Weird problem with voltages on antec truepower 430

[SenorBeef]

I've got an antec truepower 430 running an MSI K8N NEO 2 PLAT mb.


About a year ago, I was having weird problems with temporary freezing in games. In trying to figure out the problem, I noticed my voltages were low - my 3.3 was around 3.1, my 5v was around 4.8, and my 12v was about 11.9.

I went through my case trying to see if there were any physical problems. I noticed that the blue wire from my power supply, labelled "PSU FAN MONITOR" wasn't plugged into anything. Must've slipped out, because I remember needing to plug that in.

Anyway, I found the right connector for it (SFAN2) and for whatever reason, suddenly it worked fine. My voltages were all within spec.

Now, later, I'm having the same problems. Little freezeups, crashes... so I check my voltage, and yep, it's low like it was. Only this time... that PSU monitor wire is still connected. So, there's nothing to plug in to fix it - I'm not sure what to do. Anyone have any ideas?

 

[Trap05]

All that wire does is allow the motherboard to monitor the power supply fan speed. Also are you getting those voltages with a multimeter or from the bios or a windows program? If it's a multimeter then they are a tad low but nothing terrible. Did you change anything recently? Add something? Change a driver? More details would be helpful

 

[SenorBeef]

I realize that I can't identify any logical connection to the temp monitoring cable and the output votages of the PSU, but there was a definite correlation - before, when I had this problem, hooking up the wire fixed my voltages and instability problems.

I'm getting the voltages from the bios screen and MBM - I don't have a voltimeter available.

I haven't made any changes that I'm aware of when this issue cropped up. Definitely no hardware changes - and I don't think driver changes either.

The voltages aren't too low, but they seem to correspond with the freezes and crashes. I have a 6800 GT which is pretty power hungry - I'm guessing low output voltages combined with high drain are causing my problems.

I really don't know *why* it fixed it before when I hooked up that wire - I just knew my voltages were below normal, and that games were having crash issues, and then I plugged it in, and suddenly my voltages were healthy, and the crash issues went away.

 

[Trap05]

I would either replace the power supply or buy a multimeter to test it under load.

 

[SenorBeef]

It seems clear that the voltage readings are accurate - when the voltages are normal, everything runs fine - when they're reported low, I have stability problems.

I just don't know why I had this problem before, but was able to fix it, but now it's back.

I may have to replace the unit... but I don't have another computer to test with, and it'd suck if it turned out to be the MB regulator or something that's broken and the PSU was fine.

 

[RollingThunder]

It seems clear that the voltage readings are accurate - when the voltages are normal, everything runs fine - when they're reported low, I have stability problems.

I just don't know why I had this problem before, but was able to fix it, but now it's back.

I may have to replace the unit... but I don't have another computer to test with, and it'd suck if it turned out to be the MB regulator or something that's broken and the PSU was fine.

Might help to list your whole system and overclocking specs. That way we can see the peripherals you are running plus whatever overclock you may have.

 

[SenorBeef]

I've actually got everything running at stock for the time being. A64 3200+ winchester, k8n neo2 plat, corsair and mushkin 3200c2 stuff, bfg 6800 gt, audigy 2 zs.

I managed to either make things better or screw them up worse. I pulled apart the power connectors (all of them) and reconnected them, in case a bad connection was at fault. Only, really stupidly, I forgot to disconnect the PSU before I unplugged/replugged the ATX power supply. That, possibly through ESD, screwed up my bios settings, and maybe other stuff, I'm not sure.

Anyway, the strange thing is, in the bios hardware monitor, after I did all that, things were looking better. The 3.3 went from 3.08 to 3.28, 5 went from 4.8 to 4.98, and 12v went from 11.9 to 12.03 or so.

Strangely, though, when I look at the voltages from MBM, my 3.3 is 3.12-3.2, my 5 is about 4.85, and my 12 is 12.55. So, in the bios hardware monitor, the voltages are nearly perfect. In MBM, they're still off - although the 12v is actually above spec, rather than below.

MBM and the bios hardware monitor both just read values from the sensor chip, right? Why would they basically agree before, but disagree now?

Also, strangely, my CPU is a 1.4v chip. MBM used to read 2.5v for "core 0" - I figured it was just misreading my memory as my CPU. The bios screen gave the proper voltage for the cpu. But now, suddenly, MBM is reading 1.4v for "core 0" - and I didn't change any settings. I don't know why the hell that change occured.

 

[Oklahoma Wolf]

If the Truepower is fairly old and well used, it could be running into the infamous bad capacitor issue - I'd open it up and inspect all caps for bulging and leaking. If you find some that are, you'll need to replace the PSU or replace the bad caps with good quality parts from Sanyo, Panasonic, Rubycon, Nichicon, or United Chemi-Con.

 

[RollingThunder]

If the Truepower is fairly old and well used, it could be running into the infamous bad capacitor issue - I'd open it up and inspect all caps for bulging and leaking. If you find some that are, you'll need to replace the PSU or replace the bad caps with good quality parts from Sanyo, Panasonic, Rubycon, Nichicon, or United Chemi-Con.

Wolf,

What are the tolerances for voltage output for most power supplys?

 

[Oklahoma Wolf]

5% on most, 3% on the Truepowers. However, when these caps go bad, they may be letting enough ripple through to exceed this. It is very difficult to verify with a voltmeter due to how rapid the fluctuation is - you really need a scope to tell. There have been cases with CWT units with bad caps when so much ripple got through the PSU went into protective shutdown.