VOGONS


First post, by Cydernet

User metadata
Rank Newbie
Rank
Newbie

Hi all,

Would it be possible to help me with a Pentium 233 motherboard problem that I’m experiencing? The processor I am trying to install is a Pentium 233 MMX and the motherboard is a MSI MS5129 V1. The problem I’m experiencing is that I cannot get the processor to run at anything but 133mhz or 166mhz. I have correctly set the jumpers for the voltage and the SW1 switches to the relevant on/off positions according to the manual, which states that it supports the P55 cpu but cannot achieve any other speeds, especially the 233mhz my CPU is designed for. I have updated the boards bios to the final revision as well. I’m fairly stumped at this point and was wondering if anyone had any ideas?

Cheers

Reply 1 of 9, by Garrett W

User metadata
Rank Oldbie
Rank
Oldbie

Hi!

I'm not sure which board revision you have but I found the manual for revision v1.3 on archive.org. Page 20 has all you'll need for setting up the FSB and multiplier correctly. Make sure you set the DIP Switch for 66MHz FSB operation. Now 233/66 = 3.5, so that's your multiplier. Unfortunately, it looks like you can only do 1.5x,2x,2.5x and 3.0x according to the manual. Sometimes there are undocumented settings for such things, however if we are to take what the manual says at face value, there are only two switches that control the multiplier, and seeing as those switches only have two states, there cannot be more than 4 multiplier options. But perhaps you can try moving the first switch, as it seems to stay in the same position for all listed combinations. How do you have your switches set up right now?
Also, make sure you are using split voltage at 2.8V for the CPU and 3.3V for I/O, as this is how PMMX CPUs are intended to be used.

Reply 2 of 9, by SScorpio

User metadata
Rank Member
Rank
Member
Garrett W wrote on 2022-09-23, 15:40:

Unfortunately, it looks like you can only do 1.5x,2x,2.5x and 3.0x according to the manual.

The P55C 233 MMX is wired so 1.5x = 3.5x internally.

From the manual, but it could vary based on board revision. But there are six switches.

The first four switches control the bus speed, you'd want this set to 66Mhz.

The final two switches control the multiplier which can be in one of four states.
off off - 1.5x(3.5x) - 233Mhz
on off - 2x - 133Mhz
on on - 2.5x - 166Mhz
off on - 3x - 200Mhz

I believe most P55C 233 chips are multiplier unlocked. So just adjusting those last two switches should let you post at any of the four speed settings. Are you able to get all four different speeds? Have you tried running a benchmark just in case a BIOS reports the speed as different from what it really is?

Reply 3 of 9, by Repo Man11

User metadata
Rank l33t
Rank
l33t

The manual linked above shows support for a P54 @ 200 MHz, not a P55. I believe you have the same issue as Necroware had with the board in this video. The link should be at the correct portion of the video, but it's at the 5:00 mark if not. https://youtu.be/J0NLGfocviU?t=292

"We do these things not because they are easy, but because we thought they would be easy."

Reply 4 of 9, by Cydernet

User metadata
Rank Newbie
Rank
Newbie

Thank you for your replies. I’ve tested the different switch settings and checked using speedsys regarding the benchmarks which match what is reported and multiplayers which are as follows, using the last two (5 and 6 SW1 switches.
The other four switches are set to 66mhz clock.

DD: 133 mhz x2 multi
DU: 166 mhz x2.5 multi
UU: 166 mhz x2.5 multi
UD: 133 MHz x2 multi

This is puzzling as it doesn’t match what the manual is stating they should be.

Reply 5 of 9, by SScorpio

User metadata
Rank Member
Rank
Member
Cydernet wrote on 2022-09-23, 17:51:
Thank you for your replies. I’ve tested the different switch settings and checked using speedsys regarding the benchmarks which […]
Show full quote

Thank you for your replies. I’ve tested the different switch settings and checked using speedsys regarding the benchmarks which match what is reported and multiplayers which are as follows, using the last two (5 and 6 SW1 switches.
The other four switches are set to 66mhz clock.

DD: 133 mhz x2 multi
DU: 166 mhz x2.5 multi
UU: 166 mhz x2.5 multi
UD: 133 MHz x2 multi

This is puzzling as it doesn’t match what the manual is stating they should be.

Take the board out of the case, and use a multi meter and perform a continuity/diode test on the back of the board where the bank of switches are. I bet you one of the two is stuck in an open position, that's why it can switch between 133 and 166. But is unable to get the other two speeds.

The rear of the switches should be easy to find as it will be two rows of six contact points. Just bridge the same pin in the two rows to do the test. Change the switch position and retest.

* * * * * *
- - - - | -
* * * * * *

Reply 6 of 9, by Garrett W

User metadata
Rank Oldbie
Rank
Oldbie

I had forgotten about the 1.5x multiplier getting interpreted as 3.5x, good catch!

Reply 7 of 9, by Cydernet

User metadata
Rank Newbie
Rank
Newbie

Thank you everyone for your help. May try bridging the connections in time and hopefully this will correct the issue.

Reply 8 of 9, by SScorpio

User metadata
Rank Member
Rank
Member
Cydernet wrote on 2022-09-24, 08:42:

Thank you everyone for your help. May try bridging the connections in time and hopefully this will correct the issue.

Bridging won't help if that's really the problem. If the switch is stuck then the connection is already bridged and you need to break that link.

The repair will require replacing the bank of switches the question is how far apart the holes underneath are. If they are the standard distance you might be able to change from switches to 2x6 pins and just go with jumpers rather than the switches.

Actually... thinking about it a little more. You could just remove the switches altogether. All six in the off position will give you 66Mhz bus and the 1.5(3.5)x multiplier so you'll have the 233Mhz. You just won't be able to underclock.

Reply 9 of 9, by Repo Man11

User metadata
Rank l33t
Rank
l33t

I have a PCChips M520 which states that it supports P55 CPUs ( https://theretroweb.com/motherboards/s/pcchips-m520 ), and it has the exact same issue (and the same issue as the board in the video). The 3x multiplier doesn't work with P55 CPUS because of a difference in the CPUs; presumably they would work correctly with the optional VRM module. As is, the only CPU that will run faster than 166 MHz in this board would be a 200 MHz P54.

"We do these things not because they are easy, but because we thought they would be easy."