Did you use the same PSU in both cases?

That time feels like if the ATX power switch is stuck shorted.. It would power on then force a power off after a few secs.. I'd try a different PSU or momentarilly shorting the power header with a jumper and releasing it quickly.

Also, does the same happen without any CPU installed?

Shponglefan wrote on 2026-02-28, 14:35:

Did you use the same PSU in both cases?

Hi, I tried 3 different PSU's. All 3 are confirmed working.

Cheers,

wbahnassi wrote on 2026-02-28, 14:58:

That time feels like if the ATX power switch is stuck shorted.. It would power on then force a power off after a few secs.. I'd try a different PSU or momentarilly shorting the power header with a jumper and releasing it quickly.

Also, does the same happen without any CPU installed?

Hiya, I tried 3 different PSU's. All confirmed working units.

Same symptoms without the CPU installed.

Cheers,

Does it turn on the moment the ATX PSU's hard switch is turned on? Is there continuity between the two pins of the power header on the front panel connectors block?

wbahnassi wrote on 2026-03-01, 05:26:

Does it turn on the moment the ATX PSU's hard switch is turned on? Is there continuity between the two pins of the power header on the front panel connectors block?

No, I have to push the power button on the case to power the system on. No continuity between the two pins of the power switch header on the motherboard.

Cheers,

In general I find it unlikely that a CPU can cause MB failure, especially when it is a correct CPU for the platform. Your case just sounds a coincidence.

Your problem sounds like a power delivery issue, like perhaps a regulator is failed. It also just may be that the new CPU has higher power consumpition and it pushed the already failing component over the edge.

Did you test it before as a bare motherboard? Do you now have it installed in a case?

CharlieFoxtrot wrote on 2026-03-01, 06:44:

In general I find it unlikely that a CPU can cause MB failure, especially when it is a correct CPU for the platform. Your case just sounds a coincidence.

Your problem sounds like a power delivery issue, like perhaps a regulator is failed. It also just may be that the new CPU has higher power consumpition and it pushed the already failing component over the edge.

This.

It would be good to test the CPU on another board but you gotta work with what you have.
Even if the K6-III/333 isn't faulty it is still one of the more power hungry socket 7 CPUs.
I'd be checking the power delivery on the board, particularly the 5 big caps next to the CPU socket. They're from the era of limited lifetime caps.

Take the motherboard out, check for shorts on the bottom, check all metal standoffs are in the right place...

1 Congrats on a CPU that officially doesnt exist according to AMD 😀 its the year 2000 Fry's Electronics Black Friday special. Afaik available only there and then for cool $29.
2 diagram for GIGABYTE GA-5AA rev3.2 Re: 430VX or 580VPX schematics and post describing 2.2 vs 3.2 differences Gigabyte GA-5AA revision 1.1, 2.2 and 3.2 differences

on Ver2.2 CPU Vcore is generated by non stnchronous DC/DC converter. U11 controller + q8 mosfer + D10 diode + L10 coil.
You can check if everything is working by measuring voltage on L10 coil and on pin2 of U11.
Personally I would start by recapping this board. Dont fret too much about special brands, any cheapest lowesr caps with more or less similar parameters will do. Even the crappiest caps wont see hard load here and will deliver thousands of hours of continued use, more than you and your children would ever use this board 😀
TC23 _non_ lowesr 330uF 6.3V, pretty much any random cap
TC24 TC39 TC40 C114 and 5 ones along CPU socket all lowesr 1000-1500uF 6.3V

but shutting down after 3 seconds I suspect might be due to shorting to something inside the case, maybe one of the motherboard metal studs didnt line up with hole? try with mobo out of the case

Hi all,

A follow up with this issue, I managed to find another super socket 7 motherboard which was confirmed to work. Tested with the CPU and it worked. I guess it was just by coincidence the motherboard died from the time I tested it. 🙁
I can confirm I tested the board outside of the case when it refused to boot the first time. There are no extra stand offs or anything touching that could have shorted the motherboard. I suspect it's just component failure on the board.

Good news, the next step is try to repair the motherboard since it's one of the best ss7 one and deserve to be saved!

Nemo1985 wrote on 2026-03-14, 10:24:

Good news, the next step is try to repair the motherboard since it's one of the best ss7 one and deserve to be saved!

Absolutely! Would really like to get it up and running as it's the same as what I had back in the day. 😀

rasz_pl wrote on 2026-03-01, 12:06:

1 Congrats on a CPU that officially doesnt exist according to AMD 😀 its the year 2000 Fry's Electronics Black Friday special. Afaik available only there and then for cool $29.

I certainly did not buy my K6-3 333 MHz on black friday 2000. I know what I bought on that day 😀 They had these chips available through summer, if not most of the year. If they had this in their black friday 2000 sale(?), then it was the last few left. I'm pretty sure I remember someone there tell me they were trying to get rid of them on my earlier visits. Unsure why it didn't attract more attention, but I guess in 2000, Athlon was out, and getting price cuts. If anything, it would have been the Athlon on sale black friday 2000 because I know some speeds were sub $200 by that November, so if they went on sale, that would have brought people in. The K6-3 333 MHz was priced lower than the K6-2 on 100 MHz FSB, due to the low speed. This is why I bought them. It would have been otherwise a well performing socket 7 chip, priced quite low. Still beaten though, by the highest clocked K6-2 and the other K6-3 in the store, if you wanted to spend more money.

Beerfloat wrote:

Even if the K6-III/333 isn't faulty it is still one of the more power hungry socket 7 CPUs.

CharlieFoxtrot wrote:

Your problem sounds like a power delivery issue, like perhaps a regulator is failed. It also just may be that the new CPU has higher power consumpition and it pushed the already failing component over the edge.

Being the slowest clock speed in the FSB 100MHz (95MHz actually) class of chips, it would have been the one with the least power draw and heat production. I know my chip runs quite cool. The K6-2 was the hotter chip, since it was clocked higher than this one. This kind of makes me wonder, that this production batch was intended as OEM for laptops, but the buyer bailed on it, just like the K6-2+ 570 MHz, and so became a Frys store special. This chip would been the coolest 100MHz class chip until the K6-2/3+ die shrink.

It could be a bad chip, but as other people are saying, I would think that based on the behavior, I would look at the motherboard and power supply for clean power delivery. That is usually the situation for things like this.

Choppo51 wrote on 2026-03-14, 06:51:

I suspect it's just component failure on the board.

You already got a very good advice on where to start to check to identify the failure. I would just like to add, that since that Gigabyte has already BGA chipset it's also possible that due to the handling of the board during tests/assembling some ball could lose contact from chip to PCB. You can try to press with your finger on top of the chipsets to try to force chip to get better contact to PCB (if only any of the balls is a culprit). If it boots then it will require reballing for chipset.