Curious, the picture of the same board on UH19 doesn't show an oscillator there:
http://www.win3x.org/uh19/public/motherboard/ … 93508883311.jpg

The other crystals seem to all be there. Any idea what the crystal on your board connects to?

I have a Asus ISA-386C and it does not have a crystal there. Found some pics thru google images and a few show a crystal but most do not.
Re: What retro activity did you get up to today?
https://www.clous.cz/isa-386c/
Yes you need to track down where it goes !

Looks like there are some versions. Mine looks like the one from ASUS poster...

This oscilator 'clocks' XTAL2 of the P8042AHP.
The guy I got it from said that it was 80MHz oscilator... but I've got some doubts...

Nice FPU you've got there 😀

Locutus wrote on 2021-08-28, 00:10:

Looks like there are some versions. Mine looks like the one form ASUS poster... ASUS_PIC.jpg […]
Show full quote

Looks like there are some versions. Mine looks like the one form ASUS poster...
ASUS_PIC.jpg

This oscilator 'clocks' XTAL2 of the P8042AHP.
The guy I got it from said that it was 80MHz oscilator... but I've got some doubts...
image_2021-08-28_021737.png

Yep Mine is v1.1, yours is v1.0 (printed between the ISA slots.)
So it runs up to the KB controller after going thru logic up by the 85c206 ? Saw your drawing and it of makes no sense (not your fault) but to run a clock circuit that far to logic then back to the KBC just seems odd.
Am probably mis-reading what you meant but that just seems out of spec and not a normal way to do things.

Can you take a good picture of the back of your board ?

My board for reference.

Well the datasheet for the P8042 gives the min/max clock period as 80 & 613 ns. Which I think works out as 12.5MHz down to 1.63MHz. So that narrows it down a bit. Looks like they might be using the second clock input circuit from the datasheet:

Thans for all the answers !

Horun wrote on 2021-08-28, 01:41:

Yep Mine is v1.1, yours is v1.0 (printed between the ISA slots.)
So it runs up to the KB controller after going thru logic up by the 85c206 ? Saw your drawing and it of makes no sense (not your fault) but to run a clock circuit that far to logic then back to the KBC just seems odd.Am probably mis-reading what you meant but that just seems out of spec and not a normal way to do things.
Can you take a good picture of the back of your board ?

Sorry, english is my second language so....
That seems odd to me too, but I've tracked connections from the PCB .

snufkin wrote on 2021-08-28, 09:16:

Well the datasheet for the P8042 gives the min/max clock period as 80 & 613 ns. Which I think works out as 12.5MHz down to 1.63MHz. So that narrows it down a bit. Looks like they might be using the second clock input circuit from the datasheet:P8042_Clock.jpg

Exactly my point - that's why I've asked.... but I'm a 'hothead' so the storry continues fast.... 😉

This is the bag of oscillators that I got with the mainboard. The guy I got it from is a ham radio enthusiast... he took out oscillators 25years ago and (Thank God) forgot to trash the rest...
He is 100% positive that this is the set from this motherboard (they turned out not so useful after all...).
Batch dates on them seem to confirm his words.

I've done some soldering...

and...

1) The mainboard 'gave signs of life' 😉
It showed BIOS version and stopped on memory test (sometimes counted 64k..... of RAM after power on but in most cases nothing at all) and made a reset (in infinite loop)
2) To be sure I've mounted 100% verified RAM and CPU and removed FPU (same behavior).
3) I took out keyboard controller (8042AHP) and without it, the mainboard starts, tests all installed RAM and freezes with constant beep where it should display 'Press ....'(quite normal right ?).

So... I've returned to this 'enigmatic' oscillator.
Signal on the output (capacitance of the probe makes it look like sine).

This is the signal on the input of the buffer that is connected to the XTAL2 of 8042AHP chip (pin 9 of the 7407 IC)

And on the output of the buffer (this is connected directly to XTAL2 of 8042AHP) ...

[EDIT]
It was late.... and I have to repeat last measurement (trigger is not set correctly...).

This is quite interesting !
If you manage to fix it, would you like to take another picture so I can save it to win3x ?

Deksor wrote on 2021-08-28, 13:05:

This is quite interesting !
If you manage to fix it, would you like to take another picture so I can save it to win3x ?

Sure !

... the story continues...
I've checked P8042AHP in another mainboard - works fine (CLK ~7.1MHz).

I've confirmed my previous 'clock' schematic and I'm 100% sure that it looks like that:

...which brings the conclusion that this oscillator musn't be 80MHz...

Yep ! The 7404 is an inverter and 7407 is a buffer, no flip-flops/dividers if you traced accurate.
Sounds like you need a ~8Mhz crystal. Not 80Mhz.

The story ends with a happy ending 😀

Tested with external generator, P8042AHP works fine up to ~12MHz.
I'll mount 8MHz crystal to reduce it's heating.

Great ! added: I think you want a DOC 20 or DOC 70 type crystal (case size 14pin, TTL output)

Deksor wrote on 2021-08-28, 13:05:

This is quite interesting !
If you manage to fix it, would you like to take another picture so I can save it to win3x ?

There you go, complete and fully working 😀

really nice detective work... but very,, very satisfaying when all turns out ok in the end.