PowerBook 100 half screen dead

[bengi3]

Is this lcd fixable?

 

[desertrout]

Is this a suddenly new symptom? Is it consistent? My first thought is a bad connection, but would be good to know some additional details / context.

 

[bengi3]

It’s a new purchase (a whole bunch of busted stuff) so I know nothing.

 

[desertrout]

Ah ok. Do you know if the display (or the logic board) has been recapped?

 

[GRudolf94]

That panel is suffering from corrosion and you have a dead via to row driver #4. Stop running it to not make it worse. You'll need to recap it, or, if that has already been done, go over the work - carefully desoldering the row driver (which is a polyimide ribbon delicately bonded to the glass, and attached to the board by thin slivers of copper), and possibly running a jumper to driver #3 or somewhere else on the panel PCB. Look closely on this picture of a 145 panel which had the same defect - a pin on driver 4 has been jumped to 3 with thin wire for testing.

 

[bengi3]

ok, so I have to desolder and resolder all 4 row drivers? or just the last one? quite a challenge though. I have many busted lcd from PB100 so I can make some practice, but its really a challenge for me…

 

[GRudolf94]

No - removing the PCB from the drivers is not necessary, but go spot-checking things. See what pins of 1 have continuity to 2 and 3 with a multimeter, and see whether 4 has continuity as well.

 

[bengi3]

Ok, I can see that the mess is exactly on the other side of the PCB where the 8 caps are leaking. there’s still continuity between each pin of rows 1-2-3-4 and the corresponding pin. i.e. each row has 12 pins so I tested pin 1 of row 1 with pin 1 of row 2, then 3 and 4 etc... will recapping and cleaning be enough?

 

[desertrout]

Ok, I can see that the mess is exactly on the other side of the PCB where the 8 caps are leaking. there’s still continuity between each pin of rows 1-2-3-4 and the corresponding pin. i.e. each row has 12 pins so I tested pin 1 of row 1 with pin 1 of row 2, then 3 and 4 etc... will recapping and cleaning be enough?

Hard to know. Recap and see. Also best to recap the logic board too.

 

[bengi3]

I have several known good logic boards and use one of these for testing.

I recapped lcd with tantalium caps, cleaned everything with isopropanol and passed the solder over the 12 row 4 pins.

Result, half screen stil dead AND now blue screen of death…

 

[bengi3]

No - removing the PCB from the drivers is not necessary, but go spot-checking things. See what pins of 1 have continuity to 2 and 3 with a multimeter, and see whether 4 has continuity as well.

Incidentally, what is the continuity scheme?

1-row1 to 1-row2 to 1-row3 to 1-row4
2-row1 to 2-row2 to 2-row3 to 2-row4
Etc…
I noticed that pin 6-7-8 are continuous

 

[GRudolf94]

1 and two are half a screen (one field), 3 and 4 are the other half. You should have continuity on all of them except maybe two pins.
Should be 1-1 to 2-1, 1-2 to 2-2, 1-3 to 2-3, then at some point you'll have something like 1-4 goes to some part of the board, 1-5 goes to 2-4, 2-5 goes nowhere, then 1-6 goes to 2-6, etc.

This is just an example, it changes with each type of panel and I don't have either of my PB100s with me to check at the moment.
Whatever you see between drivers 1 and 2 should be the same between 3 and 4.

 

[bengi3]

no way, I tried to fix other two LCDs with the same defect (bottom half dead) and I can see that the driver connections are bad, so I guess its game over

 

[GRudolf94]

It's definitely fixable - you probably should just find a good panel to compare, or wait until I can measure one of mine.

 

[bengi3]

Thank you, in the meanwhile I'll check the bunch of busted PB100 lcds, from vinegar to dead pixels…

 

[dankcomputing]

That panel is suffering from corrosion and you have a dead via to row driver #4. Stop running it to not make it worse. You'll need to recap it, or, if that has already been done, go over the work - carefully desoldering the row driver (which is a polyimide ribbon delicately bonded to the glass, and attached to the board by thin slivers of copper), and possibly running a jumper to driver #3 or somewhere else on the panel PCB. Look closely on this picture of a 145 panel which had the same defect - a pin on driver 4 has been jumped to 3 with thin wire for testing.

I was able to do the exact repair you did on a LM64P51 using a dead LM64P58 as a guide (there appears to only be minor trace layout differences between the two). All of the horizontal TAB bridge thingies are connected to each other on each pin and I just reconnected the one that wasn't.