Memory test, voltage

jrhii

Distinguished
May 8, 2010
248
0
18,710
I have been having a crashing problem with my computer, and it got worse a couple of days ago. It freezes at times, even on boot. I managed to slightly fix the problem by raising ram voltage from 1.662 to 1.693 (one step on my Mobo,) but it still freezes at times. I was getting just freezes, then BSOD sometimes. Not after the change it does the sound loop freeze.

It shows a error on memtest86, so something is wrong. My primary question is, would lack of voltage cause an issue in memtest86? Does this sudden change in voltage need mean that something is going bad?

I know I still have a bit of testing to do, taking out individual RAM and testing on card at a time to see if I can narrow out a problem.

I have 8 gigs of ram, the 4 from the corsair in my sig, plus 4 mushkin i put in a year ago.
 

bryanl

Distinguished
Aug 31, 2009
236
1
18,715
An increase of .031V is less than a 2% difference, most likley beyond the accuracy of the monitoring hardware of the motherboard and could be due to a change in room temperature, as can the crashing problem.

Something is wrong with memory that needs even 1.662V to operate at official specifications, even if that's the factory voltage rating.
 

jrhii

Distinguished
May 8, 2010
248
0
18,710
I don't see how the ram having to be at spec'd voltage (1.65) to run at spec (1600 mhz, 7-8-7) is an issue.

http://www.newegg.com/Product/Product.aspx?Item=N82E16820145285

I can say with certainty that upping the voltage by that amount reduced the freezing problems, though they are still occasionally happening. The issue is is that they started about 2 days ago with no change in the settings, so does this new need for a higher voltage indicate that ram (or something else is going bad?
 

bryanl

Distinguished
Aug 31, 2009
236
1
18,715

In your case it is failing at spec, at a much higher rate than any DRAM chip manufacturer would consider acceptable.