Kreshna Aryaguna Nurzaman wrote on 2020-08-28, 15:18:

Thanks, PTherapist. Now, what if I use AMD K6-2 300 instead? […]
Show full quote

Thanks, PTherapist. Now, what if I use AMD K6-2 300 instead?

I will use the same setting you used; I will set FSB to 50MHz, so the CPU would run at 300 MHz. But unlike you, I will not underclock the K6-2, because it is already a 300 Mhz CPU anyway. As such, I will overvolt the CPU (because I will not use VRM), while not underclocking it, because it's a 300 Mhz CPU anyway.

Will I have problems?

The reason I'm asking this is because AMD K6-2 300 is cheaper and more plentiful than AMD K6-2 500 anyway. Well at least on Ebay.

Not entirely sure, but it may not run at 300MHz, you might end up with a clock speed somewhere between 150-200MHz. The K6-2 300 uses a 4.5x multiplier which this motherboard does not officially support, thought it might if configured to 2.0x or 1.5x - in which case you might be ok, if it's stable at that top speed.

Take the K6-2 400 in my example for instance, that will not run at 300 and at 50MHz FSB I was getting below 200MHz out of it. Had to opt for 210MHz @ 60MHz FSB. If I wanted to take it up to 300MHz, I'd have to go with another K6-2 500 CPU instead.

The Serpent Rider wrote on 2020-08-28, 15:33:

Or, in PTherapist's case, it's 3.3v.

Nah, 3.3v is too much for K6-2 for any prolonged use. Short term benchmarking - sure. Staying on 50% over default voltage (2.2v) is dangerous. And realistically it would be not 3,3v, but 3,45v on single rail boards, so already over 50%.

Well unless my board is severely underpowering the CPU, I ran a K6-2 500 with 3.3V for years and the CPU & motherboard still work fine. The PC was in active use for about 2 years with this configuration. The CPU itself never got noticeably hot, just normal working temperature. CPU was underclocked by about 200MHz, however.

PTherapist wrote on 2020-08-28, 15:41:

Kreshna Aryaguna Nurzaman wrote on 2020-08-28, 15:18:

Thanks, PTherapist. Now, what if I use AMD K6-2 300 instead? […]
Show full quote

Thanks, PTherapist. Now, what if I use AMD K6-2 300 instead?

I will use the same setting you used; I will set FSB to 50MHz, so the CPU would run at 300 MHz. But unlike you, I will not underclock the K6-2, because it is already a 300 Mhz CPU anyway. As such, I will overvolt the CPU (because I will not use VRM), while not underclocking it, because it's a 300 Mhz CPU anyway.

Will I have problems?

The reason I'm asking this is because AMD K6-2 300 is cheaper and more plentiful than AMD K6-2 500 anyway. Well at least on Ebay.

Not entirely sure, but it may not run at 300MHz, you might end up with a clock speed somewhere between 150-200MHz. The K6-2 300 uses a 4.5x multiplier which this motherboard does not officially support, thought it might if configured to 2.0x or 1.5x - in which case you might be ok, if it's stable at that top speed.

Take the K6-2 400 in my example for instance, that will not run at 300 and at 50MHz FSB I was getting below 200MHz out of it. Had to opt for 210MHz @ 60MHz FSB. If I wanted to take it up to 300MHz, I'd have to go with another K6-2 500 CPU instead.

I see, thanks. I presume AMD K6-2 at 200 MHz will still be faster than Pentium MMX at 200 MHz, no? I'm not aiming to run unaccelerated Quake at 800x600 though. I just wonder how smooth (or how jerky) System Shock at 640 x 480 on 200 MHz AMD K6-2, as well as F-22 Lightning II.

In any case, is it difficult to build our own VRM using PMMX regulator module and 30-pin IDC header connector as gdjacobs might implied?

gdjacobs wrote on 2020-08-28, 15:16:

Kreshna, have you got an existing PMMX regulator module? If it's linear, it might be straightforward to modify the set point and add some cooling to the pass element.

Once again, realistically clock speed doesn't matter much - main source of heat is massively increased voltage. Heat dissipation increase under voltage increase is not linear.
Lowering temperature also won't help much and you can quite literally kill CPU by overvolting even under top notch water cooling. Only exception to that is when you're going below zero Celsius, which cause superconductivity and upper voltage limit for transistors is much higher.

EDIT:

Well unless my board is severely underpowering the CPU, I ran a K6-2 500 with 3.3V for years and the CPU & motherboard still work fine.

Usually CPU won't just suddenly die. Borderline voltage will just slowly degrade silicon and upper working-overclocking limit will be reduced over time.

The Serpent Rider wrote on 2020-08-28, 15:54:

Once again, realistically clock speed doesn't matter much - main source of heat is massively increased voltage. Heat dissipation […]
Show full quote

Once again, realistically clock speed doesn't matter much - main source of heat is massively increased voltage. Heat dissipation increase under voltage increase is not linear.
Lowering temperature also won't help much and you can quite literally kill CPU by overvolting even under top notch water cooling. Only exception to that is when you're going below zero Celsius, which cause superconductivity and upper voltage limit for transistors is much higher.

EDIT:

Well unless my board is severely underpowering the CPU, I ran a K6-2 500 with 3.3V for years and the CPU & motherboard still work fine.

Usually CPU won't just suddenly die. Borderline voltage will just slowly degrade silicon and upper working-overclocking limit will be reduced over time.

Well if the degradation take years then it's good enough for me.

What board have you got?

Edit: NM, I see.

I think it could definitely be done. You'd probably want to use an adjustable voltage regulator chip driving an external pass element. Ideally it would have an output pin indicating operation in or out of regulation. You'd probably want two MOSFETs to float the voltage output pins and another to clamp the adjustment pin to ground when the module is disabled.

gdjacobs wrote on 2020-08-29, 02:19:

What board have you got?

PC Chips M507. Wait, I understand. You meant modifying the mobo's internal VRM, didn't you? The one that enables us to select between 3.5v and 3.3v?

gdjacobs wrote on 2020-08-29, 02:19:

I think it could definitely be done. You'd probably want to use an adjustable voltage regulator chip driving an external pass element. Ideally it would have an output pin indicating operation in or out of regulation. You'd probably want two MOSFETs to float the voltage output pins and another to clamp the adjustment pin to ground when the module is disabled.

Thanks. I still need to use the mobo's VRM slot, do I not? So, I would need a 2x15 pin female header connector for the purpose, wouldn't I?

This header should work, no?

30 pins. Should be good.

Is it worth all this hassle, though? Why not just replace witha a K6-2 capable motherboard?

gdjacobs wrote on 2020-09-04, 04:13:

30 pins. Should be good.

Thanks!

appiah4 wrote on 2020-09-04, 05:01:

Is it worth all this hassle, though? Why not just replace witha a K6-2 capable motherboard?

From strictly a strict cost-benefit point of view? I think not. In fact, I already have my DOS-capable machine, based on Pentium II CPU and 440BX motherboard. It runs non-accelerated, textured DOS SVGA games pretty well (my Voodoo5 5500 has become nothing but glorified VESA card, really). In fact, an 440BX system is more straightforward, without having to wrestle with CPU voltage, VRM, and such. You just plug the correct CPU, and voila! A system fast enough to run texture-mapped, non-accelerated DOS game in 640x480 resolution.

But where is the fun in that?

First, this board holds sentimental value. Next, I have never build anything older than 440 BX, so this is a good chance to try.