm a long-time reader of this valuable forum...

I'm in search of a bit of help, I hope a LCD tech guru can give me a hint.

Here's the problem:

I recently bought a COMPAQ Contura 3/25S whose screen suffered double-side vinegar syndrome. The screen showed the caracters (bios POST, DOS prompt etc.) only in the corners, the central part of the screen was unreadable.

I disassembled the screen and noticed a leaked capacitor on the high voltage side of the PCB. The liquid has damaged the green plastic coating that protects some copper paths. I removed the bad cap, washed the area with isopropilic alcohol and soldered a good capacitor back in. Some copper paths are now naked, without coating. The copper paths aren't interrupted (the multimeter 'beeps'), but some paths are now almost black (oxydation?) under the coating film (see picture).

attachment%5D

Now the bad part (oh well, the premises are bad too, but I mean, the really bad part)

When I turned on the notebook after the 'clean & recap' operation the screen began to act strange, intermittently showing the characters.
I tried to reseat the white ribbon that connects the motherboard to the screen PCB, but the characters disappeared definitely.

Any help? I have some soldering skills and a multimeter.

The backlight works well. The brightness sliding resistor works, it diminishes and increases the light bulb power perfectly.

The 15 pins colored-wires connector from the PCB to the Sharp screen circuitry is ok (I tested every pin).

The white ribbon from the motherboard is also OK, no interrupted paths.

Pressing FN key + F3 (screen negative/positive) does nothing.

Pressing FN key + F4 (external/ internal screen) turns off and on an external monitor connected to the VGA port, as it should be, I can see everything (bios, prompt, programs etc.), and I can see colors, too.

In the picture, the area surrounded by a yellow circle is a large copper path where the positive extremity of a SMD tantalum capacitor is soldered, it originally showed 0.02 volts but after cleaning that area with heated soldering paste it now measures 0.57 volts.
The screen contrast sliding resistor make it drop to 0.52 volts when set to the left side. This copper line ends to the 2 V-in pins of a LM337 integrated circuit. Is it normal that a so low voltage is given to this IC? Do you think that this LM337 IC controls only the screen contrast? Could it be that the contrast is so low that the screen doesn't show the characters, that are there, but remain 'invisible' because of low contrast?

Some voltages I found around in the PCB are steady: 4.9V, 18.1V, 1.2V.

Wow, long post... sorry!

Thank you in advance

Attachments

Reply 1 of 7, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

The discolouration on the traces surrounding where the capacitor leaked suggest a bad trace still. Have you tried scraping off the darkened green solder mask to see what the trace underneath looks like? Perhaps they need tinning up to get them back to normal or you might find a break.

Did you reseat that ribbon connector while the laptop was on or off? I ask because I made the mistake of running my Thinkpad 240 while I was messing with the backlight ribbon cable - which was in at a slight angle. When powered the laptop's screen would no longer light up and I had to reverse engineer / trace through quite a bit of the backlight circuitry to work out that I'd broken a transistor that would've run the backlight enable circuit. Hopefully something similar hasn't happened with yours.

IT's a mono screen right? Do you have the part code? It may be possible to track down the LCD panel datasheet or one for a very similar panel and you can work out what the voltages should be on the different pins. From what you've described it seems that the voltage to set the contrast may have ended up too low / high. I seem to recall that the contrast or LCD drive signal or something should be around negative 20 volts.
I've attached the datasheet for the LM64183P which is a similarly cursed LCD to yours in my mono AST bravo NB where the polariser has broken down - of course in trying to fix that the middle of the screen ended up damaged in the middle and I can't find any polarisers that work - mono LCDs should all work relatively the same way so this might be good to refer to:

Attachments

Reply 2 of 7, by Tommaso74

User metadata
Rank Newbie
Rank
Newbie

Thank you Thermalwrong,

I will try the scrap & tin technique as suggested.

Thermalwrong wrote:

Did you reseat that ribbon connector while the laptop was on or off? I ask because I made the mistake of running my Thinkpad 240 while I was messing with the backlight ribbon cable

You are not alone !!! Sadly yes, I reseated the motherboard ribbon while the power was ON! Ok so maybe I have blown a transistor or something else... But if I remember correctly the characters appeared and disappeared eventually on subsequent reboots after the ribbon reseat.

How can I check if a transistor is not working properly?

The model of the monochrome LCD is Sharp LM64P728.

This is similar to the one on your datasheet and other datasheet I found around, see attachment, so I have to find a negative voltage of ~ -18V on pin 4 named 'DISP', a +5V on pin 5 named 'VDD' and +25V on pin 7 named 'VEE'. I'll check and report the voltages.

What I don't understand is: there are 15 pins on datasheets, but the check holes in the driver PCB are 14 (see photo). Why the engineers eliminated one check hole on the PCB? Is the ground the missing hole?

I will further investigate and report back.

Attachments

Reply 3 of 7, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

In my case I was able to fix the damaged transistor because I have multiple Thinkpad 240 laptops and was able to work out the purpose of each pin from doing some probing. It was a transistor that ran the backlight enable signal that I'd damaged and it gave a lower voltage while my known good board gave something like 3.3v on that pin.

I'm not sure how the LCD connector could have just 14 pins, it doesn't look like there's really a pin that could be left out. Doesn't it use a flat cable connector rather than a pin header?
For now I think it'll be best to check out the inverter / contrast board since that's where the damage is visible

ooh, have a read of the last paragraph "UPDATE (Feb 2014)" in this blog post - it's the same problem as you're having and it's a damaged trace from electrolyte !! 😀
http://www.toughdev.com/content/2011/09/compa … c-80386-laptop/

Reply 4 of 7, by Tommaso74

User metadata
Rank Newbie
Rank
Newbie

Thanks again Thermalwrong,

the connector from the inverter board to the LCD driver is a 15 pins molex one, it's just strange to me that the check holes are 14 as in my previous photo. It is difficult for me to check a voltage with the multimeter directly on the molex pins, they are so little that I fear I can short two pins and damage the circuit. I want to use the check holes to measure safely the voltages. If I won't easily find the 1 by 1 relation datasheet <-> holes I can try with a small sewing needle to touch the pins directly on the molex.

And yes, I have already read that toughdev.com blog too!!! I really hope it's just a matter of damaged trace, I will try to tin the oxydated traces as you suggested.

I will report my progresses in a day or two.

Reply 5 of 7, by Tommaso74

User metadata
Rank Newbie
Rank
Newbie

Well I tried to tin some areas, desoldered and resoldered the 1 microFarad tantalum capacitor, resoldered the new capacitor, checked with ohmeter that no traces are in short with others. No difference so far.
As for the voltages, I have found that pin 4 of the connector is giving +2 V . The datasheet says pin 4 is the DISP pin, and 7th pin should have a -17 V negative voltage, but has no voltage at all. See photo, maybe the trace under the little transistor is interrupted? It seems that it is the same trace that should end to the right side of the transistor named S6AB, where a 0,21V is read.

Attachments

Reply 6 of 7, by Tommaso74

User metadata
Rank Newbie
Rank
Newbie

Trying to track down the -32V rail coming from the motherboard ribbon,:
there is -32V until the transistor circled with red in the photo attached. Then nothing.
I'm not an electronic engineer, but I think it's not normal.
All the tracks marked in yellow have no voltage, and they end to pin 7, in datasheet called Vss (some other datasheets call it Vzz) 'Power supply for LCD (negative)' .

Attachments

Reply 7 of 7, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

Is that while it's powered or off? That's probably maybe an NPN transistor (look up the markings & "sot23" to confirm) and that would break the 'rail' at that point while it's off.
If it's also doing nothing while it's on, see if you can figure out which pin is the gate of that transistor and see if that's getting voltage at all. Probing very carefully while it's on you should see a positive voltage there which would allow the -32v through. If the gate is getting a positive voltage but there's no -32v going through the transistor then maybe that's the part that's damaged. It could still be a trace elsewhere on the board that's damaged though so you should make sure all the traces that are visibly blackened still are connected from point to point.

It might be worth getting in touch with the people from that blog thread as they should know more about troubleshooting this than I can help with.