dondiego wrote:

Your motherboard doesn't support the pentium mmx aka p55c, you need either a different processor or a different motherboard. The multiplier is configured right now at 1.5 i guess for the pentium 100 and according to the manual the board supports multipliers up to 3.

ftp://ftp.fic.com.tw/motherboard/cpu_table/other/1.pdf

I think it's likely the manual states 3 as maximum multiplier because it doesn't officially support Intels MMX chips, which remap the 1.5 multiplier to 3.5.

He sais it works, but he might stress his board a bit by his current setup.

His 200MMX chip might have it's 3.5x multiplier disabled though, like many of the later MMX chips had.

Yes, this setup worked for five years until I decided to buy a new computer. But I never paid attention to it. Now I'm concerned about motherboard lifespan. The regulator gets ridiculously hot and I can't remember how it worked with the original Pentium 100.
Would you recommend me downgrading to some P54C or it's safe to stick with P55C?

The P100 does have a lower TDP than the MMX, so it might not run as hot. I'd say guetto-mod a fan nearby and be happy.

CPU temperature was never a problem. In fact it worked with a passive heatsink all those years. I'm concerned about the heatsinked transistor. It runs ridiculously hot. It burned my finger with a little touch on it. I'm afraid of motherboard lifespan because it's almost impossible to find socket 7 mobos in here, and I would be very sad if it brokens because it has much sentimental value to me.

I will try with a p54c and test it. If I see the transistor reachs the same high temps, I will consider keep the mmx on it and put some fan blower on top of transistor heatsink.

Little update on project: paint job is almost finished. I guess it will be ready for next monday. I will buy some more things and start assembling it maybe for tuesday 😎

And of course, a million thanks to all of you for beeing so helpful! It's been a long road since the last time I struggled with this kind of hardware so I don't remember so many things.

Can you please provide either the datasheet or part number for that regulator and I'll have a look at the thermal specs.

ramiro77 wrote:

Yes, this setup worked for five years until I decided to buy a new computer. But I never paid attention to it. Now I'm concerned about motherboard lifespan. The regulator gets ridiculously hot and I can't remember how it worked with the original Pentium 100.
Would you recommend me downgrading to some P54C or it's safe to stick with P55C?

I'd say just keep the MMX chip in your board and lower the multiplier to 2x so it runs at 133MHz for now.

gdjacobs wrote:

Can you please provide either the datasheet or part number for that regulator and I'll have a look at the thermal specs.

I already did that. It's a D45H8 pnp transistor. But since I don't have a temperature probe right now I can't do some measures. I will do it asap.
http://pdf1.alldatasheet.com/datasheet-pdf/vi … NSEMI/D45H.html

Tetrium wrote:

ramiro77 wrote:

Yes, this setup worked for five years until I decided to buy a new computer. But I never paid attention to it. Now I'm concerned about motherboard lifespan. The regulator gets ridiculously hot and I can't remember how it worked with the original Pentium 100.
Would you recommend me downgrading to some P54C or it's safe to stick with P55C?

I'd say just keep the MMX chip in your board and lower the multiplier to 2x so it runs at 133MHz for now.

Why 133 mhz? I understand that temperature rises most with voltage than frequency but maybe I'm wrong. And CPU temperature isn't a problem here because I'm using that fancy oversized socket 370 cooler. It's really really cold, it barely warms up. Running it at 133 mhz causes less stress to the motherboard? I'm intrigued on this.

ramiro77 wrote:

I already did that. It's a D45H8 pnp transistor. But since I don't have a temperature probe right now I can't do some measures. […]
Show full quote

gdjacobs wrote:

Can you please provide either the datasheet or part number for that regulator and I'll have a look at the thermal specs.

I already did that. It's a D45H8 pnp transistor. But since I don't have a temperature probe right now I can't do some measures. I will do it asap.
http://pdf1.alldatasheet.com/datasheet-pdf/vi … NSEMI/D45H.html

Tetrium wrote:

ramiro77 wrote:

Yes, this setup worked for five years until I decided to buy a new computer. But I never paid attention to it. Now I'm concerned about motherboard lifespan. The regulator gets ridiculously hot and I can't remember how it worked with the original Pentium 100.
Would you recommend me downgrading to some P54C or it's safe to stick with P55C?

I'd say just keep the MMX chip in your board and lower the multiplier to 2x so it runs at 133MHz for now.

Why 133 mhz? I understand that temperature rises most with voltage than frequency but maybe I'm wrong. And CPU temperature isn't a problem here because I'm using that fancy oversized socket 370 cooler. It's really really cold, it barely warms up. Running it at 133 mhz causes less stress to the motherboard? I'm intrigued on this.

Because the voltage regulator will have to dissipate all the excess heat the CPU uses and the more power the CPU uses, the more hot the voltage regulator will become and the more heat the regulator heatsink will have to dissipate.

It's true that heat by voltage increases exponentially and the heat by increased frequency only liniar, but as you can't decrease the amount of heat the voltage regulator has to dissipate by increasing the voltage supplied to the chip, decreasing it's frequency is the only other option.

I agree with what was suggested, add a fan very close to the voltage regulator heatsink so it loses it's heat much more rapidly. I think some overclockers of the old age used to replace the tiny stock heatsink with a beefier one (I would've used one from a dead PSU for example), but adding a fan will still be needed as a large heatsink will be mostly rendered useless if there is no airflow anyway.

And I'd suggest you use active cooling as the airflow from the CPUs HSF might help cool the voltage regulator heatsink.

Lower CPU current draw -> Lower regulator dissipation.

ramiro77 wrote:

CPU temperature was never a problem. In fact it worked with a passive heatsink all those years. I'm concerned about the heatsinked transistor.

Exactly... The lower the CPUs TDP, the lower its current draw, and therefore the less the regulator has to dissipate.

Allright guys! I'll give it a shot tomorrow. It's too late here. I'm so sleepy right now and that is bad for safe working jajaja.
Thanks to all of you! I'll keep updating this.

Well guys, I have bad news. I only managed to get 166 mhz with 2x multiplier. It doesn't get 133 mhz. The transistor is still getting alarming temperatures. I tried lowering the cpu voltage from 3.49v (VRE) to 3.38v (STD/VR) but still no luck. I'm starting to realise that I will have to change that little tiny heatsink. 😢

Little update: I managed to run it at 150 mhz (60 x 2.5). It seems that my SL27J is locked at 2.5x and 3.0x multipliers. I saw that on many posts in this forum. What a pity.
Major update: I set jumpers to 50 mhz bus and 2.0x multiplier (while my cpu recognizes as 2.5x multiplier). I was expecting to running it at 125mhz. But it booted at 133 mhz!!! 😕 😕 😕 I don't know why. Maybe you can put some light on this. Perhaps there is some black magic on this motherboard? Jajajaja.

Transistor is still reaching high temperatures. I don't feel any changes touching it. I'll try to get a temperature probe and do some measurements. And I'm looking for some heatsink to cut and give proper shape to fit in that space.

Hi again folks! Here is an update on this project:

I couldn't find screwed posts to attach the motherboard to the case. So I manage to do it with nuts and washers. It looks ok. I had to do some drilling and cutting. I'm not happy at this because I wanted to keep all as original as posible, but I would never go back to those horrible plastic inserts so proper screwing is a must.

Then I fitted a brand new speaker. Bought it today.

Motherboard, hard drive, vga, pci usb card and sound card mounted:

Next step: Finish assembling the power supply unit and mount it.

ramiro77 wrote:

Transistor is still reaching high temperatures. I don't feel any changes touching it. I'll try to get a temperature probe and do some measurements. And I'm looking for some heatsink to cut and give proper shape to fit in that space.

I'm not good at cutting metal, but you could cannibalize some other cooler for this (if you're able to saw it in smaller bits). Kinda like those little taller VGA memory heatsinks?
An old CPU heatsink could also work well, but you'd have to be willing to sacrifice one. One other option I considered myself is to use the (often for a PSU crappy) heatsinks that are used inside PSUs, it's no loss if you use one from a PSU that's already written off 😀

Hi tetrium! I'm struggling with that right now. I can't find anything that would fit properly. I have tools for pretty much any kind of work.
I'll reply to this later. I will not give up on this issue.

I did some other things today. I repaired an old pair of amplified speakers that matches perfectly with the case and sounds pretty good. And I finished assembling the psu.

PSUs also have regulators/transistors. You can try opening some busted PSU and see if you can find a fitting heatsink to use/cut.

Hi alexanrs! Yes, I know that. But I don't have any of those.
I have found a good heatsink to cut. I'm fighting against it right now. It's starting to look nice. I'll maybe finish it for today. Still have to finish cutting, then do some sanding and cleaning with caustic soda to remove anodized paint.
Meanwhile some pictures:

Here I am soldering the case speaker. It's brand new.

And this is the PSU. I lubricated the fan with some good synthetic oil for engines. It runs almost noiseless.

Well. It's done 😎
I went from something like this (illustrative image):

to this:

It took me something like four hours. I didn't use caustic soda, just a lot of sanding. I kept some black anodized finish between fins. Now I have to buy a 40 milimeter fan to put in there. I tested it with a noisy damaged one and the heatsink remains cold in almost every condition.
Today I also did some wire soldering in the front panel. I can't find pin connectors for the motherboard in every store I ask for it. It's a little weird, I bought a bunch of them some months ago.

ramiro77 wrote:

Well. It's done :cool: I went from something like this (illustrative image): […]
Show full quote

Well. It's done 😎
I went from something like this (illustrative image):

to this:

It took me something like four hours. I didn't use caustic soda, just a lot of sanding. I kept some black anodized finish between fins. Now I have to buy a 40 milimeter fan to put in there. I tested it with a noisy damaged one and the heatsink remains cold in almost every condition.
Today I also did some wire soldering in the front panel. I can't find pin connectors for the motherboard in every store I ask for it. It's a little weird, I bought a bunch of them some months ago.

This looks really good actually! 😁

Just beware that you should touch the VRM itself if possible as a cold heatsink may indicate poor heat transfer from the hot item to the heatsink.

Thank you tetrium! It does look good indeed 😎
And yes, I'm aware of that. I touched the regulator and it's ok. Without fan the heatsink goes a little hot. Not as the original heatsink, but I think it will be neccesary to keep some 40 mm fan as it keeps it really cold.
Now I'm doing some tests at 200 mhz with the damaged fan and it's all ok. Still waiting to find some terminal connectors to restore the front panel.

Other topic: I just found that my motherboard doesn't have turbo switch connector. But I do remember the switch beeing conected to led display AND to the motherboard with 3 pin connector in both. I just can't remember where it was connected, and the manual doesn't say anything about it. I just read this:

TB_LED: Indicates if the system speed is in normal or turbo...

So I can assume it does have some turbo function. But the only "misterious" pins are two between speakers connected to nothing (I've checked tracing the motherboard). Then two pins labelled "OUTLET" (unknown function) and five pins labelled "MS_JMP" but I think they are for mouse or something.
I'm a little confused here... I can always play with fsb, multipliers and switchs, but I do remember turbo button working on this computer. For the record, it wasn't connected to fsb and/or multiplier. Those are two pins rows and the turbo connector has three pins.