Some motherboards require healthy bios battery to post. So that is worth a shot.

If the label on the bios chip is any good then it seems to be a newer bios version than is available on theretroweb.
Getting a dump of that would be appreciated.

Agreed on the testing with a cmos battery first.
Also check to see what post code it gives without any ram and different ram if possible.

Also, add a speaker to listen for beep codes.

A few longshots:
Try another BIOS chip with an older BIOS version, even if you successfully dumped the current one.
Check the frequency of the XTALs.
Check activity on ISA data and address lines.

- Xtals and osc are fine.
- CMOS battery, circuit ok but nothing change.
- Speaker, total silence.

I will try with BIOS images (https://theretroweb.com/motherboards/s/ecs-wh386sx) and (https://theretroweb.com/motherboards/s/ecs-wh386sx-wh386sx-d).

Since I can dump the BIOS that came with the motherboard, I don't think that's the problem, but I'll try.

I think the motherboard is dead...

RafaelBR wrote on 2026-02-07, 20:28:

Since I can dump the BIOS that came with the motherboard, I don't think that's the problem, but I'll try.

Please do dump the bios.

In the photos the post code it gets stuck on is 41. Does this come on immediatly or does it show other post codes before it gets stuck on that and is it the same without any ram ?

RafaelBR wrote on 2026-02-07, 16:12:

- I tried turning it on without the cache ICs.

Was here also a jumper you had to change to tell it no cahe?

Is there a short on the reset switch (i.e. does the board think the reset switch is constantly pressed in?

weedeewee wrote on 2026-02-07, 21:02:

In the photos the post code it gets stuck on is 41. Does this come on immediatly or does it show other post codes before it gets stuck on that and is it the same without any ram ?

Error code 41 occurs the exact moment the board is powered on.

The previous code is 01.
With or without memory, the same thing happens.
The IRDY LED lights up and goes out immediately...

I checked the board with a thermal camera; only the W83C312F doesn't show any heat. I couldn't find a datasheet for it, but I can detect that it's being powered with 5V.

NeilKnows wrote on 2026-02-07, 21:13:

Was here also a jumper you had to change to tell it no cahe?

No, only size. 64 ou 32Kb

With or without cache, it's the same thing. It could be a faulty cache IC, but it's not impossible.

NeilKnows wrote on 2026-02-07, 21:13:

Is there a short on the reset switch (i.e. does the board think the reset switch is constantly pressed in?

There is no short in the reset. When the reset button is shorted, the reset LED on the diagnostic board lights up, resets the numbering, and returns to 41...

If I understand you initial post correctly, the board died approximately at the moment you connected the keyboard. It sounds like you connected the keyboard while the system was powered on. This is an operation people tell you not to do, because there is no proper power sequencing (GND and +5V pins making connection before the data pins) on the AT keyboard connector. While I also did connect keyboards while a system was powered on without ill effects, the danger oftentimes quoted: "Connecting an AT keyboard or a PS/2 mouse with the system turned on will likely damage your system" possibly sometimes is real. The keyboard connects to the Intel 8042 auxilliary processor IBM introduced with the IBM AT, which is usually called "keyboard controller", but actually does some more auxilliary jobs controlling the AT platform, and is essential for a well-behaving system. It's possible that the issues you are facing now are due to a broken keyboard controller.

If you have another 286/386/486 board that does not have a PS/2 mouse port or a pin header for a PS/2 mouse port, try using the keyboard controller from that board instead of the keyboard controller in your non-functional board.

mkarcher wrote on 2026-02-07, 22:29:

If I understand you initial post correctly, the board died approximately at the moment you connected the keyboard. It sounds like you connected the keyboard while the system was powered on. This is an operation people tell you not to do, because there is no proper power sequencing (GND and +5V pins making connection before the data pins) on the AT keyboard connector. While I also did connect keyboards while a system was powered on without ill effects, the danger oftentimes quoted: "Connecting an AT keyboard or a PS/2 mouse with the system turned on will likely damage your system" possibly sometimes is real. The keyboard connects to the Intel 8042 auxilliary processor IBM introduced with the IBM AT, which is usually called "keyboard controller", but actually does some more auxilliary jobs controlling the AT platform, and is essential for a well-behaving system. It's possible that the issues you are facing now are due to a broken keyboard controller.

If you have another 286/386/486 board that does not have a PS/2 mouse port or a pin header for a PS/2 mouse port, try using the keyboard controller from that board instead of the keyboard controller in your non-functional board.

You're correct. The keyboard was connected afterward. I'll see if I can get another keyboard controller to test, but could some other component have been damaged? I wasn't aware of this problem, so it's a valuable learning experience.

mkarcher wrote on 2026-02-07, 22:29:

If I understand you initial post correctly, the board died approximately at the moment you connected the keyboard. It sounds like you connected the keyboard while the system was powered on. This is an operation people tell you not to do, because there is no proper power sequencing (GND and +5V pins making connection before the data pins) on the AT keyboard connector. While I also did connect keyboards while a system was powered on without ill effects, the danger oftentimes quoted: "Connecting an AT keyboard or a PS/2 mouse with the system turned on will likely damage your system" possibly sometimes is real. The keyboard connects to the Intel 8042 auxilliary processor IBM introduced with the IBM AT, which is usually called "keyboard controller", but actually does some more auxilliary jobs controlling the AT platform, and is essential for a well-behaving system. It's possible that the issues you are facing now are due to a broken keyboard controller.

If you have another 286/386/486 board that does not have a PS/2 mouse port or a pin header for a PS/2 mouse port, try using the keyboard controller from that board instead of the keyboard controller in your non-functional board.

given the post code it was stuck at (A20 line was controlled by KBC on older mbs), I was wondering if that could be a cause but I missed the part about connecting the keyboard after power up.

Thanks for the explanation & nice catch !

Very rare board you have there!

On all old motherboards, all chips should be reseated due to oxidisation on the pins making bad connections. RAM slots should be cleaned with isopropyl alcohol & a toothbrush (there are no beeps/indications from oxidised RAM slots).

Some motherboards need an IDE card inserted to start up.

Check all lines near the power connector & keyboard controller.

If you don't have a 10uF capacitor at the power lines anymore, voltage dips may cause some chips not to start at the right time.

RafaelBR wrote on 2026-02-07, 22:49:

I'll see if I can get another keyboard controller to test, but could some other component have been damaged?

The keyboard is connected to +5V (usually via a fuse and/or inductor), GND (directly) and two pins of the keyboard controller directly, as well as a driver chip used to send data/clock to the keyboard, typically a 7406. It is not connected to anything else on the board. Assuming the most likely cause of damage, that is a data pin connected first and a considerable amount of electrostatic charge was dissipated through the keyboard controller or the 7406, I don't think anything else might have taken damage. If the keyboard still gets power (LEDs blink on power-up), the +5V connection works. There are inductors, capacitors and pull-up-resistors on the data and clock line, which theroretically can be damaged as well, but I wouldn't expect them to be as susceptible to ESD as the integrated circuits.

All components except the KBC are not supposed to crash the board that early during POST. If you are getting messages like "keyboard missing", "keyboard error", "keyboard clock/data line stuck" with a different keyboard controller, you need to continue troubleshooting the components I mentioned. In my oppinion, there it's not usefull to focus on possible secondary damage right now.

Have you checked for broken traces?
Sometimes I missed some.

Tried a different 7406, tried this one in another mobo?

Checked for bridges on system ICs, after soldering it happens and they are really really tiny.
Leads are soldered for real on pads? I had many get unstuck easily, i.e. no real connection or flaky.

Continuity on Address with ISA slots, Bios, 7406, RTC?
Cpu, IIRC, go through some 74Lxxx, Address + other go through RPx, if one is broken it could go to ground and kill signals, or just be broken.
I had that once.

Not the best guide, sorry 😀