You are using a standard 486DX - right? It runs with 5V core voltage so the voltages at the regulator transistor don´t matter at the moment.
The voltage drop from 5V to 2.7V indicates a serious problem with your PSU. Did you try different PSUs?
EnergyStar wrote on 2022-07-11, 21:11:I replaced the UM8663AF, UM8667 chips. Swapped UM8002. Nothing got worse but nothing helped either. […]
I replaced the UM8663AF, UM8667 chips. Swapped UM8002. Nothing got worse but nothing helped either.
The only thing that I noticed is that when the board does not start and hangs on the C7, the power supply on all systems drops from 5V to ~ 2.7V.
2.7V on the +5V line is way to low to expect any proper operation. 4.5V is absolute minimum for guaranteed operation, and everything below 4.0V surely makes the system fail. If +5V drops that low, you either have something shorting out the +5V line, or the +5V line on the board is broken near the AT power supply connector, which is not that uncommon on boards with battery leakage damage. Usually, 5V is carried on an inner layer of the board, so I suggest trying to resolder all the +5V pins of the AT connector (best to remove most of the old solder first), and using deox-IT or a similar anti-corrosive agent on the AT power supply connector pins themself.
Did you try different PSUs?
Yes, I tried on a different power supply - no changes. Still the voltage drops to 2.7V.
The power supply works fine on working motherboards.
Voltages measured on plugs with inserted graphics card, diagnostic card, RAM, processor, HDD and FDD and keyboard:
5.1V
-4.8V
12.1V
-12.0V
I also noticed that very rarely the 12V LED on the diagnostic card does not light up, but the computer boots. I have also noticed the rare cases when the monitor turns off after a short time.
I checked the jumpers again, replaced the cache from another motherboard - no changes.
I swapped the next two IC's. Nothing changes.
Sometimes it stops at C7, sometimes at E1, sometimes no indication on the diagnostic card. When it boots up normally, it sometimes gets into a restart loop with this message at the bottom.
Sometimes while saving BIOS settings, it freezes (only white bars at the top and bottom can be seen) displaying C7. It happens that the keyboard does not work fully in the BIOS, sometimes it does not work at all and the LED does not light up.
When error C7 won't go away, removing and reinserting the Dallas chip helps.
Usually, 5V is carried on an inner layer of the board, so I suggest trying to resolder all the +5V pins of the AT connector (best to remove most of the old solder first), and using deox-IT or a similar anti-corrosive agent on the AT power supply connector pins themself.
I replaced the entire power socket with another in better condition - no changes.
However, all these symptoms were present from the very beginning. So much work and no progress ..
At least any ATX power supply would be shut down by it´s own protection citcuit immediately if the 5V rail had an overload bringing down the output voltage from 5V (nominal) to 2.7V.
What PSUs have you tried? And do you measure the 2.7V voltage directly at the AT connector?
EnergyStar wrote on 2022-07-15, 23:38:I replaced the entire power socket with another in better condition - no changes.
In that case, possibly the inner layer itself is corroded at the soldering points for the power socket, so the +5V pins at that socket dont make reliable contact to the power plane. As other posters already commented, every decent power supply would go into short circuit protection mode if +5V is pulled down to below 3V in the power supply, so we can be quite sure that the drop from +5V to +2.7V happens either at the power socket of the mainboard or on the mainboard itself. As you already swapped the power socket of the mainboard and tried different supplies, it is very likely that the drop happens on the mainboard. If I am right on that diagnosis, pulling some not too skinny bodge wires from +5V pins of the AT power connector to +5V pins elsewhere on the board (like the ISA +5V pins) might cure the problem.
What PSUs have you tried?
This is the power supply I use the most. I will post a photo of the second one soon.
And do you measure the 2.7V voltage directly at the AT connector?
As far as I remember, 2.7V was on both the plug and the entire motherboard. Unfortunately, now it is difficult for me to capture the moment when this voltage drops to verify it 100%.
If I am right on that diagnosis, pulling some not too skinny bodge wires from +5V pins of the AT power connector to +5V pins elsewhere on the board (like the ISA +5V pins) might cure the problem.
I ran a thick cable from the + 5V power supply to the + 5V power supply of the ISA port from the solder side - no changes. Even when the motherboard stops at E0 / E1 / C7, mostly 5V is present on the chips.
I noticed that when BIOS automatically sets the date to 1985. then you cannot change the date but the arrows work and you can navigate through the BIOS settings. Sometimes the date is automatically set to 1995 (for some reason) and then it can be changed with the +/- keys. FDD's LED is on (properly not constantly) you can hear it running during POST, then when you get the message "disk not present, pres any key .." pressing "Enter" LED is on and it spins the floppy but does not read the boot sector. The amount of CACHE memory is suddenly given correctly, sometimes it is not recognized at all. Sometimes it recognizes the hard drive automatically, sometimes not even if the FDC Failure message is not displayed. The keyboard often does not work until the Keyboard Error message appears, but enters the BIOS when F1 is pressed. All in all, very different and strange things are happening within the disk controller, BIOS, keyboard, power supply, and CACHE memory. I don't know what you guys think, but maybe the problem is simply a damaged chipset. The problem is with most of the things that hold them together. It is a pity that I do not have a replacement chip.
Is there any way to check if the chipset is working properly?
That´s not an easy decision.
If you can rule out all the common problems like damaged traces, loose pins, corroded contacts, broken resistor arrays and everything that you can discover using your eyes, there´s often only the chipset left (one of it´s chips to be correct). local selective heating and cooling can lead you to the culprit if you´re lucky.
If you want to get to the heart of the matter and consistently proove what´s going wrong, you have to do an examination like this:
Debugging a Micronics W6-LI POST failure
If that´s not an option you need to test the chips by replacing them one by one.
