Well.. I've been looking at those from different sellers so just when ahead and purchased 2 since shipping for additional items is free.

Mine look more like this one:

I suppose the number after the N is the date of manufacture. Usually 91 to 95.

kixs wrote:

I suppose the number after the N is the date of manufacture. Usually 91 to 95.

Oh dear, then that "N1901" chip would be from the 1st week of 1919 😁

2019 and these are still produced. They are used in the Boing 737 Max for example.

derSammler wrote:

They are used in the Boing 737 Max for example.

This is... bad news...? 😕

Indeed. Their lack of speed may have caused the two crashes. At least that's the last info I did read on that topic.

Anyway, it also means that the pictured CPU is most likely legit.

it's probably the one part of the plane that's actually reliable 🤣

derSammler wrote:

2019 and these are still produced. They are used in the Boing 737 Max for example.

Where? Digikey says Renesas has EOL the part in 2017.

Renesas? We are talking about Harris.

derSammler wrote:

Renesas? We are talking about Harris.

Renesas owns Intersil, which was spun off from Harris' semiconductor division in 1999

So if no one is manufacturing Harris branded 286s in 2019, that would out this pictured CPU as a remarking or outright fake, right?

Most likely fake like many other ICs out there.

Well are they ? Someone bought them, maybe they are working and someone else make these

derSammler wrote:

Indeed. Their lack of speed may have caused the two crashes. At least that's the last info I did read on that topic.

Anyway, it also means that the pictured CPU is most likely legit.

This can't be! If they are good for F-29 Retaliator advanced dynamics simulation, then they are good for slow ass 737 as well! ;D

I bought a couple of these from UT Source some time ago. Selled one to friend of mine, and he says it's warking fine, but it's not really stable with 50MHz OSC (although it may be cause by something else - ie. chipset - as far as I remember he uses motherboard with HT12A and I'm not sure if this motherboard has dedicated OSC for AT Bus or if it's clock is derived from CPU's OSC).

The UTsource chip I bought and it definitely looks like the real deal. Passed acetone test and performs properly too.

HT12A definitely won't do 25MHz but it gets to 20MHz no problem but few extra MHz and it starts to crap out. on the board i have there's two oscillators, both go to chipset but one of them controls ISA speed and that one cannot be too high (in my case 37.5MHz was max for that oscillator, over 32MHz stock). Secondary oscillator was 25MHz and is now replaced with 40MHz and things run fine.

Tiido wrote:

HT12A definitely won't do 25MHz but it gets to 20MHz no problem but few extra MHz and it starts to crap out. on the board i have there's two oscillators, both go to chipset but one of them controls ISA speed and that one cannot be too high (in my case 37.5MHz was max for that oscillator, over 32MHz stock). Secondary oscillator was 25MHz and is now replaced with 40MHz and things run fine.

Interesting. I have a board with a socketed Harris 16Mhz but a soldered on 32Mhz oscillator. It has the HT12 chipset, not the HT12A.

I was wondering if I would be able to upgrade the CPU and from what you're saying a 20Mhz would be possible. I'd have to swap out the oscillator which is a bit of a pain, my soldering skills aren't that strong. However, once the soldering iron is out, I imagine an oscillator can be replaced with a socketed one for easy future swapping?

I find it odd that the CPU on my board is socketed and the oscillator is not. Can the 32Mhz oscillator support lower than 16Mhz frequencies, but not higher, by any chance? E.g. a 12Mhz CPU in combination with the 32Mhz oscillator?

My board:

pan069 wrote:

I was wondering if I would be able to upgrade the CPU and from what you're saying a 20Mhz would be possible. I'd have to swap out the oscillator which is a bit of a pain, my soldering skills aren't that strong. However, once the soldering iron is out, I imagine an oscillator can be replaced with a socketed one for easy future swapping?

Yeah, you need DIP14 socket just remove all but the 1,7,8,14 (corner) pins.
It's just 4 pins to desolder, very easy to do (you will need solder sucker / desoldering pump).

pan069 wrote:

I find it odd that the CPU on my board is socketed and the oscillator is not. Can the 32Mhz oscillator support lower than 16Mhz frequencies, but not higher, by any chance? E.g. a 12Mhz CPU in combination with the 32Mhz oscillator?

It's because this board was originally designed to be used with PGA CPUs as well (take a look at the underside of the board - you will see that PLCC socket is soldered in as PGA) - they were using what was at hand, and because they had PLCC CPUs - they just used PGA-PLCC adapter. It was pretty common practice. 286 motherboards with PGA sockets or soldered-in 286 PGA CPUs are semi-rare.

12MHz CPU with 32MHz OSC will be overclocked to 16MHz. With 12MHz CPU you need 24MHz OSC.

Also keep in mind, that your FPU may not keep up with the higher clocked CPU and there's not much you can do about it, because there is no space for additional FPU OSC (maybe with JP9 you will be able to change the divider for the FPU though). So with 20MHz CPU you would need 12MHz or faster FPU.

Nice board BTW.

Awesome. thanks for the info!

Looks remarked to me. Indent texture differs alot from that of the surface, it was either blacktopped or etched away and re-engraved. Original part was probably a 20MHz Harris.