Found these images of a 200,

I don’t see anything looking like configuration resistors 🙁

Apparently there is a 180mhz part, I wonder if they are in fact 3.5x 50mhz units?

Or perhaps 3x60🤔 which doesn’t help.

Some info on overdrives:

Looks like it agrees that they are locked on the multiplier.

125mhz, 150mhz and 166mhz on p54c

And

150, 166, 180, and 200mhz? On p55c variant.

https://dependency-injection.com/pentium-overdrive-roundup/

So I’m guessing this is a p54c variant of the socket 5/7 overdrive. (Really only targeted for early socket 5 I assume, mainly for boards without multiplier support I assume) though any socket 5/7/ss7 should be able to run it?

Again, not seeing a obvious way to configure a multiplier. Which is again unfortunate because p54c came in up to 3x on some variants. (Bummer) I assume those are all capacitors.

Perhaps they are configured under the gold wire lid.?

Edit:

https://en.m.wikipedia.org/wiki/Pentium_OverDrive

More info:

Looks like p55c variants were set to 3x or 2.5x no 3.5x

Still no clues as to how to possibly set 3.5x

Looking over the datasheet for the pentium mmx overdrive it seems that smp may work, I will just likely have to go and try it to confirm.

Edit: more info:
https://cpushack.com/UpgradeProcessors.html

I’m wondering as to the function of the three small gold pads.

And what is under the daughterboard

Okay, testing time…

First I start with only the overdrive.just to make sure all is working.

The tyan has a dual/single plane jumper. I removed it for single plane setting and set the apic jumper to single cpu

Nothing, just “- -“ on the diag card.. okay, suspected the voltage settigns, so I set the dual plane voltage to 3.3v and reinstalled the jumper. (Maybe I was supposed to set the vcc2 to 3.3 beforehand? Idk, I havent ran many single plane cpus in this mobo over the years. anyway. With the vcc2 set to 3.3 and single/dual set to dual…?? It works! We have one overdrive detected! (And only one installed)

Okay, next step,

I only have one overdrive, so this is gonna be difficult…

The core its self is most similar to a pentium mmx obviously, but electrically as far as the voltage is conserned, it’s about the same as a p54c as far as the motherboard is concerned anyway.

So I figured it was safer to put a 3.3v sy045 pentium 200 (non mmx) in there than a pentium 200 mmx. So this is what I will try first.

Installed cpu, moved apic jumper, crossed fingers and…..

Bios reports two pentium overdrive cpus. 🧐🤔

So, it seems that while the data sheet says no, the processor says yes. 😀 this will have to be confirmed further though under actual OS conditions. Preferably with two overdrives, but we’ll see what the future holds.

Next I tried booting from a hard drive with a smp windows xp install on it.

It got to the mouse cursor and hung.

Suspecting the mmx vs non mmx to be the issue I rebooted and put a mmx 233 in the slot.

The board should still be set to 2x multi btw

It booted to the desktop and I noted that the mouse was still hung probably from usb driver not being installed (this is the case) I still think that it was probably hung with the p54c secondary though, I can try again later.

Anyway, the mobo so far as I know is still set to 2x multi from when I had the k6-3+ in it. So the overdrive should be at 200mhz, and the mmx 233 which is significantly overvolted right now and I advise against doing this, is running at I assume 2x for 133mhz

Okay, I got to windows, it’s not frozen, and all looks as I would expect.

Cpu-z just assumes that both are running at 200 therefore shows both doing so, even though I don’t believe this is the case and will try to prove that with a super pi test.

During screenshot, windows was still installing drivers so that kept the cpus going 100% 🤣

Not sure what happened but system remained very slow 100% cpu usage, did a reboot. This time I’m trying the p54c again. This time the mouse is working (because driver is installed now) in hindsight I think I just didn’t wait long enough, and it would have finished booting.

And…. Spoke too soon, It froze. Got to the blue welcome screen though.

Back to the pentium mmx we go.

Okay, may just have a buggy OS because two p54c chips are acting the same way. I’m gonna put the OS back in the dual 233 mmx it came from and see if it works there. And return this unit to the single processor os and see what we have going on. (i have three tyan s1564 mobos)

Conclusion:

A pentium mmx 200 overdrive seems to exhibit MPS compatibility.

The test was not ideal though, if you try this you should try it with two overdrives. I consider the test incomplete until someone does that.

It’s possible my slowness issues may have been the OS bugging out and a fresh install may have been fine? Unsure

The p54c didn’t seem to like the confuguration.

And I don’t recommend trying to overvolt your pentium mmx like that.

So yeah, if you wanna try it, try it with two overdrives.

After the test,

I tested the same mobo with a single processor install on the overdrive all seemed fine,

Had a issue with post card, ended up fixing it by reseating it.

Put k6-3+ back in system, all seems okay nice and fast again on the single processor install of xp.

Put that dual Processor xp install back in the system from once it came (233mmx x2), and it boots normally at normal speed now. All seems back to normal

The one test I should have done but didn’t is dual 233 mmx in the motherboard I tried the test in. I’m sure it’s fine though.

I’ll revisit this when I get a second overdrive.

A few years ago I tried dual P200 MMX overdrives on my dual socket 5 system, but the motherboard would only see and work with one CPU.

feipoa wrote on 2022-06-06, 09:20:

A few years ago I tried dual P200 MMX overdrives on my dual socket 5 system, but the motherboard would only see and work with one CPU.

Interesting. Which mobo?

I wonder if the socket 7 apic (if different) makes it work.

It’s also possible that it was a false positive somehow. Maybe somehow it was partially working. And would fail entirely if I put a second over drive in.

Setting both to 200mhz setting didn’t seem to solve the slowness issues. Forgot if I mentioned that.

I never got around to testing the two cores with super pi because of the slowness

This system, NEC Proserva V Plus: Re: NEC Proserva V Plus (dual socket 7 system) (have a lead on multiple of these)

It works in dual mode if I use regular P233 MMX CPUs from within Evergreen or Kingston interposers. MMX overdrives have SMP capability disabled.

feipoa wrote on 2022-06-06, 09:43:

This system, NEC Proserva V Plus: Re: NEC Proserva V Plus (dual socket 7 system) (have a lead on multiple of these)

It works in dual mode if I use regular P233 MMX CPUs from within Evergreen or Kingston interposers. MMX overdrives have SMP capability disabled.

I think I see a 430hx chipset there… 🤔 didn’t realize they put it on socket 5 systems.

I have a bad feeling that when Install two overdrives I’ll get your result. It may be that the mmx chip was forcing the overdrive to work or work partially with smp.

Which kinda makes me wonder, if a pin mod of some sort might be possible if that is the case.

What configuration would allow two processors one being a overdrive and one being a p55 or p54 to show as smp, but for it not to work if there were two overdrives?

Cpuid pin? D/p pin? 🤔

https://www.pchardwarelinks.com/586pin.htm

Picd0 I bet…

🤔

Okay, I'm coming up with some answers here it seems.

It seems that I may have actually stumbled upon the only way to actually dual a Pentium mmx overdrive.

This is a quote from the pentium mmx overdrive datasheet

1the CPUTYP signal pin
2on the Pentium OverDrive processor with MMX
3technology is internally not connected to the
4package pin. The core is internally tied to VSS.
5The pin on the package is defined as INC. Any
6external connections to the package pin will not
7affect the processor core because the core is
8physically disconnected from the package pin.

and this is a quote from the pentium mmx datasheet about that pin

1CPU type distinguishes the Primary processor from the Dual processor. In a
2single processor environment, or when the Pentium processor with MMX
3technology is acting as the Primary processor in a dual processing system,
4CPUTYP should be strapped to VSS. The Dual processor should have CPUTYP
5strapped to VCC3.

So it looks like the way they disabled smp was at the gold wire. the secondary must be pulled high to vcc3 but a overdrive is pulled permanently low and the pin is disconnected. which explains how I was able to do it. my pentium mmx overdrive was in the primary slot where it should be pulled low. and my pentium mmx was in the secondary where it should be pulled high and since it has that pin it is able to be pulled high.

mystery solved.

And... this also suggests how one might be able to enable it: With micro surgery on the gold wires 🤣.

more info

did some tests:

-The cputyp pin is indeed internal no connect on the mmx overdrive.

-The motherboard seems to be what drives this pin as a input to the processor.

-On the primary socket with cpus installed I have 25 ohms to ground and 100 ohms to vcc3

-On the secondary socket with cpus installed I have 100 ohms to ground and 25 ohms to vcc3

So this is all making 100% sense now.

You can use a mmx overdrive (in mps) in the primary socket, but not in the secondary socket, because then both would have cputyp driven low because of the internal pull down and disconnected pin.

-I can also confirm that all four BF pins are internal no connects as well. Bf0, bf1, bf2, and intel bf2 so you won't be setting the multipliers of this cpu externally. probably again another gold wire trick.

At this point it looks like if someone has the skills, they could do some gold wire surgery and probably enable mps and 3.5x but I won't be doing that to mine.

I'm gonna do a new test, this time with a kingston interposer in the second socket. I didn't try this before because I don't have a heatsink clip for it so I have to glue the heatsink on. Which is what I'm doing.