Problems with ASRock 775i65G, anyone still running one?

Vinny N

Platinum Member
Feb 13, 2000
2,277
1
81
Everything is running at stock or underclocked (E6300 @ 1.76ghz, DDR400 at DDR333 as a result to conform to SPD timings rather than chance the DDR355 at unknown CAS specs)

The 3.3v rail is sagging to 3.25 typically with lows of 3.23 and occasioanlly 3.2v. I had this problem before and had to RMA the power supply (it was a Antec basic 380W, they sent me a new EA430).

The 380W PSU's 3.3v was barely giving 3.15-3.20 by the time I had to RMA it...but it showed a fine 3.3v rail in another system...

I remember the 775i65G on the new PSU it started off around 3.28v in BIOS (speedfan reported it as 3.26).

There was a time during the RMA that I used another PSU from an old P4 system. It reported the highest 3.3v rail I saw (3.29/3.30 in bios, 3.28 by speedfan). But I was way underclocked to 1.4ghz to enable the onboard video, (didn't have a pci express power connector for my AGP HD2600). I didn't run it long enough to see if it would start sagging too...

I don't see any bloated, leaking or popped capacitors on the board and it was otherwise stable for over a year on the EA430 until yesterday when it had trouble posting after a reboot.

Is it possible the motherboard is slowly killing this PSU?

I think multimeter reads were showing as normal before in all cases...and thought perhaps it was just the ATX power connector but it shows no signs of overheating/burning (I remember seeing singed 5v connectors back in the day on some bad boards killing PSUs...)
 
Last edited:

tcsenter

Lifer
Sep 7, 2001
18,345
256
126
The lowest voltage discrepancy you've reported is well within spec. And using the built-in hardware monitor (whether by BIOS or other software) to compare voltage supplied by the PSU across different motherboards is highly flawed, since the discrepancy or margin of error of the hardware monitoring circuit used on most motherboards exceeds the variation allowed by ATX specs.

If direct multimeter reading shows more consistency (within normal values), put a higher confidence on that and discount the fickle motherboard hardware monitoring values.
 

Vinny N

Platinum Member
Feb 13, 2000
2,277
1
81
Ah, I had a feeling you might say that about the tolerance of the spec.

Guess I'll be firing up a multimeter tonight to check it. Thanks!