It's interesting that it has an empty oscillator socket. Does it have something to do with the 80287 socket?

Yes, this empty socket is for a 80287.

It should work. The cpu runs at 1/2 of 50mhz and the coprocessor clock is 1/3 of the cpu clock.

Yes, for the i287 that's true (2/3 CPU or 1/3 SYS). Except for 287XL and clone chips.

The guide about copros says:
"The Intel 80287XL, the Cyrix 82S87, and the IIT 2C87 contain the internals of
a 387 coprocessor, but are pin-compatible to the original 287. These chips
divide the system clock by two internally, as opposed to three in the
original 80287. Since the 80286 also divides the system clock by two, they
usually run synchronously with respect to the CPU, although they can also be
run asynchronously."

This topic is also discussed here.

I hope you don't mind I'm qouting this.. I'm not saying this to sound smart or something,
I'm just quoting this because I'm a bit worried about math coprocessors. They get damaged easily.
Especially the original ceramic 80287 made by Intel gets really hot and requires a lot of power (I own one myself)!
Never try to overclock this one! It will either calculate wrong numbers or gets damaged.
If you have to use one, I strongly recommend to add a larger heatsink and make sure your PSU is still strong..

It's interesting... the quartz is 32MHz so this board runs at 16MHz. Someone has already changed the CPU but forgot or didn't know that the quartz has to be changed too.

Anyhow... about 80% of 20MHz Harris cpus should overclock to 25MHz without problems. But this 25% clock increase overclocks everything on the motherboard so it's not 100% that it will take it. Especially with fast I/O enabled on this board.

I have a similar project planned soon. In my case the 20MHz Harris is soldered on the board so I'll take a chance and hopelly it will work at 25MHz with all the bells and whistles of HT12/A chipset 😀

Yup... that 32MHz OSC needs to come out before the benefits of anything faster than a 16MHz CPU can be realised.

That other empty OSC socket, would that be separate for the 287? Interesting if it runs off a separate clock!

The old HT12 chipsets do support 286s upto 16MHz and about 4MB of RAM (maybe even more, like 8MB).
Specs

Jo22 wrote:

The old HT12 chipsets do support 286s upto 16MHz and about 4MB of RAM (maybe even more, like 8MB).
Specs

Ok, good to know. Thank you for this link.

I have already a Nos Harris286-25 with new oscilator.

I think i will look for an other 286 motherboard with 286-25.

Aw, sorry. 🙁 I didn't mean to discourage you with that project. I just googled for a reason this board had a 32MHz oscillator installed..
Maybe it will work on a higher frequency, nevertheless.
Datasheets for older hardware are often more refraining because of safety reasons (back then computers were a serious thing!)

I think there was also a discussion at dosforum.de about headland stuff.
Maybe some of those guys know more about this.

Besides, many years ago I read an article about an high-end 286 computer in a magazine (late 80s issue).
That PC also used a Harris 286-25 and a cooling system made out of cardboard! 😁
The system was overclocked, of course. I also remember, the experts were a bit shocked about this.
They couldn't belive this machine was intended for series production (cardboard ? no way!), they thought it was only a prototype made for them exclusively.

I'm not an overclocker, but I think it must be at least possible to overclock 286 chipsets, because it has already been done in the 80s.
Except for some VLSI cipsets, maybe. I heard they are more sensitive than usual. Same for cache/parity chips.