restoring a battery-bombed IIcx

[Phipli]

I added a few drops of oil to the bearing in the IIcx power supply fan. I don't think it made any difference though. Check out the before and after and tell me if this sounds normal.

before:

after:

Well, it's spinning faster

I usually run them for a while face down (out of the case if the case doesn't really allow for it) so the oil can work into the dry grease.

It usually works, my LC fan went from loud to near silent.

 

[bigmessowires]

I'm still plugging away on this IIcx. I think it's nearly finished. Two more repairs:

I desoldered UM2 - the 74HC132 in the startup circuit that controls power-on behavior. There was a whole lot of funk under there. I cleaned the pads, and replaced it with a new chip. That cured my cold-start instant power-on weirdness! I removed my previous fix/hack with the extra resistor.

I also took another look at the bad RAM that was originally in the machine, which were 4 x 4MB SIMMs. Now that I also have a working LCII which can use the same RAM but only needs two SIMMs instead of four, I realized I could test different pairs in the LCII to identify which one of the four SIMMs was the bad one. No surprise: it was the one with the visible corrosion, that had been sitting in a puddle of battery goo. I tested for continuity between the SIMM pads and the RAM chip pins and found five pins on one chip with no connection. After lots of cleaning and scraping and fluxing and resoldering, the bad SIMM is working again! That gives me a complete set of 4 x 4MB which is really nice.

 

[Phipli]

I'm still plugging away on this IIcx. I think it's nearly finished. Two more repairs:

I desoldered UM2 - the 74HC132 in the startup circuit that controls power-on behavior. There was a whole lot of funk under there. I cleaned the pads, and replaced it with a new chip. That cured my cold-start instant power-on weirdness! I removed my previous fix/hack with the extra resistor.

I also took another look at the bad RAM that was originally in the machine, which were 4 x 4MB SIMMs. Now that I also have a working LCII which can use the same RAM but only needs two SIMMs instead of four, I realized I could test different pairs in the LCII to identify which one of the four SIMMs was the bad one. No surprise: it was the one with the visible corrosion, that had been sitting in a puddle of battery goo. I tested for continuity between the SIMM pads and the RAM chip pins and found five pins on one chip with no connection. After lots of cleaning and scraping and fluxing and resoldering, the bad SIMM is working again! That gives me a complete set of 4 x 4MB which is really nice.

20MB is absolutely perfect for a IIci running 7.1 or 7.5.3

They're excellent boxes. Extremely expandable. They sold so many that they are the machine that most accelerators target, so you can upgrade it fairly easily with a fast 030, an 040 or a PPC. Getting something is easy, but they're pricy these days.

 

[bigmessowires]

After a crazy amount of time struggling to clear the through-hole for the negative terminal of the battery holder, I finally succeeded with the help of the needle desoldering tools that @macuserman recommended. Thanks for the tip! I got the new battery holder installed, and also the all-important blue LED for power. The blue LED is tough to photograph, because it's so bright that it blows up the photo exposure and looks mostly white. When seen in person it's a vivid blue. I think it makes the IIcx faster, although benchmarks don't show it.



That's the end of the rework for this IIcx. I verified that it still boots and runs normally, now it's time to give it a bath and then install everything back into the case.

 

[jmacz]

I think it makes the IIcx faster, although benchmarks don't show it.

 

[cgp]

I think it makes the IIcx faster, although benchmarks don't show it.

I believe blue LEDs post-date vintage benchmarks. They were certainly developed well after the IIcx. So the benchmarks are blind.

 

[macuserman]

I finally succeeded with the help of the needle desoldering tools that @macuserman recommended. Thanks for the tip!

Awesome! I’m glad they worked for you too!

 

[JC8080]

Interesting hand writing here... where did this machine come from?

View attachment 63859

I wouldn't bet my life on this, but... WDL is the normal abbreviation for Washington Driver's License, and that string of numbers and letters follows the pre-2018 drivers license numbering format.

First 5 letters of last name, first name initial, middle name initial, 2-digit number consists of 100 minus birth year, then 3 checksum digits (1 number followed by 2 letters).

I have seen DL numbers engraved on the outside of a computer before, I suppose people have also put them on the inside.

 

[bigmessowires]

So I have the Mac IIcx formerly belonging to R. L. Dorma-xxxx, born 1952, resident of Washington state, now age 71? The most likely completion is Dorman. I need to find this guy and give him hell for not removing the battery before he put the computer into storage.

EDIT: according to this explanation of the algorithm, his birthdate is March 13 1952.

 

[JC8080]

So I have the Mac IIcx formerly belonging to R. L. Dorma-xxxx, born 1952, resident of Washington state, now age 71? The most likely completion is Dorman. I need to find this guy and give him hell for not removing the battery before he put the computer into storage.

EDIT: according to this explanation of the algorithm, his birthdate is March 13 1952.

Or buy him a beer for not tossing it into the trash 20 years ago! And then a stern scolding on the battery.

 

[bigmessowires]

That's true, I like your positive spin on not dumping the Mac in the trash. I'm pretty sure I found the guy - all the info checks out and I have his address and phone number, but no email. I don't think he probably wants a phone call from a stranger asking about the Mac IIcx that he owned 30 years ago, so I'll probably let it be.

 

[Mk.558]

I would consider it respectful to obscure his info on the internet. While I am not him and cannot speak for him, I wouldn't want my info plainly plastered on the internet. The name I wouldn't worry about so much, but the birthdate would.

Good to see you got it working. How does that blue LED work out? I once had some trashy external speakers that had a harsh clear blue light for the power indicator, and I eventually swapped in a red LED with a resistor because it got on my nerves. Soft diffused blue light doesn't bother me much though.

 

[Sideburn]

I just found this thread so I am late to the party and need to read up but I have repaired two startup circuits already and have worked on / am working on 3 iicx boards. One is so bad on the battery area that it basically my needs to be redone completely. Mostly startup circuit in that area.

My first board that wouldn’t startup, I just bypassed the whole circuit first with an ATX power supply and re capped the board and fixed broken traces until I got some activity out of it.

The issue with the startup circuit was a bad trace of three and a bad chip. I can let you know which ones and what areas.

 

[bigmessowires]

Good point. The edit window is closed but I'll ask the mods to obscure that part of R.L.D.'s info. The blue LED does seem quite bright, but I haven't actually tried it reinstalled into the case yet. There's a current-limiting resistor on the bottom side of the motherboard under the LED, so you can swap that for a higher-value resistor to make it less bright.

The soft-power startup circuit is way too clever for its own good: three ICs, six capacitors, four diodes, two transistors, and ten resistors all just so you can turn the computer on and off. A hard power switch looks pretty good in comparison. Bypassing the whole circuit is a great idea if you simply can't get it to work.

 

[JC8080]

Good point. The edit window is closed but I'll ask the mods to obscure that part of R.L.D.'s info. The blue LED does seem quite bright, but I haven't actually tried it reinstalled into the case yet. There's a current-limiting resistor on the bottom side of the motherboard under the LED, so you can swap that for a higher-value resistor to make it less bright.

The soft-power startup circuit is way too clever for its own good: three ICs, six capacitors, four diodes, two transistors, and ten resistors all just so you can turn the computer on and off. A hard power switch looks pretty good in comparison. Bypassing the whole circuit is a great idea if you simply can't get it to work.

I also wouldn't object if they edited my post to make it vaguer or remove it entirely. I was not thinking of the potential privacy issues at the time.

 

[bigmessowires]

Well... a step backwards with this IIcx. I washed the motherboard with hot soapy water, then submerged the whole thing in 99% isopropyl alcohol for about 20 minutes while I scrubbed with an ESD-safe brush. I then set the motherboard on a floor furnace vent for 24 hours to dry in the warm air. Once it was dry, I installed the 4 x 4MB SIMMs that I fixed and tested yesterday in my LCII.

Now the IIcx no longer boots. It turns on and makes some distorted buzzing tones for a few seconds - something that might be garbled chimes of death. At first they were normal volume, but over repeated booting attempts the noises became fainter until they're mostly inaudible now.

I'm thinking maybe I didn't wait long enough for the motherboard to dry, even with it sitting on the heat vent, and there's still some water trapped in the sound chips or elsewhere. Any other brilliant suggestions? I can also remove the 4 x 4MB RAM and return to the 4 x 1MB that I had before, but the RAM clips are brittle and I don't like to touch them any more than necessary.

 

[Phipli]

Well... a step backwards with this IIcx. I washed the motherboard with hot soapy water, then submerged the whole thing in 99% isopropyl alcohol for about 20 minutes while I scrubbed with an ESD-safe brush. I then set the motherboard on a floor furnace vent for 24 hours to dry in the warm air. Once it was dry, I installed the 4 x 4MB SIMMs that I fixed and tested yesterday in my LCII.

Now the IIcx no longer boots. It turns on and makes some distorted buzzing tones for a few seconds - something that might be garbled chimes of death. At first they were normal volume, but over repeated booting attempts the noises became fainter until they're mostly inaudible now.

I'm thinking maybe I didn't wait long enough for the motherboard to dry, even with it sitting on the heat vent, and there's still some water trapped in the sound chips or elsewhere. Any other brilliant suggestions? I can also remove the 4 x 4MB RAM and return to the 4 x 1MB that I had before, but the RAM clips are brittle and I don't like to touch them any more than necessary.

Drown it in IPA again for 5 minutes, then somewhere with good ventilation, blast it with a hairdryer that has a cold or not very warm setting. Go over it carefully inch by inch.

24 hours was probably plenty long enough, so I suspect something marginal has failed. I'd start by inspecting QFP chip pins - are any out of place?

Does the CPU get warm?

Are the clocks running?

Does the reset line behave as expected...

Feeling of deja vu

 

[bigmessowires]

Yes, definitely a feeling of deja vu! Perhaps I should have left it alone and not washed it, but I was concerned about hidden battery goo that might cause further problems in the future.

Fortunately all is not lost. After removing the 4 x 4 MB RAM and replacing it with 4 x 1 MB, the computer boots and runs, but the sound remains very quiet and scratchy. Probably there's still trapped water under the sound chips, or I damaged something there... at least it's a localized problem that I should be able to fix.

For the 4 x 4 MB RAM, it works in the LCII so I'm not sure why it doesn't work here. I can try some deoxit on the RAM SIMM sockets and the RAM itself. If that doesn't work then I'll just go with the lower RAM amount.

 

[bigmessowires]

I have the 4 x 4MB installed in bank A, and 4 x 1MB installed in bank B, which I think is the correct way to do it.

 

[Phipli]

For the 4 x 4 MB RAM, it works in the LCII so I'm not sure why it doesn't work here. I can try some deoxit on the RAM SIMM sockets and the RAM itself. If that doesn't work then I'll just go with the lower RAM amount.

Possibly a dodgy higher address line that isn't used by 1MB SIMMs.

I have the 4 x 4MB installed in bank A, and 4 x 1MB installed in bank B, which I think is the correct way to do it.

I believe it doesn't matter in the IIcx really. Not enough to cause issues.