• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

New theory on the cause of the infamous Tunnel Vision problem - Testers needed!

3lectr1cPPC

Well-known member
@bwinkel67 on the TinkerDifferent forums proposed a new possible cause behind the infamous Tunnel Vision issue on monochrome and grayscale Active-Matrix LCDs on our precious PowerBooks! I offered to crosspost here to get more attention on it, and to hopefully get some people to test this new proposed fix out!

If you've been keeping up with the 8 page long tunnel vision thread here, you'll know that youtuber TheBasement (@Johanncerecke on these forums) did a video trying out the baking method on a 170 LCD with some success. Prior to this, he had tested another claimed fix of bending some tabs on the display in further, which unsurprisingly did nothing, proving this fix doesn't work.

Or does it?

First off, I'll copy in what @bwinkel67 said on the TD thread:
So I had an idea. I know that Johann did a video on his YouTube channel, The Basement, where he a) disproved the quick fix method of tightening the tabs, and b) baked the LCD with some success. The quick fix, I believe, came from another YouTuber in an under 5 minute video where he claimed that you could tighten the tabs and that would solve it (he also had a second method of reprogramming the logic chips which seems a bit questionable...dunno).

In any case, that got me thinking about the entire problem. Is it actually a moisture issue or is it just that the layers are coming apart due to failed glue after so many years? If so, then tab-tightening technique could work and when baking, perhaps it is the glue being re-flowed and not the moisture being driven out?! The temperatures listed are in the range of some glues softening, and if the layers are adhered together could that be what fixes it? Leaving it in longer allows the glue to soften more evenly perhaps. What I've seen from some finished pics/videos is that it returns more slightly and not as evenly as before (at least one video where a guy tried this with a different machine).

So if it is an adhesion problem then the technique of tightening the tabs could work as it is pushing the elastomeric strips to the circuit board which pushes back and perhaps presses the panel into the metal bracket forcing the layers to stay together when the panel warms up. Elastomeric strips do have some give (I'm assuming the LCD panel is using elastomeric strips since I haven't taken the panel completely apart). So I gave it a shot the other day and not only further bent the tabs inwards but also tried to get a pair of pliers to hold them firmly and and push the tabs towards the circuit board with a little torque. The screen still darkened but it seemed to have taken much longer and didn't seem to get as bad. Will do more experiments with it in the future. I'm thinking that instead of pliers perhaps a small hammer and pushing the board into the panel with a firm grip and then gently tapping the tabs might yield a tighter result...will see. Not too tight as I don't want to damage the elastomeric strips, which seem (on other devices) to be made of some type of silicon.

In any case, what do people think? Could it be an adhesion problem or has anyone tested specifically for moisture. I know some have said silica and vacuum helps, but then again, vacuum could also press the panel tightly together, re-adhering it for a bit. The fact that someone else said storing the LCD with a ton of silica packets (what I was trying) for half a year in a sealed bag had absolutely no impact just got me thinking, since that ought to drive out some moisture at the edges at least. I guess vacuum sealing it is supposed to squeeze the moisture out...trying to wrap my mind around if that makes sense. Btw, the guy who created the TRS-80 LCD's also tried baking them with no success.
For more context on this, here's the link to the thread itself: https://tinkerdifferent.com/threads/tunnel-vision-on-pb170.778/page-2#post-9513

So, a couple of questions:

1. Has anyone actually proven that moisture is the cause, or has it always just been assumed due to the symptoms present?
2. Can anyone here with a tunnel vision-affected display try this new idea out?
3. This has only ever been proposed for a 170 LCD, do the LCDs used in the PB 180, Active-Matrix Duos, and the 540 also have these tabs?

I'd try it out on my own 170, but my LCD actually doesn't have the issue, so I can't.

That's all I've got, let me know what you can find!
 

Johnnya101

Well-known member
I can answer 1. From my experience and time being here, it has always thought to have been moisture because putting the screen in some rice or baking/drying it has fixed it (maybe temporarily).
 

3lectr1cPPC

Well-known member
Has rice specifically worked for you? Baking helping can be explained that the heat re-softens the glue a bit which causes the screen layers to stay together better, but rice not so much.
 

techknight

Well-known member
yeah desiccant/rice does help solve the problem to a point. i ran this experiment over 10 years ago.

It is entirely possible there are multiple causes of the failure but moisture is one of them, and its because the panel seals fail and it gets in. Along with that, the layers can separate because again the panel seals fail.

It could also be contaminated liquid crystal that has slowly degraded over time, possibilities are endless...

Could take the panel apart, split the layers, redo the crystal if its possible and put it back together. but idk the manufacturing processes involved to do this.
 

3lectr1cPPC

Well-known member
Thank you for your input! I agree that multiple causes make sense for this, after all, we see different forms of the problem. Some report inverted colors in the affected areas while others report all black, for example.
So far, all that we've come up with to fix this curse is temporary and partial fixed (ex: oven). Could a more permanent solution (such as a method to reseal the display) be possible? What form does the seal take, is it something that goes around the edges of the panel, or is it an adhesive like what the polarizers are glued on with?
 

techknight

Well-known member
Thank you for your input! I agree that multiple causes make sense for this, after all, we see different forms of the problem. Some report inverted colors in the affected areas while others report all black, for example.
So far, all that we've come up with to fix this curse is temporary and partial fixed (ex: oven). Could a more permanent solution (such as a method to reseal the display) be possible? What form does the seal take, is it something that goes around the edges of the panel, or is it an adhesive like what the polarizers are glued on with?

honestly I wonder if speaking with someone overseas who makes/manufactures LCD panels could tell us.
 

bwinkel67

Active member
So I've had some success with the tunnel vision issues on my PowerBook 180. My screen starts out ok for 15 minutes and then develops the black areas. It goes slowly but after 90 minutes it gets bad. So I baked the screen at different temperatures with different results. First I tried YouTuber TheBasement's approach and put it in an oven for 10 hours at 185 degrees farentheit. The screen looked good for 15 to 20 minutes and went right back to as bad as it was at the 90 minute mark (so little to no change). I then tried YouTuber i80386sx approach of 4-5 hours at 215-220 degrees farenheit and after 90 minutes the screen hasn't changed. So the key is to get it right at 100 degree celcius mark, or slightly above.

I'm reluctant to attribute it to water, even though the boiling point is at 100 degrees celcius. I know there's been anecdotal evidence that suggests putting it in rice/silica helps the screen but I've also read others that said that storing in with rice/silica for 3 years had no change. With regard to baking, it may just be that things at that temperature re-bond (re-flow) the layers and their adhesives. What was really promising is that i80386sx did an 8-month follow-up video with no change to the screen (btw, he was not working with a PowerBook, but a PC laptop with similar problems). I'm curious to learn how LCD panels are made. Is there a baking process involved? Was I repeating what is done at the factory? If so, would be interesting to find out what temperatures are used.

Other things I learned....I don't think my tapping down the metal tabs idea was sound. The metal tabs are pretty soft so I don't think they would hold out for a long time if tapping them all towards the motherboard would be a fix (i.e. they would just rebound). The bezel is already pretty tight (when you release one side of the metal tabs, it immediately pops out). So there is likely no gain is holding the LCD more firmly in place. Also, after power down at the 90 minute mark, with the screen not showing any tunnel effects while on, I was able to confirm with the unpowered LCD not showing any signs either (usually you'd see dark spots). It did show a remnant of the menu bar so I don't know if that's normal behavior. It did eventually fade away.

I will update in a few months to see what the status of the screen is, though I don't plan on using it daily. I do live in a central AC'ed house so that may cause the humidiy level to stay low (i.e. if it is indeed a moisture issue). I don't plan to pack it in silica just to give it a fair test. If I can repeat i80386sx's results, that would be promising. I may also, eventually, look to see if the LCD could be repalced with a modern one -- similar to how YouTuber Noel's Retro Lab replaced an Amstrad PC's LCD with a modern one and used a converter to interface with the LCD's input.
 
Last edited:

bwinkel67

Active member
Oh, for those that want to bake it....some ovens can vary 30 degrees in both directions (so a 60 degree farenheit swing). We had an old gas stove that only swung about 10 - 15 degrees farenheit. One thing that helped was adding a couple of pizza stones that, once heated, helped keep the temp level and in the end the temp swung less than 10 degrees.
 
Last edited:

bwinkel67

Active member
Interesting video on making of TFT LCD. There is a baking process but that happens after each etched layer. It seems that the sealant is created to keep the liquid crystals contained. So maybe there is a leak between layers? I know there was a YouTuber that was creating their own LCD's at home. I will try and find them and ask some questions.

 

3lectr1cPPC

Well-known member
Over on the TinkerDifferent thread, we're taking a look at whether screen manufacturer has anything to do with it. So far, we know that the 170 and 180's LCDs were made by Hosiden, but what about the PB Duo 280 and the grayscale 540?

And of course, the Mac Portable, which we know is affected by the issue, but much less commonly than later LCDs found in PowerBooks and PC laptops for that matter. @techknight, do you know the manufacturer of the portable LCD?

Edit: Just got confirmation that the portable LCD is a Hosiden as well. Strange that it isn't affected as bad.
 

Cedsrepairs

Well-known member
If this is a moisture ingress issue, it's surprising nobody tested sealing the screen (with like silicone or something like that)

It would probably solve the problem ocne and for all on affected screens.

I'm affected but my units are in way too good cosmetic shape for me to risk anything on them. My PB180 is as good as new, apart from the tunnel vision (which only appear after a good 20min of use, I heard other are way more affected)

However, if I had a PB180 or so that's in a sorry state, it would be worth trying re sealing the border of the screen (or smth like that)
 

compu_85

Active member
On my PB180, I was able to reduce the bad tunnel vision some by just leaving the whole machine out in the sun for several hours.

I plan to do a few more rounds of this when it's not quite so hot, and not so humid.

-J
 

Cedsrepairs

Well-known member
When you say reduce , for how long ? How long does it take for tunnel vission to come back as compared to before ?
Can you give more details ?

Again it just seems that once dried out, something has to be resealed again (with glue or anything else) so that this problem disappears for good...
 

compu_85

Active member
Initially the screen was getting black around the edges after 15 minutes. Now it takes about 30 minutes. The amount of "black ring" is also less, instead of completely covering the menu bar, and making the apple and file menus difficult to find, now the apple logo is just mostly obscured but the menu is findable.

The screen does develop a "memory effect" after it's run a while. That also seems reduced, but not cured after leaving it in the sun.

-J
 
I read in some other thread that tunnelvision is actually caused by the failing TFT transistors, which have, over the years, been damaged by an unwanted DC offset on their AC supply. This would make sense in that physical pushing and bending doesn't fix tunnel vision symptoms, but heat reverses them to some extent.
 

bwinkel67

Active member
I read in some other thread that tunnelvision is actually caused by the failing TFT transistors, which have, over the years, been damaged by an unwanted DC offset on their AC supply. This would make sense in that physical pushing and bending doesn't fix tunnel vision symptoms, but heat reverses them to some extent.

That sounds interesting. Can you point me to where you saw that?
 

piernov

New member
Like a lot of these machines, my PowerBook 170 had the "tunnel vision" problem (1st picture). It later on developped a lot of "stains" as well, in just a couple of weeks, and it was closed but I may have left it powered on (2nd picture, taken after a long rest so no tunnel vision yet).
Both of these combined made the screen look very bad (3rd picture, taken after a few dozen minutes of running).
Decided to try baking the screen, tried less than 100°C at first but it made no difference to the "stains".
Going for broke, I pushed it above 100°C, probably 120°C, and the stains started changing shape.
Sadly, after a few hours, I was left with 2 big round spots, but surprisingly no tunnel vision whatsoever (4th picture, taken a few dozen minutes after turning on). When taking it out of the oven you could clearly see something solidifying again progressively between the pieces of glass.
"Tunnel vision" issue could come back, but apart from these 2 large spots (and a handful of other pixels), the image now looks absolutely perfect. Panel was ghosting more than it should have been, and looks like that's solved as well.
Unfortunately I think it's game over for these 2 large spots, but the LCD was already pretty back to look at before…
 

Attachments

  • 20220716_162745_e.jpg
    20220716_162745_e.jpg
    289.2 KB · Views: 55
  • 20220904_192749.jpg
    20220904_192749.jpg
    759.8 KB · Views: 53
  • 20220910_184549_e.jpg
    20220910_184549_e.jpg
    284.8 KB · Views: 63
  • 20220911_001559.jpg
    20220911_001559.jpg
    663 KB · Views: 61

3lectr1cPPC

Well-known member
That’s a rather strange failure mode, looks like some really bad pixel rot. What’s strange is that they originally didn’t show, and just appeared as you said. Also odd how they changed with the bake. If you did indeed leave it on for that time it’s possible that that somehow permanently damaged the pixels. Not sure.
 

3lectr1cPPC

Well-known member
That sounds interesting. Can you point me to where you saw that?

@techknight on page 1. That theory has been pretty much thrown away in recent years due to the moisture theory, but maybe there’s some truth to it. I don’t know nearly enough about electrical components to deduce this myself. I’d assume you’d need an affected and unaffected panel side by side to test on.
 

3lectr1cPPC

Well-known member
Here’s an interesting symptom that my affected 540 has - screen burn in

When I look at it head on, it’s fine. From an angle though…
5BA6D436-474C-4521-BF8D-4F930360A0D1.jpeg
This is before the visible tunnel vision has even started. Ignore the yellow bottom corners, that's some other issue, hopefully just some contamination on the backlight diffuser that I can clean.
 
Last edited:
Top