jonnyGURU Forums
Home Site Search Reviews Articles Contest Links PSU FAQs  


Go Back   jonnyGURU Forums > Computer Hardware > PC Power Supply Discussion

PC Power Supply Discussion Troubleshooting and discussion of computer power supplies

Reply
 
Thread Tools Display Modes
  #1  
Old 06-17-2012
siana's Avatar
siana siana is offline
micro ATX User
 
Join Date: Aug 2011
Posts: 13
Thanks: 0
Thanked 0 Times in 0 Posts
Default Slow start-up, ground noise - signs of trouble?

Helloes,

last time i posted, it was back in autumn i had a very frugal computer and FSP350 PSU and was thinking of a major upgrade. Then in middle of January, a surprise GTX295 came in my mail, sent to me by a fan for free, so i rushed out to get a PSU which would be able to feed a 300W+ dual-board GPU. Those are slow weeks when people don't buy anything, so the shops weren't stocked very well. The cheapest thing which didn't seem like total garbage was bequiet! L7 630W, so this is what i bought. The device is ODM by HEC-Compucase.

First day or so, i noticed the PSU took quite a bit of time, on the order of magnitude of 3 seconds between pressing the button and the fans spinning up, light going on. Decided to ignore this for a while, until a few months later i read pwr_good signal MUST come at most half a second after ps_on is activated. So i tried to reproduce the issue, and surprise, now the PSU has been turning on quickly, within less than a third of a second. Wonder whether it not doing this when it was really new means anything, and whether there is a plausible mechanism due to which this behaviour and its change can be explained.

A few days later, i tried to listen to some music, and noticed noise from the speakers, quite annoying in fact. I plugged in headphones into PC instead of amplifier, the sound was clean. I went ahead and opened up the audio amplifier, which is a miniature Chinese off-brand one, Dynavox. It turned out to be based on TDA2004 IC, with a somewhat... questionable construction. All INPUT- of the IC and the RCA connector sleeve are connected straight to internal GND, and GND was connected to protective earth through the heatsink tab of the IC connected to earthed enclosure, and only on that point. So i went ahead and slid a piece of thermal interface film between heatsink and enclosure, and replaced screws with which heatsink was connected to enclosure with Nylon ones. Effectively, i disconnected GND from earth and floated the amplifier on the input which comes from the computer. Not saying it was an AWESOME idea, and yet, the amplifier is noise-free since. I'm not super conserned with my safety, because amp enclosure is metallic and it's securely connected to protective earth. Well, with my old computer configuration, the noise used to be barely noticeable, but with new, it was really bad. The computer and the amplifier are connected into the same wall outlet, and not connected to any devices connected to any other. The said wall outlet is in an expensive apartment house from late 1960ies in southern Germany, i'm quite sure the earth is properly connected - electric safety is taken very seriously here.

I wasn't very sure whether floating the amp on the PC would have any safety implications for hardware, so i tried to undo this mod, connected the GND back to earth, and instead inserted a resistor of a couple hundred Ohm into the GND coming from computer into the amplifier, and i wasn't happy with the result, the buzzing was still there, though there was less of it but not enough, so i undid that and floated the amp again. Eventually i will want to redesign the whole amplifier with it being properly grounded and with proper differential input.

So i'm really wondering, first, did i do anything horribly horribly wrong, and second, what's a good way to measure common mode, ground noise coming from the computer, third, whether it means something is wrong with the PSU, and fourth, what kind of reasonably safe mods can i do - which don't involve opening up the PSU - to mitigate the issue?

The reason i am asking around now, is that yesterday i ran into this review, sorry in German:
http://www.planet3dnow.de/vbulletin/...399965&garpg=4
Which descibes a similar HEC-built PSU where thermal interface film between secondary-circuit rectifier diodes and earthed heatsink has been applied somewhat carelessly, which opens a chance that noise current is injected, if i understood this correctly at all.

Common components:
ASRock N68 C-S UCC, based on NForce 630a
2 SATA hard disks
1 SATA optical disk drive
Dynavox miniature "Hi-Fi" amplifier, TDA2004 based

Old PC configuration:
Athlon II x2 3GHz, <65W
Gainward GeForce GT220, <35W

New PC configuration:
Phenom II x6 2.8 GHz+Turbo, 95W
MSI GeForce GTX295 (old version), 300W

Last edited by siana; 06-17-2012 at 03:43 PM.
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 10:32 AM.


Powered by vBulletin® Version 3.8.1
Copyright ©2000 - 2013, Jelsoft Enterprises Ltd.