I didn't see your question answered but the 3 gold pads is the power, ground, rpm contacts for the fan just like the POD for socket 3.

Sphere478 wrote on 2022-06-04, 07:44:

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

These gold pads are for the fan

EDIT: Looks like chrismeyer6 beat me to it xD

chrismeyer6 wrote on 2022-06-06, 11:53:

I didn't see your question answered but the 3 gold pads is the power, ground, rpm contacts for the fan just like the POD for socket 3.

I'm talking about the other three. I know the big ones are for the fan, that's obvious 🤣...

-

Result of test with kingston interposer:

Windows hangs on dontgo.sys

-Removed overdrive and replaced with 233 mmx. So two mmx, one inthe interposer ins second socket, one not in interposer in primary socket. (no overdrive)
same result.

-Removed interposer, installed two mmx as normal (no overdrive, no interposer) works now!

-Replaced the newly installed mmx with the mmx from the interposer still works.

So it seems my interposer is the problem. I may have it set wrong? Will have to confirm the voltage setings.

Edit: there definitely seems to be a issue with the kingston interposer in the second slot. I'm not going to be able to run the test with it. I'm guessing it's not transmitting a mps pin or something because I know it works with my k6-3+. I verified the voltage and tried a few multipliers, no dice.

I think this test could work though with a different interposer. perhaps the one Feipoa was using? Feipoa, ya wanna run the test for me?

Kingston TurboChip socket 7 voltage regulator related questions
my interposer btw

Sphere478 wrote on 2022-06-04, 07:44:

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

On the socket 3 overdrive they controlled the multiplier in the following way: if fan is working (not shorted?) then 2.5x, else 1x.
Here is a video on that

https://www.youtube.com/watch?v=JVpmTlndEco&a … hannel=Atheatos

perhenden wrote on 2022-06-06, 13:09:

On the socket 3 overdrive they controlled the multiplier in the following way: if fan is working (not shorted?) then 2.5x, else […]
Show full quote

Sphere478 wrote on 2022-06-04, 07:44:

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

On the socket 3 overdrive they controlled the multiplier in the following way: if fan is working (not shorted?) then 2.5x, else 1x.
Here is a video on that

https://www.youtube.com/watch?v=JVpmTlndEco&a … hannel=Atheatos

oh that's wonderful!

okay wow, so this is all assumption, but I'm gonna guess that that pad is a BF pin... because that would be the easiest thing for intel to do. but that's assuming a whoooole lot.

Question is, what does a 200mmx do when you pull the fan?

It seems like maybe the 200mmx overdrive was more of a package swap, where as the socket 3 pentium overdrive was actually a package and silicon level job.

-

Edit: nothing happens (according to post screen) when you pull the fan on a 200 mmx OD. The dang thing doesn't seem to make much heat 🤣 they may be running these at less than 2.8v

-

Edit 2: cpuz also says 200 mhz but man is this overdrive freaking slow...

maybe I'm just spoiled with k6-3+ and dual 233 mmx.

But probably not:
according to cpu-z benchmark I'm like 1/4-1/5th the performance of a 200 mmx right now. So something isn't right.

-

Edit 3: I plugged the fan back in and performance instantly went up to where it should be. that confirms that even though the bios and cpuz say that it is at 200 it is in fact doing something odd.. perhaps disabling cache? or maybe the core clock is being detected wrong. hard to say...

I wonder if this is what was happening to me earlier. Maybe a dirty contact?

With respect to windows XP being unstable and/or 100% CPU with dual procs - make sure it is NOT trying to use ACPI. You may also need to change the MPS version from 1.4 to 1.1 in your BIOS or vice versa I can't remember which one was better.

maxtherabbit wrote on 2022-06-06, 14:35:

With respect to windows XP being unstable and/or 100% CPU with dual procs - make sure it is NOT trying to use ACPI. You may also need to change the MPS version from 1.4 to 1.1 in your BIOS or vice versa I can't remember which one was better.

The behavior with the fan is too similar to what it was doing earlier. It may have just been a dirty contact.

In any case the xp install that I used was native to a dual 233 mmx so xp should not have cared that I swapped one of them with a mmx overdrive. i can see how it might have cared about the p54c though.

The mps version is a interesting idea. You think that may be why the p54c failed to fully boot?

I'm kinda bummed that the interposer didn't work. I really think that setup should work with a different interposer. (As a permanent solution.)

In digging through my notes from the NEC Proserva system, I determined that the POD200 MMX needed to be placed in the Primary CPU slot and the Kingston P200MMX is in the Secondary slot. I wrote down some Quake 3 scores, so looks like the timedemo finishes. I then sourced and obtained a second POD200, but it would only work with one CPU. Keep in mind that I am using two different versions of the POD200, the v2.0 and v2.1. S-Specs were SL2FF and SL2RM. Perhaps someone would be willing to attempt the de-cap and wire mod to see if they can get two POD200's running in SMP.

feipoa wrote on 2022-06-08, 19:55:

In digging through my notes from the NEC Proserva system, I determined that the POD200 MMX needed to be placed in the Primary CPU slot and the Kingston P200MMX is in the Secondary slot. I wrote down some Quake 3 scores, so looks like the timedemo finishes. I then sourced and obtained a second POD200, but it would only work with one CPU. Keep in mind that I am using two different versions of the POD200, the v2.0 and v2.1. S-Specs were SL2FF and SL2RM. Perhaps someone would be willing to attempt the de-cap and wire mod to see if they can get two POD200's running in SMP.

Thanks for confirming the findings 😀

I think if someone decided they were willing to attempt a wire mod they should also de cap a ceramic pentium mmx as well, as I believe it could be used as a guide to figure out which wires were missing/changed.

To be clear if you (people reading this) aren’t aware, these wires are impossibly small, it would require extremely precise equipment and a microscope to do this, if even it is possible.

Unsure what core these pics are of, but appear to be p55 or p54 in a standard non overdrive package.

We should get fritz involved here 😀

If I haven’t mentioned it already, my thoughts are this:

I think intel made this package able to do smp but chose to disable it through gold wire, so I think the traces exist, though I have no evidence of that, it is a hunch.

There in theory should be three pads of interest to us,

One for the smp

And two bf pins.

We will likely see these pads have some or at least one wire in them.

I bet they ran a wire to vss from the smp core pad instead of to the smp pin pad on the package

I suspect the same for whatever configuration sets 3x multiplier.

Setting 3.5x may be as simple as cutting a wire. Or adding one.

The smp core pad, we could re route it to the pin pad on the package or simply move the wire to vcc3 to make it permanently a secondary socket cpu.

Sphere478 wrote on 2022-06-08, 20:30:

We should get fritz involved here 😀

Have read your PMs and sorry I don't even have a Pentium MMX Overdrive yet to take pictures of the CPU (have only standard ceramic Pentium MMX).

Fritzchens Fritz wrote on 2022-06-19, 17:55:

Sphere478 wrote on 2022-06-08, 20:30:

We should get fritz involved here 😀

Have read your PMs and sorry I don't even have a Pentium MMX Overdrive yet to take pictures of the CPU (have only standard ceramic Pentium MMX).

We could use some high res ceramic mmx shots. If you have better ones than I posted above. That would be helpful. Thanks Fritz!

Oh btw, I see four levels for those gold wires, looks like two for traces, and two for power?

Can you confirm those lower two are power planes?

If you wanna go really crazy, you could map out the pin/pad associations. Specifically the three in question.

I found out why my Kingston interposer isn’t working. It’s missing the ADSC# pin. 🙁 looks to be broken off.

Sphere478 wrote on 2022-06-23, 11:10:

I found out why my Kingston interposer isn’t working. It’s missing the ADSC# pin. 🙁 looks to be broken off.

You should be able to fix the broken pin with a fine piece of wire.

chrismeyer6 wrote on 2022-06-23, 11:50:

Sphere478 wrote on 2022-06-23, 11:10:

I found out why my Kingston interposer isn’t working. It’s missing the ADSC# pin. 🙁 looks to be broken off.

You should be able to fix the broken pin with a fine piece of wire.

I’ve been contemplating how to fix it actually, it’s tough because it’s inside the interstitial socket which is quite permanently attached.

I think someone around here had one they pulled apart. Wonder if they would be willing to send me some pins?

Btw, I learned something today. The overdrive is actually a dual plane processor. But both planes are designed for 3.3v.

Found this