Repairing a ST Book.

[stephen_usher]

But why not mention that on the manual?

Because that back-up battery on the motherboard would never go flat during the expected product lifetime so no-one would need to know.

[AndresPlazaR]

Because that back-up battery on the motherboard would never go flat during the expected product lifetime so no-one would need to know.

Ah! You were talking about the internal battery (not the battery pack)... I see, it make sense.

[AndresPlazaR]

Let's continue:

Now that the notebook is disassembled, I inspected the main board carefully. I did not notice any capacitors leaking, and the board seems in pretty good shape. I noticed the internal battery:

s_PXL_20240618_151418979.jpg (311.09 KiB) Viewed 1395 times

and, of course, there was no voltage in the terminals (hey, after more than 30 years...). So I decided to replace it with a battery holder:

s_PXL_20240619_013857128.jpg (379.68 KiB) Viewed 1395 times

(this is the one I bought)
I needed to file some of the corners, so it would fit in between the other ICs. It is easy to do, and holds the two ML2016 batteries.

In the article about the 4 bombs, Robson mentions that the cause is the failure of the hard disk. I removed it and it was a Conner one... I think it is the same brand used in the STacy, and suffering from the same "decease": the seal liquefied and it was leaking. Fortunately, the hard disk bracket held most of it, and there was none in the case or in the main board. The obvious replacement was a CF adapter on a 2.5" HDD form:

s_PXL_20240617_014308836.jpg (326.81 KiB) Viewed 1395 times

(this is the one I bought)

To prepare the CF card I used the Phoenix that I built:

s_PXL_20240619_165225189.jpg (340.38 KiB) Viewed 1395 times

But, even after all that, I was still getting the 4 bombs. Shoot.

While I was trying to think what could be the problem (the official "illegal instruction" meaning is so vague), I decided to change the TOS to the English version (as I mentioned before, this is a German version with a German TOS). I bought this chip and got the English TOS for the ST Book from this website.

Although not directly related to Atari of the ST Book, I should mention that I was using one of those TL866II Plus programmers (with the included adapter for PLCC-44), but it was not reading or writing (just in case, I also wanted to backup the German TOS). It seems that many people had the same problem and there is even a thread about this and how to fix it here in Exxos: viewtopic.php?t=2929 .

And guess what? Replacing the TOS with the new one I made, fixed the booting problem! I'm assuming the original was corrupted) This is how the screen looked while in GEM:

s_PXL_20240622_030459920.MP.jpg (288.78 KiB) Viewed 1395 times

This gives a better idea on what's happening to the screen (the reason why the 4 bombs look weird): the bottom half of the screen shows every other two columns. Here it is a somewhat blurry detail:

closeup.jpg (19.69 KiB) Viewed 1395 times

OK, more reasons to open the screen. That will be the next post.

Andres.

[sporniket]

Looking good.

If I may request something while work is still in progress, it would be a scan of the bottom side and the top side, as it seems that the motheboard would fits inside a A4 scanner ?

[AndresPlazaR]

If I may request something while work is still in progress, it would be a scan of the bottom side and the top side, as it seems that the motheboard would fits inside a A4 scanner ?

Yes, I will do that. Hopefully this week. Will post the picture here.

Andres.

[AndresPlazaR]

Now it is time to address the screen issues.

There are three things I need to do:

1. Check the problem with the vertical lines
2. Check the problem with the bottom part of the screen not showing some vertical lines
3. Find a way to remove the hinges

First, I needed to remove the back cover of the screen (the one that says "Atari"). Robson mentioned that the two parts (the back cover and the front with the LCD) are hold together by plastic tabs along the edge and, starting from either side of the "lock" of the lid, carefully separate the tabs with a plastic tool. I was nervous about this step, but it turns out not that difficult. One thing to keep in mind is to move the support (that attaches the LCD screen to the main body of the computer) and the LCD screen to a "close" position, so the back cover can be remove easily once the tabs are separated. As a reference, here it is a picture and the pink papers show the location of each one of the tabs to be disengage:

s_PXL_20240623_023758145.jpg (390.55 KiB) Viewed 1339 times

Without the back cover, you can see that each hinge has 4 screws (plus the 2 screws each, from the other side). You need to remove all 12 screws, remove the plastic caps on each side (one allows the cables to pass through) and then wiggle each hinge out. Here are some pictures, including one hinge already removed:

s_PXL_20240623_033438323.jpg (252.89 KiB) Viewed 1339 times

s_PXL_20240623_033405596.jpg (309.6 KiB) Viewed 1339 times

So, now, the bottom support can be removed and appreciate the damage:

s_PXL_20240623_134514683.jpg (438.33 KiB) Viewed 1339 times

Those two holes there should have bronze inserts to hold screws of the right hinge. That needs to be fixed with something that will hold the mechanical pressure/stress of the spring of the hinge.

But now I can focus on the electrical problems in the screen.

As mentioned in the article about the weak display and bars, the capacitors of the LCD screen leaked:

s_PXL_20240623_024425006.jpg (305.16 KiB) Viewed 1339 times

s_PXL_20240623_024358483.jpg (269.86 KiB) Viewed 1339 times

s_PXL_20240623_024351761.jpg (282.3 KiB) Viewed 1339 times

So, I disconnected the ribbon cable connecting the screen to the small PCB, removed the metal grid that supports the actual screen to the plastic frame, and very carefully removed the LCD to work on it:

s_PXL_20240629_145650675.jpg (315.05 KiB) Viewed 1339 times

On the back, I could see some brown looking liquid:

s_PXL_20240629_145703175.jpg (254.12 KiB) Viewed 1339 times

Was that also the leaking from the caps? Not sure, but I lifted the clear protection and, using alcohol and a cotton swab, I cleaned it up.

As suggested in the article, the best way to remove those leaking caps is actually cut their terminals as close to the caps as possible, and then remove the terminals from the PCB contacts carefully. After removing, cleaning and installing the new caps, it looked like this:

s_PXL_20240629_145125725.jpg (429.03 KiB) Viewed 1339 times

As a reference, these are the caps I used:

4x https://www.digikey.com/en/products/det ... 220/160547
1x https://www.digikey.com/en/products/det ... 0B/6194660
1x https://www.digikey.com/en/products/det ... 4R7/160576

I did, also, bought these ones to see if they fit:

https://www.digikey.com/en/products/det ... 03/5728008
https://www.digikey.com/en/products/det ... 01/5728006
https://www.digikey.com/en/products/det ... 04/5728044

It turns out that the SMD caps are about half millimeter taller than the through hole ones (when these ones are installed sideways). Because I was lazy and I did not wanted to check if that half millimeter would fit, I just used the ones you see installed.

At this point I tried the LCD screen again, and the caps replacement solved the problem with the weakness of the screen and the vertical lines on the top of the screen, but not the problem of the bottom half of the screen. So, I needed to dig deeper and understand how this LCD screen worked.

The LCD screen is an Epson ECM-A0572. It seems that it was exclusively used for this notebook, because many people have tried to find if other devices used it, but no luck. And, unfortunately, there are no schematics about it.

Looking and looking, I found the datasheet of the Epson SED1355 LCD controller, that made me understand how this LCD works and, possibly, the problem I was having.

It turns out that the LCD screen used in the ST Book is an "8-bit dual monochrome passive LCD": "passive" because it does not have a backlight, "monochrome" because it only has pixels that are on or off. The very interesting part is the "8-bit dual": it means that the data is transmitted using 8 lines (8-bits), but separated in 4 bits for the top half of the screen (the UD0 to UD3 lines from the Shadow IC) and 4 bits for the bottom half (the LD0 to LD3 lines from the Shadow IC). There is a timing diagram that explains how the data is transmitted, but basically it send several 8 bits of data (4 for the top, 4 for the bottom, and each bit represents one pixel) to complete 2 lines (one top, one bottom).

If you look carefully at the closeup picture of my bottom screen, that I posted before, there is a pattern on each line, showing two pixels and then not showing two pixels... So, a pattern of 4 bits.

Looking at the PCB of the LCD screen, there are 4 ICs: one Epson, one C324G (and amplifier, I'm guessing to control the brightness of the screen) and two 74HC365... one located to closer to the top of the screen, and one closer to the bottom. Tracing the LD lines, they all go to IC2, that looks very crusty, I assume because of the leaking of the caps close to it.

Using the oscilloscope, this is what I saw (yellow is the input to IC2, blue is the output from IC2; they should be the same):

LD0:

IC2_LD0.jpg (90.79 KiB) Viewed 1339 times

LD1:

IC2_LD1.jpg (87.71 KiB) Viewed 1339 times

LD2:

IC2_LD2.jpg (85.89 KiB) Viewed 1339 times

LD3:

IC2_LD3.jpg (89.99 KiB) Viewed 1339 times

There you have it: evidence that IC2 is not working properly. I used kapton tape to protect everything around it and, using a reworking station with hot air, removed it, cleaned the area:

s_PXL_20240705_023910556.jpg (254.23 KiB) Viewed 1339 times

and installed this part: https://www.digikey.com/en/products/det ... NSR/476057. It worked!!!

s_PXL_20240713_140829285.jpg (233.1 KiB) Viewed 1339 times

If anyone is experiencing a similar problem, but in the top part of the screen, you need to replace IC3 (it is not labeled in the PCB, but it is the IC above the JP1 label):

s_PXL_20240705_154801124.MP.jpg (223.34 KiB) Viewed 1339 times

I think the only thing left to test, related to the electronics, is the keyboard. I know that some keys work, but I have not tested all of them.

And now I can focus on repairing the case.

So, at this point, we are up to date on the repairs. That means that I will only update when there is something interesting to report.

Andres.

[sporniket]

good job

If I may request something while work is still in progress, it would be a scan of the bottom side and the top side, as it seems that the motheboard would fits inside a A4 scanner ?

Yes, I will do that. Hopefully this week. Will post the picture here.

Andres.

Thanks.

[HigashiJun]

Well done !

The screen fully lives again.

[troed]

Great work bringing back these immensely rare and historically important devices back to life!

[mikro]

What a great thread. Congrats on your progress!