Deunan wrote on 2025-06-29, 12:24:

Remove co-processor from socket. Re-seat (remove and put back) the BIOS ROM chip in the socket. Put some RAM into the board. Although it should output some codes even without RAM, maybe it will help. In general having RAM populated should not prevent the mobo from booting, even if the RAM is damaged (as long as it's not completly shorted).

Check all the CPU jumpers again. If the on-board CPU is not disabled properly when the 486 is added, or if it's not enabled with 486 missing, obviously the mobo will not run properly.

Putting the POST cards backwards might have routed 12V in weird places, and killed something. Hopefully not the chipset itself. But as I've said if there are no POST card codes, then you need a scope of some sort to probe around the mobo for signals. There are some cheap scopes that might work well for this task, you won't really need anything more than 40MHz, and many things around ISA bus don't go above 10MHz.

As i said earlier, all the jumpers were left like before this "accident", and the mbo worked well. I tried to reseat BIOS chip, will do the same with Co-processor. It stopped working before i tried with reverse POST card 🙁

zuldan wrote on 2025-06-29, 12:09:

Just a wild guess here…sometimes motherboards require the battery to have the proper voltage before posting. Can you try get the voltage reading of the battery by turning over the motherboard. Maybe the battery finally died.

Battery is on 2.89..... i don't know, this seems to be enough maybe.... and maybe not
Will try with "CR2032 mod" 😀 (4th pin -, 8th pin +) and let You know.

I think i'm ready for some mental institution!!! Enough for today, i did connect CR2032 to Dallas battery, and now it is 3,2V, but i have connected Dallas battery wrong way into socket, and of course, there was some cracking sound before i managed to turn it off. Did remove the battery for some minutes, there is still 3,19v on pin 8 and 4, but who knows what else i have done to the board now.....
I think that is enough for me today, didn't slept for 2 days, did so much damage, i don't know what else i could do if i continue....
Will be asking for help when i get myself together....
Thanks to all for replies, if someone had any idea what to do, i would be grateful....

After a good weekend, and fixing Olivetti M290, i think i'm ready to try again with my love, Alaris Cougar. I will try to stay calm, not to hurry anymore, and read twice every advice of Yours. Does anybody have a clue where to start, no codes on POST card? Can we do something with multimeter, or do i need scope for this?
Thanks in advance!!!!

butjer1010 wrote on 2025-07-07, 17:32:

After a good weekend, and fixing Olivetti M290, i think i'm ready to try again with my love, Alaris Cougar. I will try to stay calm, not to hurry anymore, and read twice every advice of Yours. Does anybody have a clue where to start, no codes on POST card? Can we do something with multimeter, or do i need scope for this?
Thanks in advance!!!!

You can check to make sure the CPU socket is getting the required voltage. Just start start around the power regulators and then work into the socket.

That and checking for trace damage is all you can really do, Necroware has some good repair videos though I'm not sure I've seen one where he did a repair from no post codes.

https://www.youtube.com/watch?v=3wMpnzfVc4Y

SScorpio wrote on 2025-07-07, 18:13:

You can check to make sure the CPU socket is getting the required voltage. Just start start around the power regulators and then […]
Show full quote

butjer1010 wrote on 2025-07-07, 17:32:

After a good weekend, and fixing Olivetti M290, i think i'm ready to try again with my love, Alaris Cougar. I will try to stay calm, not to hurry anymore, and read twice every advice of Yours. Does anybody have a clue where to start, no codes on POST card? Can we do something with multimeter, or do i need scope for this?
Thanks in advance!!!!

You can check to make sure the CPU socket is getting the required voltage. Just start start around the power regulators and then work into the socket.

That and checking for trace damage is all you can really do, Necroware has some good repair videos though I'm not sure I've seen one where he did a repair from no post codes.

https://www.youtube.com/watch?v=3wMpnzfVc4Y

Ok, thanks, will start with that! Need to find pinout of this CPU first....

I have several of these boards - they are flaky in multiple ways.
All of them at some point exhibited the next problem - BL3 CPU pins get disconnected from the pads underneath, no matter how carefully the board has been handled.
Use a sharp tool to carefully examine the pins by gently pushing them sideways to see if any are loose.
Actually, a better option is to just switch the board to use a 486 CPU and bypass the BL3 one. It is one jumper change - the jumper on the right side of the CPU socket. Just make sure you use a 5V 486 or Pentium Overdrive.

pshipkov wrote on 2025-07-07, 22:01:

I have several of these boards - they are flaky in multiple ways. All of them at some point exhibited the next problem - BL3 CPU […]
Show full quote

I have several of these boards - they are flaky in multiple ways.
All of them at some point exhibited the next problem - BL3 CPU pins get disconnected from the pads underneath, no matter how carefully the board has been handled.
Use a sharp tool to carefully examine the pins by gently pushing them sideways to see if any are loose.
Actually, a better option is to just switch the board to use a 486 CPU and bypass the BL3 one. It is one jumper change - the jumper on the right side of the CPU socket. Just make sure you use a 5V 486 or Pentium Overdrive.

Hi,
i tried with another CPUs, but i'm not sure if those 2 CPU are working at all. I need to test them on working board to be sure. It didn't change anything.
Thanks for the tip, i will try to examine CPU pins 😉

Tried to find cold solder joint or CPU pin leg "detached", but no luck. Now i was back to resistors, and found 3 resistors that aren't ok IMO. RP13, RP 19 and RP24. Now i'm struggle to find something like that - 13 have writing AFL7K O, and 19 and 24 have AF10K O.
On RP13 first (common leg) and leg 2,3,4 and 5 have resistance of 4,67KO, and first and 6,7 and 8th leg have resistance of 3,2KO.
On RP19, first and all the legs except 7th, have normal resistance of 10KO, but first and 7th leg have resistance of 3,64KO.
On RP 24 first and 2,3,and 4th leg have 3,2KO, and first and 5,6,7,and 8th leg have normal resistance of 10KO.
Did i measure that also wrong, or could it be that the resistors are failing, resulting dead motherboard?

you're wasting your time with this resistor stuff, through hole passives almost never just "fail" unless they're burnt up or corroded or something

Too bad it was not this.
Another place to probe is the VRM between the two VLB slots (on their left side). Just be careful with the probes because it can burn by shortening the 12v line (pins are very closely packed together).

maxtherabbit wrote on 2025-07-08, 13:15:

you're wasting your time with this resistor stuff, through hole passives almost never just "fail" unless they're burnt up or corroded or something

I have fixed around 5 Voodoo cards changing faulty resistors arrays, so i thought maybe this would be easy job like those cards 😀

pshipkov wrote on 2025-07-08, 14:47:

Too bad it was not this.
Another place to probe is the VRM between the two VLB slots (on their left side). Just be careful with the probes because it can burn by shortening the 12v line (pins are very closely packed together).

You meant upper one from those two i circled on picture?

4th and 6th pin are ground and vcc, as You said, i need to be very careful, pins are as close they can be. Can I measure anything else except 4th and 6th pin (maybe is better to press one probe on tab instead 4th pin)?

any gnd will be fine.
i sometimes stick the black probe into a black cable slot of the AT connector, then place the red probe in the corner of the regulator where the pin enters it the plastic body. this way my entire attention is on keeping one hand steady.

as silly as it sounds after burning 2-3 of these out of momentary negligence now i know better.

pshipkov wrote on 2025-07-08, 19:01:

any gnd will be fine.
i sometimes stick the black probe into a black cable slot of the AT connector, then place the red probe in the corner of the regulator where the pin enters it the plastic body. this way my entire attention is on keeping one hand steady.

as silly as it sounds after burning 2-3 of these out of momentary negligence now i know better.

Hi to all,
I was so busy for last month, i didn't want to destroy anything else. so i waited until finally I have more time to deal with this board!
I tried Your method, and was really easy - black probe into ground wires on AT connector, and than was really easy to put the red probe to the 6th pin! I got readings of 1.67 - 1,68V? Is this ok, or could the problem be there? 1.68V sounds pretty weird to me. Pin 1 shows 12.03V (input).
Id description of that LT1076CR says output should be from 2.5 - 50V ?

Pin 1 is OK, this is input. This is a switching mode step-down (buck) regulator though and Vc (pin 6) is just some stabilty network. The voltage value here is not really important (unless it shows GND, input or output) but it must be stable.

That octagon-shaped thing below the regulator chip is a coil, but seems to have more than 2 leads. Perhaps it's a split winding for some reason? One pin of that coil should be connected to pin 2 of the regulator. That's the switching side. One other pin of the coil will be output, this one should go to the CPUs. Note you need to verify the connections with your eyes, using ohm meter here will be almost useless as the coil should have very low resistance. You want to:
1) verify that coil output does indeed go to the CPU (the BL3 and also most likely the PGA socket)
2) make sure the output has correct voltage, it's probably hardwired to 3.3V

That violet electrolytic cap near the chip and coil is probably the output filtering, might be easiest to first check the voltage there, between the leads.

As You said, all 4 leads of this coil "beeps". Upper and right are connected to pin 2 on regulator, and lower pin is going thru this thick via to this violet cap near cpu. There is a 4.01V on positive lead of cap. This cap is connected to at least 15 pins on CPU

butjer1010 wrote on 2025-07-08, 16:56:

maxtherabbit wrote on 2025-07-08, 13:15:

you're wasting your time with this resistor stuff, through hole passives almost never just "fail" unless they're burnt up or corroded or something

I have fixed around 5 Voodoo cards changing faulty resistors arrays, so i thought maybe this would be easy job like those cards 😀

right, SMD resistor arrays tend to crack from board flex or other physical stress - through hole ones don't

butjer1010 wrote on 2025-07-30, 12:53:

As You said, all 4 leads of this coil "beeps". Upper and right are connected to pin 2 on regulator, and lower pin is going thru this thick via to this violet cap near cpu. There is a 4.01V on positive lead of cap. This cap is connected to at least 15 pins on CPU

4.1V might be good enough, the mobo will generate a bit higher voltage to offset any drop in the traces (the regulator is quite far from the CPU) and it also helps the stability a bit - at the cost of higher CPU temperature.

You can check the connection to the PGA socket, use this page to find the right pins: http://ps-2.kev009.com/eprmhtml/eprmx/h12203.htm
I'm not 100% sure now if that is view from the top or bottom but it should be the top. All the red dots should have the either 5V or that 4.1V from regulator, depending on CPU setup jumpers. There is also pin J1, that one should always have 5V on it, even for low-voltage CPUs. This is the I/O reference voltage.