Andry wrote on 2025-04-11, 15:15:

Hi to everyone. […]
Show full quote

Hi to everyone.

I'm playing with my Soyo SY-5EH5 1.0 motherboard.

https://theretroweb.com/motherboards/s/soyo-sy-5eh5-1-0

I want to replace current CPU AMD K6-2 500 CPU by a K6-3 450.

I suspect that CPU voltage is less than set by jumpers, because K6-2 does not boot at stock voltage of 2.2V, but it is needed to set 2.3V. K6-3 requires 2.4V, but it boots starting from 2.6V and it is unstable.

I'd like to measure CPU core voltage output, but I'm not able to understand where take the voltage on the MB.

Does someone know where I can get VCore output value?

Thank you,

Andrea

Hi,

So, I don't know if that motherboard has secret VCORE settings, but it could be possible to modify the VCORE, to get the 2.4V necessary for the K6-3, however if you want to measure the voltage of the CPU core, you can do it fairly safely, putting the black tip of the tester, in one of the molex of the drive, the red tip goes on the top of the Mosfet, in one you will find about 2.2V, that's the VCORE.

I forgot that I checked the manual a little while ago, and I found the page, where there are all the voltages of the VCORE, and the relative positioning of the switches.
As you can see, you can select from 2.0V up 🔝, if you select 2.4V and the measurements on the Mosfet, are far below 2V, it is obvious that there is a problem in the VRM, but this problem can have different causes, so first thing 😨, you have to try to see if changing settings, you have the correct voltage 😬 VCORE.

PC@LIVE wrote on 2025-04-11, 19:39:

I forgot that I checked the manual a little while ago, and I found the page, where there are all the voltages of the VCORE, and the relative positioning of the switches.
As you can see, you can select from 2.0V up 🔝, if you select 2.4V and the measurements on the Mosfet, are far below 2V, it is obvious that there is a problem in the VRM, but this problem can have different causes, so first thing 😨, you have to try to see if changing settings, you have the correct voltage 😬 VCORE.

Hi and thank you very much for your help!

My MB version is 1.0, so CPU VCore is set by jumpers and not by switches. However, supported range is ever from 2.0V to 3.5V.

Well, these are the average measurements on the Mosfets (it is slightly unstable, +- 0,1V)

(K6-3 450 CPU 2.4V VCore):

2,0V VCore -1,78V / 4,60V (no boot)
2,2V VCore -2,12V / 4,74V (no boot)
2,4V VCore - 2,43V / 4,82V (post only)
2,5V VCore - 2,38V / 4,73V (boot, unstable)
2,6V VCore - 2,62V / 4,80V (boot, unstable)
2,7V VCore - 2,68V / 4,83V (boot, unstable)

(K6-2 533 CPU 2.2V VCore):

2,0V VCore -1,90V / 4,76V (450 MHz boot, stable)
2,2V VCore -2,13V / 4,74V (500 MHz boot, stable)
2,2V VCore -2,13V / 4,74V (523 MHz FSB 95 MHz boot, stable)
2,2V VCore -2,13V / 4,74V (550 MHz Win98 does not boot)
2,3V VCore -2,22V / 4,90V (550 MHz boot, unstable)
2,4V VCore - 2,38V / 4,70V (550 MHz boot, unstable)

I seems that the K6-3 450 is fried.

K6-2 533 is not stable at 550 MHz. I didn't try over 2.4V, it is OK at 500 MHz.

What do the voltages behind the slashes mean?

majestyk wrote on 2025-04-12, 13:17:

What do the voltages behind the slashes mean?

In my opinion they are the measurements on the two Mosfet, you can see them in the photo of the first messages, apparently everything should be quite normal, but if there is instability, maybe 🤔 could it be another problem?

Andry wrote on 2025-04-12, 12:49:

Hi and thank you very much for your help! […]
Show full quote

PC@LIVE wrote on 2025-04-11, 19:39:

I forgot that I checked the manual a little while ago, and I found the page, where there are all the voltages of the VCORE, and the relative positioning of the switches.
As you can see, you can select from 2.0V up 🔝, if you select 2.4V and the measurements on the Mosfet, are far below 2V, it is obvious that there is a problem in the VRM, but this problem can have different causes, so first thing 😨, you have to try to see if changing settings, you have the correct voltage 😬 VCORE.

Hi and thank you very much for your help!

My MB version is 1.0, so CPU VCore is set by jumpers and not by switches. However, supported range is ever from 2.0V to 3.5V.

The attachment Screenshot 2025-04-12 100953.png is no longer available

Well, these are the average measurements on the Mosfets (it is slightly unstable, +- 0,1V)

(K6-3 450 CPU 2.4V VCore):

2,0V VCore -1,78V / 4,60V (no boot)
2,2V VCore -2,12V / 4,74V (no boot)
2,4V VCore - 2,43V / 4,82V (post only)
2,5V VCore - 2,38V / 4,73V (boot, unstable)
2,6V VCore - 2,62V / 4,80V (boot, unstable)
2,7V VCore - 2,68V / 4,83V (boot, unstable)

(K6-2 533 CPU 2.2V VCore):

2,0V VCore -1,90V / 4,76V (450 MHz boot, stable)
2,2V VCore -2,13V / 4,74V (500 MHz boot, stable)
2,2V VCore -2,13V / 4,74V (523 MHz FSB 95 MHz boot, stable)
2,2V VCore -2,13V / 4,74V (550 MHz Win98 does not boot)
2,3V VCore -2,22V / 4,90V (550 MHz boot, unstable)
2,4V VCore - 2,38V / 4,70V (550 MHz boot, unstable)

I seems that the K6-3 450 is fried.

K6-2 533 is not stable at 550 MHz. I didn't try over 2.4V, it is OK at 500 MHz.

Well honestly I didn't notice the difference, however the final result is identical, instead of moving the switch, you have to move the jumpers.
The results of the measurements, they don't seem strange to me, maybe 🤔 the instability you notice, could have other causes, like some hardware or software problem, sometimes there are big problems between drivers and cards, the ETEQ chipset maybe it's an VIA (which they changed their name), there are several problems that can arise, only in some special cases can be solved, personally I have similar cards that I haven't yet ⚓️ put in place.
Usually the main problem is with some video cards, one among many the TNT2, while a GF2MX should work, but only by installing the drivers in the right order, and using those that work in similar PCs, not all versions are good 😌, and it is not said that the latest version is the best.

If the K6-III won´t run stable, but the K6-2 does my guess would be defective electrolytics in the VRM section.

In PC@Live´s picture above they are all defective. Andry, could you take some pics of your mainboard, please?

majestyk wrote on 2025-04-13, 18:58:

If the K6-III won´t run stable, but the K6-2 does my guess would be defective electrolytics in the VRM section.

In PC@Live´s picture above they are all defective. Andry, could you take some pics of your mainboard, please?

That could be the problem, but I put the image you see, to show where to measure the VCORE.
Undoubtedly, if Andry puts pictures of his MB, you could see if there are electrolytic capacitors to replace.

majestyk wrote on 2025-04-13, 18:58:

If the K6-III won´t run stable, but the K6-2 does my guess would be defective electrolytics in the VRM section.

In PC@Live´s picture above they are all defective. Andry, could you take some pics of your mainboard, please?

Caps seems good at visual inspection, but I will do.

However your assumption may be right, because the same K6-3 CPU runs fine on another motherboard (I tried that yesterday).

Caps seems to be good.

After reading many threads I've come to the following conclusions:

- if the cpu isn't supportd by the bios, it can hang at various stages and/or be unstable in DOS
- if supported, voltage has issues with electrolytics and/or VRM circuit
- also, bad crystal oscillator/ frequency generator can create issues (usually not with post)

This is a partial recap (no pun intended) of suggestions.

Personally I ended up replacing all VRM caps on similar boards.

Andry wrote on 2025-04-18, 11:34:

The attachment Caps.jpg is no longer available

Caps seems to be good.

They don't seem bad to see them.
But it does not mean that they are in perfect condition, they could now be at the end of their life, a slightly high ESR, or a capacity out of tolerance, they could be the signs that someone or everyone is to be replaced (?).