This is kind of a unique test we have worked on for a while.  One of the more important qualities of a power supply is to make sure it does not produce harmful interference to the components.  Perhaps the most sensitive of these components is the memory (video card, RAM, and CPU cache), which is just millions of little transistors locked in an on or off position, held in place only by a delicate balance of electricity. 

The driving theory behind ECC memory is that it corrects errors that occur in memory.  Perhaps a transistor is faulty and flipped for the wrong reason, or a faint electrical signal pulled a transistor into the incorrect position.  While researching this review and others, we have noticed the largest factor for incorrect memory blocks is faulty power supplies. 

How it works:

We started off by modifying MemTest86.  We bumped the delay between tests up to several hours, rather than seconds.  MemTests86 works by writing various patterns into the memory.  After our programmed delay of six hours, the program goes back over the memory and analyzes how many of those bits were flipped out of the original pattern.  Both the memory and motherboard used passed the unmodified MemTest86 without any problems.  

Click to Enlarge
Click to Enlarge

Not only is this a good measure of shielding on the PSU, but also a good measure of shielding on the motherboard and memory.  However, we used the same motherboard and memory for the entire test.  This way we assure that the small amount of interference coming off the motherboard is consistent with each power supply.  Unfortunately, it took 6 hours to run this test and we ran it three times for each supply to assure accuracy.  On the next page is a table showing the trial data.

Noise Testing Interference Results
Comments Locked

70 Comments

View All Comments

  • Anonymous User - Monday, August 18, 2003 - link

    I cast another vote for us to be able to get a copy of the modified version of memtest86. I'd love to check to find out how much bit-flipping is happening over time on my various PCs. In addition, it seems to me that it would be a good way to see if ECC memory is actually doing what it should be. (If a bit gets flipped on a board with ECC memory that's supposed to support ECC memory then there's obviously something wrong).

    Also, you should give a copy of your modified source to the memtest people so that they might include the long delay time as an option in a new version.
  • Anonymous User - Sunday, August 17, 2003 - link

    Any chance of a review of the silentmaxx fanless 350W. This thing has no fans so in theory it should be 0Dba! Not sure where the poewsupply is up to the job though on the poewer front - a review owuld be great. Cost as you probably guess is on the high side...

    http://www.silentmaxx.net/silent_products/power_su...
  • Anonymous User - Wednesday, August 13, 2003 - link

    To me the most interesting data from this roundup was the instability a PS can cause to a system. I think this subject is worth a dedicated article. Also how can we reproduce this data at home? Where can we get the modified memtest86?
  • Anonymous User - Wednesday, August 13, 2003 - link

    Hello!!??? Seasonic power supplies?? These have to be the quiestest power suplies I have ever (NOT) heard. Appears to be pretty well constructed. These should realy be included in any decent, comprehensive power supply round-up.
  • Anonymous User - Monday, August 11, 2003 - link

    They didn't look at the seasonic brand. recommended here

    http://www.silentpcreview.com/modules.php?op=modlo...
  • Anonymous User - Monday, August 11, 2003 - link

    I got a PC P&C Silencer 300 a while back, and I was very unimpressed with its noise level. It was hardly quieter than the cheapo PSU it replaced. My Enermax 365 and Antec Truepower 350 are much better.

    How about reviewing Seasonic? I hear they're super quiet. A little hard to find, though...
  • Anonymous User - Monday, August 11, 2003 - link

    Untill all power supply manufacturers get it together and sheath their cables, we are pretty much stuck with what they offer. I solved this in my window case with some electrical conduit from the auto parts store. there are a few color choices including your basic black, but any of them make a world of difference hiding those unsightly P.S. cables. That and a little electrical tape over the white connector and they almost disappear.
  • Anonymous User - Monday, August 11, 2003 - link

    Another extremely happy user with a PCP&C Silencer power supply. I do have to question just a bit why the reviewer didn't find out about their existence on his own, noise being the primary complaint in his review (though I imagine the sheer number of power supplies being reviewed and perhaps deadline pressure could have been factors).
  • Anonymous User - Sunday, August 10, 2003 - link

    This is actually the third power supply review on Anandtech. Not the second as you wrote.

  • Anonymous User - Thursday, August 7, 2003 - link

    #16, please check out http://www.wikipedia.org/wiki/CPU_power_consumptio... for information on power consumption of several common CPUs (especially AMD).

Log in

Don't have an account? Sign up now