It might well be that your power supply doesn't deliver enough power on the 5V rail. What is it rated at?
Try to clock it manually at the same speed or lower than you Duron to confirm the CPU is ok, 600MHz is a good test for example.

Did you have a look at https://theretroweb.com/motherboards/s/soltek-sl-75kiv#bios for the BIOS versions? Is one newer than yours?

If cpu was not confirmed to be working by the seller, it probably is dead.

it would be nice to submit your board photo to the retroweb since they don't have one.

This is probably not relevant to your issue, but it's reminded me about my experience with a similar Soltek board back in the day, the SL75DRV2. That thing was so unstable with my Athlon Thunderbird 1000@1200 it was embaressing, and I'd had no problems up until then running that chip in an ABIT KT7. I remember distinctly how unhappy I was with it and returned it for an ABIT KR7A, which of course was great.

Anyway, hopefully it's not just a Soltek board doing Soltek board things and my experience was an exception!

Minutemanqvs wrote on 2024-02-28, 17:14:

It might well be that your power supply doesn't deliver enough power on the 5V rail. What is it rated at?
Try to clock it manually at the same speed or lower than you Duron to confirm the CPU is ok, 600MHz is a good test for example.

Did you have a look at https://theretroweb.com/motherboards/s/soltek-sl-75kiv#bios for the BIOS versions? Is one newer than yours?

The PSU is a Corsair CX650 with 25A on the 5V rail. I thought that was enough, but just to double-check, I’ve removed all other PCI devices, stuck a potato TNT2 M64 instead of the GF2 GTS, and tried firing it up with auto settings, and then also with 6x multiplier - sadly the results are the same. I would assume with those conditions, the PSU would be enough, right?

Regarding the BIOS I checked it in the wrong place, the revision is actually from the 5th of January 2001, so that could do with an update, still, it’s weird that the Duron 850 was released later, and still works.

Oh and the seller sold it as working, so I’m just trying to consider every possibility before deciding to send it back.

botond87 wrote on 2024-02-28, 22:23:

The PSU is a Corsair CX650 with 25A on the 5V rail. I thought that was enough, but just to double-check, I’ve removed all other […]
Show full quote

Minutemanqvs wrote on 2024-02-28, 17:14:

It might well be that your power supply doesn't deliver enough power on the 5V rail. What is it rated at?
Try to clock it manually at the same speed or lower than you Duron to confirm the CPU is ok, 600MHz is a good test for example.

Did you have a look at https://theretroweb.com/motherboards/s/soltek-sl-75kiv#bios for the BIOS versions? Is one newer than yours?

The PSU is a Corsair CX650 with 25A on the 5V rail. I thought that was enough, but just to double-check, I’ve removed all other PCI devices, stuck a potato TNT2 M64 instead of the GF2 GTS, and tried firing it up with auto settings, and then also with 6x multiplier - sadly the results are the same. I would assume with those conditions, the PSU would be enough, right?

Regarding the BIOS I checked it in the wrong place, the revision is actually from the 5th of January 2001, so that could do with an update, still, it’s weird that the Duron 850 was released later, and still works.

Oh and the seller sold it as working, so I’m just trying to consider every possibility before deciding to send it back.

This is indeed more than enough power. The only last thing to try is indeed updating the BIOS (I doubt it would change anything), and then if you can test the cPU on another mainboard. But it's probably dead.

It's hard to tell what it is , it might be bad vrm that cannot deliver enought power , or bad cpu or kind of incapabilites.

any slight discoloration around the cpu core? what about chipped corners? pictures?

Thanks for all the input, I’ll try my luck with a BIOS update.

stanwebber wrote on 2024-02-29, 07:44:

any slight discoloration around the cpu core? what about chipped corners? pictures?

I didn’t see anything that would indicate physical damage, take a look if you can spot any possible problems:

Untested Athlons and XP's have a high defect rate in my experience. It probably died to overheating. There are often no visual signs of damage or discoloration, it simply won't POST anymore. No oher way to find out except testing it in another board.

If the seller sold it as working, that often doesn't mean much unless he's also an enthusiast. Could be someone who sells everything as working, and simply refunds anything that doesn't work. I've seen Athlons with completely chipped off dies sold as working.

do the bios update, if only to verify the duron still boots in the motherboard. is there a motherboard jumper for 100/133mhz fsb operation? the A1000AMT3B has a multiplier of 10x so it has to run at 100mhz fsb unless unlocked.

stanwebber wrote on 2024-02-29, 09:40:

do the bios update, if only to verify the duron still boots in the motherboard. is there a motherboard jumper for 100/133mhz fsb operation? the A1000AMT3B has a multiplier of 10x so it has to run at 100mhz fsb unless unlocked.

Yes it has a switch, it was set to 100Mhz by default, tried with both 100 and 133 (just to check if it changes anything), none of them worked.

Well after reading about the flash procedure and risks (the oldest board I flashed thus far was an LGA775), I might just get a refund and call it a day. Idon’t want to brick a working board for a CPU that’s the fifth of the price of the mobo, might just try my luck with another CPU. Unfortunately I don’t have another socket A board to mess around with.

The seller specialises in retro CPUs, so I’m willing to believe that he tested it, but I did see in his (100% positive) feedbacks, that some part was not working and issued a refund. Might have gotten simply unlicky with this one.

Thank you all for your help, much appreciated!

Can happen - I recently got a 9800SE AIW. Seller tested it and even sent me a video.
Installed it into three different boards - artefacts all over. Didn't get a refund or to send the card back so I just keep it in the event that I can reball it.

as long as the bios chip isn't soldered to the board the risk of flashing is non-existent. even a bad flash is recoverable so long as the bootblock is intact. in the event of a catastrophic failure you can still recover the bios in an external flasher or in ANY other motherboard with a compatible socket via a hot swap flash.

i have easily flashed over 100 firmwares and have never had an (unintentional) failed flash. i have, however, had numerous bios chips spontaneously corrupt on me from doing nothing and have hot flashed all of the original chips with success.