Tunnel vision on PB170

cb5890

New Tinkerer
Jul 5, 2022
7
2
3
Thanks for clarifying! Interesting that the portable LCD doesn’t have this issue as much. What makes it different? You don’t hear about Portables with this issue. Techknight has one with the problem, but that’s the only example I know of. Not sure if his affected display is a backlit or non backlit though,
I'm backing a display for a portable with backlight upgrade right now, so the one with backlight is definitively affected. This one becomes completely dark after just a couple minutes of use.

But it's the only one of the four backlight ones that I have that is affected. The other 5 non-backlight ones appear just fine. So it's apparently not that common though - if you believe my small sample :D
 
  • Like
Reactions: 3lectr1c

cb5890

New Tinkerer
Jul 5, 2022
7
2
3
When I stored the portable in its anti-static bag a couple months ago, it was fully working. As I pulled it out today, nothing worked. So after some unfruitful debugging, I concluded that only moisture - maybe together with some flux residue and/or electrolyte could still be playing tricks. Hence, I cleaned the board again and then also backed it. (Maybe it was also just jealous because it wanted to be baked as well like its screen? That could be arranged :D) That did the trick and now it's back :)

You'll maybe notice the missing hybrid module. The little breadboard on the left side suggests a valid battery voltage to the power management IC. This lets the machine stay on without a hybrid, since the PMU IC and the OS believe that the battery is charged. In fact, right now they believe the battery is being charged, because I was too lazy to pull the CHRG_ON signal back low again... But that doesn't matter for now. I'm very excited that the board is going to receive one of @Androda 's hybrid replacement modules soon. Then this temporary fix will become obsolete :)

The board and the LCD are now running for more than 45 minutes without any darkening at all! :-O Previously, I couldn't even let the LCD run for a couple of minutes without becoming unusable. I'm absolutely amazed by the result. It actually worked - at least until now :-O

The two dark spots are really really unfortunate. The LCD was already damaged there, but the spots were basically invisible before. Backing somehow lead to them turning completely black with an area of very dark, but working LCD surface around it. Still, I would count the current state of the LCD as a massive improvement.

However, I haven't tested the LCD with backlight yet, since the backlight upgrade needs a dedicated expansion card to work... and to put in the logic board would make the entire setup even more fragile - given the current stacked layout... :D
 

Attachments

  • IMG_1298.jpeg
    IMG_1298.jpeg
    2.1 MB · Views: 113
  • Like
Reactions: 3lectr1c

cb5890

New Tinkerer
Jul 5, 2022
7
2
3
More than 90 minutes now. Not even the slightest hint of tunneling in sight. Really amazing :D

I have to try with the backlight again, but if the results are the same, then I would deem this screen fixed - until further notice.
 

bwinkel67

New Tinkerer
Jun 1, 2022
31
14
8
More than 90 minutes now. Not even the slightest hint of tunneling in sight. Really amazing :D

I have to try with the backlight again, but if the results are the same, then I would deem this screen fixed - until further notice.

Cool. What temperature did you bake it at and for how long? It's been about 2 months since I baked mine so I want to wait longer before I test to see how it currently is. Just stored on a shelf normally so if it's moisture related (I'm hesitant to believe that) then the screen should get to its bad-self again when I turn it on since it's been pretty humid here over July and August.

I just saw on 68kmla.org another theory which intrigues me. That the TFT transistors on the screen are going bad due to unwanted DC offset on their AC supply and baking them is somehow reversing that. I know there have been some folks (Adrian's Digital Basement) that have experimented with baking bad chips to see if some could return to use with mixed results, so interesting concept.
 

cb5890

New Tinkerer
Jul 5, 2022
7
2
3
Cool. What temperature did you bake it at and for how long?

I just saw on 68kmla.org another theory which intrigues me. That the TFT transistors on the screen are going bad due to unwanted DC offset on their AC supply and baking them is somehow reversing that. I know there have been some folks (Adrian Black form Adrian's Digital Basement) that have experimented with baking bad chips to see if some could return to use with mixed results, so interesting concept.
I baked it at about 100 degrees Celsius (observed temperature range was from 90 to about 110 degrees) and for about 4 and a half hours. Initially, I wanted to go for 6 hours, but as the temperature repeatedly overshot 100 degrees I didn't want to overdo it... My girlfriend only allowed to use me her old pizza oven, so temperature control wasn't that ideal :D

I also have to say that the screen looks kind of different than before. Usually, the Portable's backlight screen has a bright white-ish color. This one goes more into the green now. I imagine that could be the backings effect on the polarizer though. I don't mind it, but it's interesting what a trip to the oven can do to a LCD Panel :D

Hm, about the TFT transistors, that's also an interesting theory. I remember the episode from Adrian's digital basement, but his results were quite discouraging IIRC. However, in one of his recent episodes, he then actually accidentally fixed a bad chip while desoldering it. That caught him by surprise, so you never know :D
 

cb5890

New Tinkerer
Jul 5, 2022
7
2
3
Oh, right and I also wanted to share a picture on how it looked with backlight. Here it is :)
 

Attachments

  • IMG_1312.jpeg
    IMG_1312.jpeg
    1.8 MB · Views: 96

bwinkel67

New Tinkerer
Jun 1, 2022
31
14
8
I baked it at about 100 degrees Celsius (observed temperature range was from 90 to about 110 degrees) and for about 4 and a half hours. Initially, I wanted to go for 6 hours, but as the temperature repeatedly overshot 100 degrees I didn't want to overdo it... My girlfriend only allowed to use me her old pizza oven, so temperature control wasn't that ideal :D

That's exactly what I did...for about 5 hours. I think my temperature fluctuated a little less since we put pizza stones in the oven, which helped keep the heat. I first tried it for longer at lower temperatures (85 celcius) and found that it didn't do the job but the higher temp at shorter time did the trick. It's one reason I don't think it's moisture but maybe re-flow of adhesive or something. Though the TFT transistor idea would also make sense.
 
Last edited:
  • Like
Reactions: cb5890

bwinkel67

New Tinkerer
Jun 1, 2022
31
14
8
I also have to say that the screen looks kind of different than before. Usually, the Portable's backlight screen has a bright white-ish color. This one goes more into the green now. I imagine that could be the backings effect on the polarizer though. I don't mind it, but it's interesting what a trip to the oven can do to a LCD Panel :D

My screen did not seem to change in appearance, just no more tunnel vision (PB 180).
 

bwinkel67

New Tinkerer
Jun 1, 2022
31
14
8
I’ll have another test subject coming in next week - I just bought a grayscale PowerBook 540 off of eBay. It supposedly works so we’ll see whether it has Tunnel Vision or not.

Did it have tunnel vision? Another experimented I wanted to try, before baking it, was to put it in a refrigerator and have it turned on for an hour to see if cooler temperatures would affect the tunnel vision. This could then lead to either adhesive separating as it warms up or TFT transistor issues I suppose.
 

cb5890

New Tinkerer
Jul 5, 2022
7
2
3
That's exactly what I did...for about 5 hours. I think my temperature fluctuated a little less since we put pizza stones in the oven, which helped keep the heat. I first tried it for longer at lower temperatures (85 celcius) and found that it didn't do the job but the higher temp at shorter time did the trick. It's one reason I don't think it's moisture but maybe re-flow of adhesive or something. Though the TFT idea would also make sense.
Oh, the pizza stones are a great idea!

At first, I was really skeptical of the oven, because empty the temperature fluctuated way over the set temperature. However, when testing it with the LCD in it at a lower temperature, the temperature was much more controlled... But I didn't have the idea to just put more mass in there to balance the temperature better. Nice! :)
 

3lectr1c

Active Tinkerer
May 15, 2022
629
294
63
the United States
www.macdat.net
Did it have tunnel vision? Another experimented I wanted to try, before baking it, was to put it in a refrigerator and have it turned on for an hour to see if cooler temperatures would affect the tunnel vision. This could then lead to either adhesive separating as it warms up or TFT transistor issues I suppose.
Sorry, forgot to post an update on that. Yes, indeed it does! It begins around 20 minutes after it starts running, an interestingly it’s the inverted kind, so you can actually still see stuff in the affected areas.
A13178CB-5174-4416-8E11-988E94C9C6A9.jpeg
 
Last edited:

bwinkel67

New Tinkerer
Jun 1, 2022
31
14
8
It's only been about 2 1/2 months since I baked my PowerBook 180's screen (5 hours 100 degrees celcius), but since I recently brought it out again to try my newly acquired external monitor cable with RGB-VGA converter, I got to see my screen after sitting on the shelf all that time. I had the screen on for about an hour while trying to figure out my UnimacFly converter and there was no tunnel vision, so that's promising. In fact after a while I forgot my laptop had that problem and was focusing on the nice crisp color display I was getting. Didn't realize I could close the lid and the machine would stay on, so if I can find an old keyboard I can use it like a desktop.

IMG_2402.JPGIMG_2401.JPG
 
  • Like
Reactions: Paolo B and retr01

Paralel

Tinkerer
Dec 14, 2022
115
47
28
The model for the Powerbook 540 is: HOSIDEN MTFT HY4xxxxx27UA. It is a rather unusual panel, in that it was a 6-bit active matrix grayscale panel. The terminology for them at the time was "MTFT" standing for "Monochrome Thin Film Transistor". Unfortunately, at this time, Hosiden was not actually labeling real part numbers, they used a hybrid marking of part and manufacturing identifier. Three documented example of which I am aware are labels on the back of three different panels with the identifier "HY41403727UA", "HY4150BF27UA", & "HY4050FE27UA". Later Hosiden panels have distinct part number labels, as well as manufacturing identifiers. These panels are just about impossible to find outside of actual 540's and NOS mac repair parts. Attached below is a picture of the back of one of these panels.

All the Hosiden models are impacted. It's the seal. From what I understand this was not an uncommon issue with active matrix screens from this time period, apparently the art of sealing all the way down to a small molecular level hadn't been perfected yet.

It's almost certainly moisture since it is one of the few ways to explain why the images in the tunneled area remain on the screen even after power is removed, and even if the actual panel is removed from the system. The water interacting with the electrical element in the screen basically creates a capacitance effect, which continues to drive the screen in the absence of power being externally supplied, which is why the image is retained.

From all the work that has been done on this issue, time and temperature seem to be the critical factors. Temps of around 100C seem to work best, temps that people have tried below that, even for long period of time seem to work, but not as well as the results at 100C. 5 hours also seems to be a good benchmark for time. The only thing that can be done is determine the optimal time using trial and error. 100C or ever so slightly above any water trapped in the screen will want to become vapor, and as a result the vapor pressure will want to force the water vapor out of the nearest exit in the seal in order to expand to its proper volume, essentially acting as a driving force for the water to leave the screen. As for time, as @bwinkel67 has shown, 5 hours works well, so the only things to do is go in both directions from that point until improvement stops or deterioration is seen to determine the best benefit vs. lowest risk. If any results below 5 hours at the predetermined temp quickly show they are inferior, then only more time needs to be tested. Same for more time, if it is quickly shown to be inferior, then only less time needs to be tested. Once the best time is found, the problem will have a best case treatment. After that the only work that needs to be done is to see if there is any way to reinforce the seal so water can't get back in again in the future, and if that can be accomplished, the problem will have a functional cure.

dsc_5801.png
 
Last edited:

Paralel

Tinkerer
Dec 14, 2022
115
47
28
I've got to bake my 540 soon. I'm waiting until I've got a 3d-printed part to fix the hinge before I do that though, I've got to get on that soon.

If that doesn't work out, let me know, I have a literal graveyard of 540's that I've accumulated over the years trying to determine the source of the problem with the screen and how to fix it. So, I have a literal ton of spare Blackbird parts sitting around.
 
  • Like
Reactions: 3lectr1c

bwinkel67

New Tinkerer
Jun 1, 2022
31
14
8
If it is moisture, why does it spread from the outside to almost the entire screen if left on long enough? Wouldn't there be constant volume and it could only impact its area?