I would guess that none of the jumpers in that chart let you change the CPU multiplier. JP7 only sets the bus speed. As the system was only designed to accept 75,90 and 100MHz CPUs, perhaps there was no need to have such a jumper. However, it could be undocumented, so look for information from a 3rd party source to see if it exists.

If not, there are ways to force it by soldering the bottom of the CPU socket.

*edit*
I just found the TH99 manual
http://www.uncreativelabs.de/th99/m/C-D/33436.htm
No multiplier jumper is listed, so it probably doesn't exist. Still pretty interesting to have a P54C board with a UMC chipset though!
So you either have to solder, use an overdrive or a 3rd party CPU adapter to get 133 going.

Soldering isn't even necessary. You can just wrap a single wire strand around Vss (ground) and a BF0 pin to set CPU multiplier to 2x, inserting a thin wire between Vss and BF0 in the socket would also do the trick. I used this trick on a IBM 330-p75 to set CPU multiplier to 3x. If your 133MHz P1 was manufactured in 1997+, it has a good chance of running @ 166 or even 200MHz freq.

|BF0|BF1|mult.|
| 1 | 1 |1.5x |
| 0 | 1 |2.0x |
| 0 | 0 |2.5x |
| 1 | 0 |3.0x |

*1 - high, 0 - low.

SSTV2 wrote:

Soldering isn't even necessary. You can just wrap a single wire strand around Vss (ground) and a BF0 pin to set CPU multiplier t […]
Show full quote

Soldering isn't even necessary. You can just wrap a single wire strand around Vss (ground) and a BF0 pin to set CPU multiplier to 2x, inserting a thin wire between Vss and BF0 in the socket would also do the trick. I used this trick on a IBM 330-p75 to set CPU multiplier to 3x. If your 133MHz P1 was manufactured in 1997+, it has a good chance of running @ 166 or even 200MHz freq.

|BF0|BF1|mult.|
| 1 | 1 |1.5x |
| 0 | 1 |2.0x |
| 0 | 0 |2.5x |
| 1 | 0 |3.0x |

*1 - high, 0 - low.

I will admit im not very familiar with that! Would live to try it, have any photos/picture references i could go off of?

Anonymous Coward wrote:

I just found the TH99 manual http://www.uncreativelabs.de/th99/m/C-D/33436.htm

You are SUCH a savior! Someone actually documented these old oem boards 😲

Exactly what iv needed! I am keeping this for reference. Thank you!

I've marked pins of interest in red:

Here's the whole datasheet for P54C pentiums, you can find multiplier selection table in pg. 22.

Yuuker wrote:

Back with the Compaq 7170 again, trying to upgrade the p90 inside to a p133.

After a bit of playing around i found that jumping JP7 1-2 and 3-4 would get me to 120mhz.

Was it really running at 120Mhz? It could also be a BIOS reporting wrong speed.

As this would be only possible 2x 60Mhz. I'd play with the jumpers some more. Try all eight combos of J7.

kixs wrote:

Yuuker wrote:

Back with the Compaq 7170 again, trying to upgrade the p90 inside to a p133.

After a bit of playing around i found that jumping JP7 1-2 and 3-4 would get me to 120mhz.

Was it really running at 120Mhz? It could also be a BIOS reporting wrong speed.

As this would be only possible 2x 60Mhz. I'd play with the jumpers some more. Try all eight combos of J7.

The bios doesn't list the speed at boot, i needed to go into phil's benchmark pack and use tools to identify the speed.

8 combos? I only could figure out 5. All speeds where verified with CHKCPU and Speedsys,

JP7:
1-2: 75mhz
3-4: 90mhz
1-2 and 5-6: 100mhz
3-4 and 5-6: 60mhz
3-4 and 1-2: 120 mhz.
Jumping all 1-2, 3-4, and 5-6 caused the system to not boot.

Any jumper settings im missing?

SSTV2 wrote:

I've marked pins of interest in red:

pin_side_view.png

Here's the whole datasheet for P54C pentiums, you can find multiplier selection table in pg. 22.

Thank you. I'l hopefully be getting a chance to try it soon

How were you able to get 60 and 120MHz with only a 1.5x mult.? Either your motherboard's PLL supports 40 and 80MHz or it, in fact, has a undocumented 2x multiplier setting. Write down the model of your PLL chip, it must have a 14.318MHz quartz nearby.

SSTV2 wrote:

How were you able to get 60 and 120MHz with only a 1.5x mult.? Either your motherboard's PLL supports 40 and 80MHz or it, in fact, has a undocumented 2x multiplier setting. Write down the model of your PLL chip, it must have a 14.318MHz quartz nearby.

Hmm, i must not be good at finding or seeing things 😵 Didn't recognize any pll manufacture on the board.
One of these chips i presume. Numbers should be visible if you zoom in.

Jackpot! If PCI clock @ 80MHz FSB also gets overclocked, then system should perform sligtly faster @ 120MHz than @ 133MHz (2x66MHz). You got some nice potential here with OCing, technically you can run a P55C @ 280MHz (3.5x80MHz) in that early pentium board.

Clock generators on Socket 7 and older motherboards

80MHz probably won't be stable on socket5.

SSTV2 wrote:

Jackpot! If PCI clock @ 80MHz FSB also gets overclocked, then system should perform sligtly faster @ 120MHz than @ 133MHz (2x66MHz). You got some nice potential here with OCing, technically you can run a P55C @ 280MHz (3.5x80MHz) in that early pentium board.

Clock generators on Socket 7 and older motherboards

mx8315.pdf

Ohhh intresting! I'll need to dig deeper into these types of overclocking but my brief look tells me these are for the wire trick huh.

Anonymous Coward wrote:

80MHz probably won't be stable on socket5.

I plan on running a Voodoo 1 in this build. Aren't those suppose to be notoriously bad with pci overclocking?

Right now im looking at my upgrade paths for this machine, i have 256k of cache on the way. CPU wise i'd like to push it a bit further, maybe even to a MMX.
MMX Overdrive chips look expensive on ebay though, anyone know where i could find one of those voltage-converter boards, maybe rip one off one of those evergreen 200mmx upgrade kits? I think the ones used for 586s looked like they could be removed or something? Never used one before

WOW... the overdrive CPUs have gone way up in price... glad I got one that came with a board for cheap. Same for the 83Mhz one for 486 boards.

Anyway, Kingston did make an upgrade that had either a K6-200 or an Intel 200MMX.
http://www.cpushack.com/UpgradeProcessors.html

Edit: Here is an eBay listing lot that has an Evergreen Spectra with a K6-2 400Mhz CPU. (not my sale)
https://www.ebay.com/itm/8-Ceramic-CPU-PGA-AM … ry/143386301027

The one in that listing is a revision C... and apparently has less voltage settings than my revision B K6-2 333Mhz one.

Only problem with these is that if you don't have the original cooler, you have to rig something up because they raise the CPU up a lot so a standard cooler bracket won't work.

cyclone3d wrote:

WOW... the overdrive CPUs have gone way up in price... glad I got one that came with a board for cheap. Same for the 83Mhz one for 486 boards.

Anyway, Kingston did make an upgrade that had either a K6-200 or an Intel 200MMX.

Regarding the Intel Overdrives for a second, is it true that the 200mhz PODPMT66X200 options do not have the required voltage converters for socket5?

Getting mixed messages from articles online that suggested it does not, while others seem to show pictures of it with a regulator. Cpu World lists it only as socket 7 whereas the 160 OD is listed as both S7 and S5 😕

Yuuker wrote:

The bios doesn't list the speed at boot, i needed to go into phil's benchmark pack and use tools to identify the speed. […]
Show full quote

kixs wrote:

Yuuker wrote:

Back with the Compaq 7170 again, trying to upgrade the p90 inside to a p133.

After a bit of playing around i found that jumping JP7 1-2 and 3-4 would get me to 120mhz.

Was it really running at 120Mhz? It could also be a BIOS reporting wrong speed.

As this would be only possible 2x 60Mhz. I'd play with the jumpers some more. Try all eight combos of J7.

The bios doesn't list the speed at boot, i needed to go into phil's benchmark pack and use tools to identify the speed.

8 combos? I only could figure out 5. All speeds where verified with CHKCPU and Speedsys,

JP7:
1-2: 75mhz
3-4: 90mhz
1-2 and 5-6: 100mhz
3-4 and 5-6: 60mhz
3-4 and 1-2: 120 mhz.
Jumping all 1-2, 3-4, and 5-6 caused the system to not boot.

Any jumper settings im missing?

SSTV2 wrote:

I've marked pins of interest in red:

pin_side_view.png

Here's the whole datasheet for P54C pentiums, you can find multiplier selection table in pg. 22.

Thank you. I'l hopefully be getting a chance to try it soon

All off
All on
1-2
3-4
5-6
1-2,3-4
1-2,5-6
3-4,5-6

But you might not be needed as SSTV2 pointed out it's actually 80x1.5 and not 60x2.

I'd put a 233MHz P1 or an unlocked, lower clock P55C CPU into that system. P55C interprets 1.5x multiplier of a P54C as 3.5x, so no wire mods would be needed, also, you don't have to worry about slightly higher core voltage, +0.5V for a P55C is in a safe range.

Hmm... P233MMX in a very early socket5 board. Its sounds kind of sketchy. You're sure that running the core at 3.3V is "safe"? It might work if the BIOS doesn't reject it or the VRMs don't burn out, but I don't know if that's something I'd want to do on old hardware that I plan to keep around.

Yuuker wrote:

cyclone3d wrote:

WOW... the overdrive CPUs have gone way up in price... glad I got one that came with a board for cheap. Same for the 83Mhz one for 486 boards.

Anyway, Kingston did make an upgrade that had either a K6-200 or an Intel 200MMX.

Regarding the Intel Overdrives for a second, is it true that the 200mhz PODPMT66X200 options do not have the required voltage converters for socket5?

Getting mixed messages from articles online that suggested it does not, while others seem to show pictures of it with a regulator. Cpu World lists it only as socket 7 whereas the 160 OD is listed as both S7 and S5 😕

I have one. It does have voltage regulator to step the voltage down.

See here as well:
POD 200Mhz

If your motherboard supports 3.3v, you could just use a regular 233MMX chip in it as long as you have a good heatsink/fan and good airflow through the case.

Anonymous Coward wrote:

Hmm... P233MMX in a very early socket5 board. Its sounds kind of sketchy. You're sure that running the core at 3.3V is "safe"? It might work if the BIOS doesn't reject it or the VRMs don't burn out, but I don't know if that's something I'd want to do on old hardware that I plan to keep around.

It's safe for the CPU at least, both P54C (133MHz+) and P55C use 0.35um technology, the only downside of running a P55C at higher core voltage is slightly higher power dissipation of CPU and VRM. If figures here are correct, then current flowing through VRM on that motherboard would increase by more than ~65% @ 233MHz (because of +0.5V), compared with a 0.6um 100MHz P1 and that should be taken into consideration, before doing extreme OCing 😀

They both use the same fabrication process, but the P233 is probably pushing the envelope of what .35 micron could deliver, which might at least partially explain why the core voltage was lowered. BTW, the maximum vcore rating for this chip is 2.9V, at least according to CPU-World. I'm sure there are examples of overclocked chips functioning at all kinds of voltages and frequencies, but it's just something to consider.

Personally the bigger issue for me would be protecting the motherboard. At least check out the vregs and see what kind of current they can handle before dumping a P233MMx in there.