System Stability Issue (PSU Related?)

[etruscan]

Over the weekend and through monday I had repeated hard freezes (both requiring hard resets and unfreezing after 10-30 seconds) while playing World of Warcraft.

I tried to duplicate the freezes with 3dMark05, and Prime 95, both running separately and simultaneously. I had no luck

I called the BFG tech support line, and the tech asked me to install PCWizard, and check PSU voltages. 3.10, 4.85 and 11.80 were reported with the same reported in BIOS. The tech told me it was my Power Supply. I went and grabbed a multimeter and plugged it in the next day . I've had it plugged in, (and had no crashes) for about 2 days. It's plugged into my 12v rail. Now, it's reporting 12.13-12.23 volts, however, PCWizard is also reporting 12.20 volts. The 5v rail is 5.12, and the software reports 4.85 still.

Are software monitoring programs relatively accurate (i.e. do they report correct values when calibrated to past values)? Because if my software monitoring program is reporting .35 volts higher now, that would appear to be the source of my problem. If software monitoring is just worthless, then I'm confused and this should perhaps not go in Power Supplies.

System specs:
AMD 3500+ socket 939
MSI K8N Neo2 Platinum
BFG Geforce 6800 GT
2 GB memory (1 GB OCZ fancy-ram, 1GB value-ram from kingston)
2 Hard Drives, both 7200 RPM
1 DVD/CD burner
1 DVD Player
Vantec random-cooler on the CPU
Zalman 7700 cooler on the GPU

 

[davidhammock200]

(1) What PSU?

(2) mobo sensors, BIOS & software are NOT very useful for reading voltages, trust your DMM.

 

[etruscan]

1) Yeah, way to go me:
Tagan Noise-free TG-480-U22
http://tinyurl.com/7bmud

2) I know they're not very useful, but if I see a .4 v jump in reported voltage along with increased system stability is there any information there for me? I didn't have my DMM hooked up when the software monitoring was reporting low 12v voltage.

 

[davidhammock200]

1) Yeah, way to go me:
Tagan Noise-free TG-480-U22
http://tinyurl.com/7bmud

2) I know they're not very useful, but if I see a .4 v jump in reported voltage along with increased system stability is there any information there for me? I didn't have my DMM hooked up when the software monitoring was reporting low 12v voltage.

That PSU should be fine.
You could try switching between single & dual +12V rails, however either should work.

 

[etruscan]

So just to really belabor a point. You would say that the software monitoring system's low voltage reported during instability and correct voltage reported during stability (DMM only connected during period of stability) is meaningless, and that I can discount my PSU as a probable source of instability?

Or are you just saying that my PSU should be adequate for my purposes, and that the Software monitoring is meaningless, but not that I should be able to discount my PSU as a source of instability?

 

[J-Mag]

The problem with software based probes is that the sensor they read from is not calibrated really, so your baseline will be off. The good thing is that they are useflu in detecting changes that occur, so if your 5v rail said 4.85 just ignore that fact and make sure that it remains stable under load. If the software reader dips then that is information that is relatively accurate and useful.

 

[davidhammock200]

So just to really belabor a point. You would say that the software monitoring system's low voltage reported during instability and correct voltage reported during stability (DMM only connected during period of stability) is meaningless, and that I can discount my PSU as a probable source of instability?

Or are you just saying that my PSU should be adequate for my purposes, and that the Software monitoring is meaningless, but not that I should be able to discount my PSU as a source of instability?

Your PSU should be fine, ignor the software & test with DMM during problems.

 

[etruscan]

Your PSU should be fine, ignor the software & test with DMM during problems.

So, I was curious if the Software reporting was really as bad as all that.

I ended up restarting my computer with the DMM plugged in about 15 times. Each time the software reported the same voltage on the 3.3 and 5.5v rails (though the voltage it reported was massively different from the voltage measured by the DMM). However, the 12v rail appears to choose a pseudo-random value as a base on every startup both with MSI core-center and PCWizard 2006.

The DMM has been steady between 12.13v and 12.23v during every boot, stress test/whatever. The pcWizard/Core Center monitoring sets to a value between 11.5 and 12.3v base, then shows no fluctuation (the DMM updates about 3x a second I think, maybe just two). It appears I can't use software monitoring for relative values, stability or anything. Ah well. Lesson learned.