Post pics, in HQ, of your board.

Does Reset work correctly?
IRDY and FRAME?

Why would you replace keyboard controller or RTC IC?

Most of the issues with 286 boards that I experienced were either
- fatal (hidden) damage caused by leaked battery,
- unsupported RAM configuration (some boards support only modules with parity),
- damaged BIOS,
- or "forgotten" turbo switch.

Yes, I didn't send one of the most important informations: battery did not leak and still have 2,8V. Photo will be today, later.
Reset is working like it should (after turning on reset is high for a moment). I don't know IRDY and FRAME should behave.
I used SIMM modules with three and 9 chips and I also used separately 4256 chips (8 pieces). Some motherboards have jumper to turn on or off parity but not this one.
I replaced keyboard controller and RTC IC just because they are in sockets.
"Forgotten turbo switch" - what does it mean? Does it matter whether turbo is on or off? On my motherboard is off for now because I didn't connect switch or mount any jumper.

I guess it's probing time.
Gonna take huge amounts of time.

Basically you have to determine if all the signals are good, if voltages are correct and if all components get the right supply.
Before that check thoroughly for broken traces, it could save you hours.
If all components are properly soldered, no hanging legs or broken solder joints.

Turbo switch gives the hi or lo start speed, sometimes it must be jumpered to start hi (full speed - not your problem now).

pvlst wrote on 2025-01-19, 18:34:

- fatal (hidden) damage caused by leaked battery, - unsupported RAM configuration (some boards support only modules with parity) […]
Show full quote

- fatal (hidden) damage caused by leaked battery,
- unsupported RAM configuration (some boards support only modules with parity),
- damaged BIOS,
- or "forgotten" turbo switch.

1. even some small leaking can cause great damage - one small stain can eat away a big trace
2. this one is why I bought huge amounts of old ram 🤣
3. I don't even check it anymore, I just burn one and retest - "bit rot" is the best advice I got when starting my dwellings here
4. this affects speed, not POST, unless I'm wrong

Broken keyboard controllers are also a problem, not so uncommon. I also test with known working components if socketed, or desolder and socket.
No stones unturned 😀

Tomek TRV wrote on 2025-01-19, 17:45:

Post always stop on 140C codes (BIOS AMI). (...) replaced keyboard controller (...)

Did you check if it is getting correct clock? It should have external crystal resonator, probably that one near the battery. Those can die with age, and any excessive dirt (this also means battery residue or spilled capacitor guts) between their pins can prevent them from starting up. Preferably test with a scope but good quality modern multimeters can also be used - a DC voltage of about 2.5V on the pins would usually mean the oscillator is running.

So I did some measurements.
Reset as I wrote before - ok.
CLK, IRDY, FRAME - on but not so bright like voltage indicators.
All components have +5,2V power supply.
Little crystal resonator 32,8kHz and the same is on OSC1, OSC2 on RTC chip.
CPU is 16MHz AMD 286 and 16MHz on pin CLK; 5,2V VCC. Chip is hot. Reset is ok.
Keyboard controller XTAL1, XTAL2 - 8MHz

Please, take pictures without the keyb controller and RTC.
Remove the battery, cut the legs or desolder, and check all the traces underneath.

Check all the traces with a magnifying glass thoroughly, also on the back.
Looks like one board I had that would stop because of a missing trace (I soldered a wire to the wrong component pin).
It was a chipset pin signal.

The cpu is hot but not hurting, that's normal. 286 get hot but not scalding, should be around 50°C.

Do you see irdy and frame blinking at any given time?

Is BIOS size jumper correctly set? J4 (just try, noting will explode, maybe)
J5, I/O speed x2 x3, test it too.

All legs of the chipset are firmly soldered?

Tomek TRV wrote on 2025-01-21, 21:12:

All components have +5,2V power supply. Little crystal resonator 32,8kHz and the same is on OSC1, OSC2 on RTC chip. CPU is 16MHz […]
Show full quote

All components have +5,2V power supply.
Little crystal resonator 32,8kHz and the same is on OSC1, OSC2 on RTC chip.
CPU is 16MHz AMD 286 and 16MHz on pin CLK; 5,2V VCC. Chip is hot. Reset is ok.
Keyboard controller XTAL1, XTAL2 - 8MHz

That is good, except the CPU should have 32MHz on the CLK2 pin. So now it's running at 8MHz but possibly that's normal, BIOS might be booting in de-turbo mode.
Remove RAM from bank 3, leave only bank 2 populated. Make sure you are using 1MB SIMMs as these mobos usually don't support anything else. If you are not 100% sure the sticks are good then try swapping them around, it's unlikely all 4 are completly dead.

Also inspect the bottom of the mobo, especially under CPU and chipset and under RAM, where there are a lot of thin traces. There might be deep scratches that affect the signals, or even completly open connections. Do you have a scope? Probe the BIOS ROM /CE and /OE pins. A good multimeter can also help, a switching signal will show as somewhere between 0V and 5V depending on duty cycle. Not as good as scope but better than nothing.

Yesterday I removed battery and checked that it is no difference if I change J4 (27C256 or 27C512). I also checked all traces under magnifying glass and they are looking ok. Error code 14 for AMI bios means that there is a problem with first 64kb of RAM. I will desolder all slots and replace them with new ones but before I will do it can someone tell me what is between chipset and RAM? If there are some for example 74xxx chips then for sure easier is to replace them than desolder these 30 pin slots. Someone made a HT12 board replica but I can't find schematics. Whole chipset is just one IC so it makes this troubleshooting easier.

Oh it does spit out error codes? I missed that. Skip the BIOS probing then and focus on RAM. You don't need to desolder the slots (unless you just like to solder stuff), simply test with a meter all the pins on the SIMM slot. All should be connected to chipset or power rails. Usually on these single-chip mobos the RAM is connected directly to chipset. If not there will be some ALS or F buffers between them but that is rare and often only used as amplifiers for the address lines (one direction only). There might be (in fact should be but some mobos skip those too) resistor packs to terminate the address and data lines near the slots. Usually in hundreds of ohms. A damaged resistor pack could be causing problems - but I don't see any here. Those between ISA slots are probably for the ISA bus, at least there's that.

BTW did you try to test the mobo without RAM at all? Does it also stop on the same POST codes?

I didn't try this mobo without any memory. I have to do it but now I am going to work for two weeks so further troubleshooting must to wait. I tried to start this mobo with simm modules with 9 and with 3 chips. I have a few sets of them but nothing is changing. The best would be if I could see some schematics for HT12 mobo.