PowerBook 100 screen corrupts after 1-2 hours

[Bolacore]

Hi there, I'm trying to track down an odd issue on a PB100. It's been recapped and cleaned (LCD and inverter too) and works flawlessly at first, but after it's been on for an hour or two any open windows start turning black like this:

If i move or close the windows it lightens up in that area, but the contrast on the rest of the screen stays very weird. Any new opened window does the same thing, and it gets progressively worse and more corrupt looking the longer it's left on.

Changing the contrast via the knob kinda helps, but it makes the rest of the screen washed out looking to make the windows visible again. Leaving it off for a while makes it work great again until it hits that hour mark.

Any idea on what could be going on here? Not sure if the LCD itself is just failing or something else. Thanks!

 

[3lectr1cPPC]

Issues that appear after the system has been on for a while usually seem to be caused by bad/flaky solder joints somewhere. I'd start by inspecting the entire thing, especially around where any capacitors were and reflowing anything you can.

 

[Bolacore]

I've checked the ones I've replaced already and so far it seems ok. It seems to be the exact same symptoms as this person here: https://goughlui.com/2016/09/24/repair-apple-macintosh-powerbook-100-gotta-re-cap-em-all/

The board was really in great shape as far as i saw, but I'll get it under the microscope tomorrow for a better look, i see a few previous posts about the inductors in the corner being spotty. Thanks!

 

[alexGS]

Strangely, my PowerBook 100 LCD started doing something similar the other day. It is different though, because mine is flickery all over and doubles up/washes out pixels and lines, rather than having some of the display perfect and other parts all black. That seems like a logic board fault to me.

Mine had worked fine when used occasionally over the last year or two, but after an hour of use, it suddenly went like this. Now it does it every time, unless left overnight, after which it will be ok for a couple of minutes. Shame because there were no dead pixels or trackball mark etc.

I tried swapping the panel itself with my other ‘spares’ PB100, and the other one works fine, ruling out the multicoloured cable, interconnect board, the inverter board, and the logic board etc. as the cause of the problem.

I would like both of these machines to work, having put lots of effort into unyellowing plastics and replacing electro caps etc., but it’s starting to feel like I’m going to have to strip one for parts, as it’s going to cost more for the screen than the rest of the machine is worth. I think?

-Alex

 

[alexGS]

The most bizarre thing happened today…

After numerous attempts to find a fault with the LCD, shortening the input ribbon cable, cleaning/reflowing the driver IC pins (you can see why in a photo below, I felt sure that was a short!) and an attempt to wash under the same driver ICs in case ‘capacitor juice’ had got in there, none of these things made any difference.

I even tried desoldering some of the LCD’s ribbons so as to clean and inspect under them. When I saw a crack in the corner of the glass (where contacts are), I thought that was ‘game over’. But then, I reckoned that would knock out rows or columns of pixels, which wasn’t the case.

I put it all back together and it looked as bad as ever - but then, all by itself, it came right. The flickering stopped and the contrast returned to normal. Restarting does produce the ghosting/flickering on the ‘Welcome to Macintosh’ box, but then it crisps up. It feels as though something is marginal and comes to life when the brightness gets a boost during the startup.

Very strange to me, I’ve never seen an LCD fix itself. Perhaps my washing under the ICs was successful and it just took a while to dry out (might still be drying out). I had already replaced the SMD electro caps with MLCCs, so I think once it is fixed, it should stay that way. Apart from the mystery of the cracked glass (I don’t think I did that)…

 

[alexGS]

Changing the contrast via the knob kinda helps, but it makes the rest of the screen washed out looking to make the windows visible again. Leaving it off for a while makes it work great again until it hits that hour mark.

Any idea on what could be going on here? Not sure if the LCD itself is just failing or something else. Thanks!

Sorry to have hijacked your thread, Bolacore - but I do wonder whether we have related difficulties - my ghosting/contrast problems were also split vertically (top half/bottom half), like yours. Can I suggest in light of my accidental success today, try standing the display on end and spray contact cleaner under the two surface-mount ICs on the LCD board? I think mine had ‘capacitor juice’ under there, as an amber liquid ran out. I kept spraying until it ran clear.

At first there was no difference, but after the display had been on for a little while, it suddenly cleared up (I was really surprised).

-Alex

 

[alexGS]

And two days later, it was back to how it was.

Plus, my other LCD now has a similar problem, but only affecting the bottom quarter of the display, with the other three-quarters completely unaffected.

These LCDs do my head in!
I wish I knew what was failing.

I start to wonder whether the MLCC caps are ‘wrong’ for the voltage divider IC; whether they have wrong ESR (too low) which overloads some part of the op-amp circuits in that IC…? I’m clutching at straws. Does anyone know how to fix these problems?

 

[techknight]

You cannot use ceramic caps in an LCD. too many varying voltages, Ceramics will cause issues with its internal capacitance changing against the applied voltage.

Not saying this is your cause, but ceramics work in alot of things but only under certain cases.

I tried using ceramics in my 145 long ago and it was still drift-happy and weird. Ripped it apart and changed them with tantalums and its fine now.

 

[3lectr1cPPC]

Interesting to hear that, I used ceramic caps on my PB 145 and it has always worked flawlessly. My 180c on the other hand was a mix of tantalum and ceramic caps and so I wonder now if that’s what is causing my display contrast issue.

 

[techknight]

Interesting to hear that, I used ceramic caps on my PB 145 and it has always worked flawlessly. My 180c on the other hand was a mix of tantalum and ceramic caps and so I wonder now if that’s what is causing my display contrast issue.

Yeah that has not been my experience. brightness still drifts all over the map with ceramics.

 

[3lectr1cPPC]

I'll make note of that in the future, and I also many go back and replace the few ceramic caps on my 180c with tantalums at some point to see if that fixes the display contrast problem it has.

Also worth noting about the PB 100 screens is that the guy on the website mentioned earlier had the problem before recapping as well, so I doubt that ceramic caps are the issue here.

 

[alexGS]

You cannot use ceramic caps in an LCD. too many varying voltages, Ceramics will cause issues with its internal capacitance changing against the applied voltage.

Not saying this is your cause, but ceramics work in alot of things but only under certain cases.

I tried using ceramics in my 145 long ago and it was still drift-happy and weird. Ripped it apart and changed them with tantalums and its fine now.

Thanks I feel this is a likely cause - when it failed, it was intermittent and sensitive to displaying certain patterns (like the startup checkerboard background) which would then make it ‘fix itself’ for a while. This suggests to me that electrolytics would perhaps ‘cushion’ these transient voltages.

I also feel it’s now too late - I popped out to buy some generic electrolytics today (too large to fit properly; can’t buy tantalums locally in 3.3uF) and swapped them in, but no difference. In the last few days, it’s gone from intermittent to permanently faulty.

I tried measuring the signals at the line driver/demultiplexers but couldn’t find anything different or missing.

Hard to see in the second photo, since I didn’t bother refitting the backlight while trying to clean and test the board. I can’t find anything wrong so far; but I think tomorrow I will unsolder the LA5316 voltage-generating chip - there are traces underneath it that I think I need to check. I’m also considering swapping the LA5316 chips between my two faulty LCDs, to see if that changes symptoms.

Basically, the contrast is wrong and there’s a shimmering appearance to the bottom quarter of the display. Is that likely to be the fault of the row driver chip on the flex-ribbon? Seems impossible to replace…

 

[alexGS]

So in the end, on the display where the bottom quarter intermittently and then permanently failed, it turned out the capacitor juice had rotted through almost all of the flex ribbon traces where they are soldered to the PCB.

I hoped I could repair the connections and gave it a try, but some of the very fine traces were broken. It’s basically game-over. I imagine replacing these chip-on-flex things is practically impossible; you’d have to re-bond the 100 tiny connections to the LCD glass edge. And, where do you get a replacement from… perhaps an LCD with broken glass?

I transferred the PCB over to the other, flickering and ghosting display, and that solved those problems. There were some dead-lines where one of the chips seems to lose connection to its flex ribbon. Apart from assembling it with some foam tape to press on the chip, I don’t have a solution for that one.

At least I now have one working LCD of the two.

 

[3lectr1cPPC]

Ah yes, that will do it. This is only going to get more and more common as time goes on.