Seemingly unique ASRock C2750D4I voltage problem

[rayrayrayraydog]

I am fully aware that this board is now ancient and had read up on the C2000 bug, but I thought I would try it anyway since I found a "pulled from a working server" one on eBay for very cheap.

Unfortunately, it won't boot up. The IPMI will start and I can log into it. The firmware version is at 0.30.0 so it seems I should not be stuck with the C2000 death. What is happening though, is the sensors for Vcore, +1.0, and +1.1 are all reading at 0 volts. I have tried with two known working power supplies, replaced the CMOS battery, and tried reverting to factory defaults with no luck. The IPMI thinks the system is powered on, but there is no POST or display of any kinds. I can power cycle the system from the menus but still nothing.

Has anyone seen an issue like this and/or been able to recover from it?

 

[rayrayrayraydog]

The only variable I cannot eliminate here is the RAM, so I've ordered a pair of ECC DDR3 UDIMMs explicitly listed on ASRock's HCL. Hopefully that helps!

 

[rayrayrayraydog]

Sorry to say, it's not the RAM. I've installed a pair of CT102472BD160B.18FED ECC UDIMMs straight from the ASRock memory QVL with no luck. Any ideas? I may try to reflash the BIOS chip but it seems most likely the board took some damage.

 

[Ericloewe]

The firmware version is at 0.30.0 so it seems I should not be stuck with the C2000 death.

Why do you say that? There is no correlation between those two things.

What is happening though, is the sensors for Vcore, +1.0, and +1.1 are all reading at 0 volts.

If the system doesn't post, I would not be surprised to see sensors out of whack. That's very typical, as many sensors are only enabled once the host is running.

Sorry to say, it's not the RAM. I've installed a pair of CT102472BD160B.18FED ECC UDIMMs straight from the ASRock memory QVL with no luck. Any ideas? I may try to reflash the BIOS chip but it seems most likely the board took some damage.

Not surprising, as the system would likely have beeped in response to bad RAM.

There's really no reason to suspect anything other than a dead LPC clock or two on the SoC, as is widespread in C2000 units. The repair isn't too difficult on the ASRock boards, so it's something you may want to investigate.

 

[rayrayrayraydog]

Thank you for replying! I had read that boards on BMC 0.30.0 and higher were not affected, but sure, that is probably wrong.

Is this the fix you mean? https://www.truenas.com/community/threads/how-to-fix-asrock-c2750d4i-with-c2000-bug.90451/ It seems like it must be but just to be sure. I should be able to do this. :)

 

[Arwen]

Their were 2 problems with the ASRock boards, one from the Intel SoC, and the other from the firmware.

If I remember the firmware problem correctly, is that it writes to the EEPROM far too much, exceeding the 100,000 writes that the device can handle. The updated firmware stops that. I don't remember anything else.

 

[rayrayrayraydog]

Thanks everyone. I'm going to try a couple fixes and then cut my losses. This board is from 2014 anyway. I bought a Jonsbo N3 case and should have instead gone MicroATX of course. For ITX and support for 8 SATA drives it seems this is the only game in town for something new in 2023: https://www.amazon.com/N5105-N6005-Demon-SATA3-0-I226-V/dp/B0BZ3DJF7J/ I'll start the research on these but if anyone knows of a better option I'm all ears! :)