loading

My computer locks up with an audio loop, help!

I recently upgraded my hardware (I put in two 8600 GT's in SLI, and an E6850, both used), and shortly after, my games began to lock up. Whatever was on the screen would stay there, and the last second of audio would loop. The only way to get out of it was a hard reset. Then I noticed when I unplugged my other two hard drives, games wouldn't freeze up anymore. So I replaced my 400 watt PSU with a 550 watt, with 50 amps on the 12 volt rails. Everything seemed to clear up, until today, when I was browsing Newegg and listening to Audioslave (not very resource intensive), and it locked up with the same symptoms for no apparent reason. I'm getting worried now, as I shelled out 75 bucks for a decent PSU and the same stuff is happening. My CPU idles in the upper 30's to lower 40's, and peaks at 60 C in games. I know it's a bit toasty, but it shouldn't be hot enough to cause this sort of problem, right? So if anyone has any ideas about what's going on, any help is appreciated.

sort by: active | newest | oldest
lemonie7 years ago
What made you think it was the power supply, and how did you decide on 550W? The 8600 can be quite hungry, so although you've added an extra 150W it's not all that much more. Are the cards fully compatible with the mainboard, and do they have good drivers? What happens if you turn down the settings so that they work less hard? L
RelaxedSoup (author)  lemonie7 years ago
I have three 7200 RPM SATA drives. When I unplugged all but my primary drive, games stopped crashing. My hypothesis was that my current PSU was barely under spec for my new hardware, and that the load of two extra drives was too much for it to handle during more intensive activities, such as gaming. It only have one 12 volt rail with 22 amps, and the recommended for SLI is 36 amps. I now have a PSU with two 12 volt rails and 50 amps (25 amps per rail), with all of my drives are plugged in, and I haven't had a game crash yet.

I decided on 550 watts because it was A.) Cheap, B.) Enough (as I understand, watts don't matter as much as amps) and C.) From a reliable manufacturer, OCZ (here's a link).

Yes, the cards are compatible (as far as I know), it's a P5N-E SLI board (link), I had a 8500 GT in before the dual 8600's and everything ran fine, the 8600's have only minor changes (32 stream processors instead of 16, slightly faster clocks). I flashed the latest BIOS on the board, and I just updated to the 186.18 drivers, although I see that nVidia has released yet another update since yesterday, I'll have to update once I'm done... I've also underclocked the cards by about 50 MHz to see if that would solve my overheating problems with CoD4, which it did. No crashes occurred when the underclock was in effect, but I didn't keep it there long.

However, I doubt that the problem here lies with my video cards, as I haven't had any problems with them since my new PSU arrived. I re-applied the thermal paste on one of them (it was getting a bit toasty) and they both idle in the low 50's and peak at 68 C. My hypotheses for what's going on are A.) My CPU is causing problems, it run a bit hotter than I'd like, and I got it used from a friend, B.) Something's gone wrong with my OS or possibly my drivers, or C.) Sleep mode causes some sort of problem with my hardware setup (more on that below). I have Windows 7 installed on a separate partition, so I could see if it's a problem with Vista, but I have no way to replicate the conditions that cause a crash, since the causes of this morning's crash are still a mystery to me.

I realize this reply is getting a bit lengthy, but I have just one more thing to cover. I ran a disk defrag on my C drive and a storage drive this morning, when it was finished, I put my computer into sleep mode. Upon resuming Windows, I loaded Windows Media Player and Firefox, and proceeded to browse the web. Then the crash with the looping audio occurred. So maybe the problem has something to do with sleep mode.

I apologize for any eyestrain you may have after reading all this, but the better informed one is, the easier it is for one to come up with a solution. Thanks.
the better informed one is, the easier it is for one to come up with a solution

Too true. It does sound like it might be a power problem, you've obviously done quite a bit of troubleshooting on this.

Things getting hot indicate high load, but keeping them cooler isn't going to change that. (Generally, too much thermal paste is used rather than too little, a small smear is enough)

Sleep mode starting or waking might cause a problem, but right out of it shouldn't.

The audio looping could be coming from the player, so I'm thinking: "why is the player behaving in that way?" Or is this the onboard sound (not too sure about that)?

There's nothing overclocked in this is there?

(still thinking)

L
RelaxedSoup (author)  lemonie7 years ago
No, although I'm planning on overclocking, once I get this all sorted out. Speaking of thermal paste, I did have some issues applying it. First, I put a thin line down the center of the die, as recommended by Arctic Silver, but it didn't really cut it, my temps hit 90 C before I noticed. Several re-applications later, I took a friend's advice and smeared a nice, thin, even layer over the entire surface of the die with a business card, which seemed to work quite well. I had good results with that method on my second graphics card as well. I'm about to buy this Zalman heatsink in a effort to drop my CPU temperatures, but I might not if you think a re-application of thermal paste might let me get by with the stock heatsink.

Also, when the crashes occur, it's as if the whole system is caught in a loop, and not just the media player. The mouse is not responsive, and Alt+F4 and Ctrl+Alt+Del have no effect. I'm running two plugins with WMP, a Windows Live Messenger Music Plugin, and this Lyrics Plugin, but I doubt either would cause a crash of this magnitude. Perhaps the onboard sound is causing problems, I'm using a splitter to power two sets of speakers... it might be an excuse to get a decent sound card :).
If you've spent some time on the heat paste it would sound like you've got the best out of what you have already, the issues having been sorted out already. Maybe it's something you can spot by keeping an eye on running processes (task manager). If you can find something happening in there you might get a clue? (still don't know) L
RelaxedSoup (author)  lemonie7 years ago
I'll try that, I'll also run my hardware monitor in the background with logging enabled, so if a crash does happen, I can go back and see what the temperatures/loads were, etc.