System is totally powering itself down with no warning

kofrad

n00b
Joined
Jan 1, 2006
Messages
29
Over the last week or so I've noticed that my install of XP is starting to hate me. I produce music using MixMeister Fusion 7, which seems to be where I get the issue (mostly).

Essentially what happens is I will be working on some music and within a few minutes, sometimes even less, the entire system just shuts off. Like the power was cut from the wall. No more lights, no noises or warnings just off. I then have to flip the switch on the back of the computer and wait for the power to drain before I can get it to turn back on. I have on several occasions experienced this issue in XP while not making music, so I don't think the problem is exclusive to MixMeister. I have yet to experience this issue in Ubuntu, only in XP.

My first guess was that the system was overheating, so I opened up the case and pulled out all the dust. Took the heatsink down from the CPU and re-applied some thermal paste, did the same to my northbridge and video card. All my temperatures dropped by at least 10 degrees. I re-sat all cables that I could think of, and the issue is still persisting.

Is there anyone who can help me? I am not sure what to do to remedy this problem. Any help is greatly appreciated, thank you.
 
Thats most non triumphant.
I hope this isnt bogus dude.
;)

What temps are you getting and what is your hardware, is it in your sig?
 
if it's not a CPU/heating issue then my next guess would be a power-supply issue.

Download memtest, make a CD, and test your ram http://www.memtest.org/ Probably isn't your ram but can't hurt to check.

Next, download Prime95 http://www.mersenne.org/freesoft/

Use this in conjunction with a temperature monitoring program. What are your temps under 100% load? Also, this will cause max power consumption from your CPU so if your powersupply can't handle it, this will also test that.
 
This definitely isn't bogus. I couldn't make this up if I tried. I haven't fully been able to check out my temps idling and under load as I don't have anything installed at the moment to help me out with that. The only temp checking I have done was under the BIOS or Ubuntu. Prior to dusting, I was able to get the CPU up to about 155F in Ubuntu. The power supply looked like it was holding out in the BIOS when I checked out the voltages. After dusting everything out, my idle temps in Ubuntu were about 130F with the case at about 95F. This may sound a bit hot, but it has been about the normal operating temperature of my system for the last three years. The ambient temperature in the room is usually about 80F.

The hardware in my sig is the same setup I have now with the exception of the power supply. Now I have a CoolMax 450W. I'll probably run memtest overnight tonight, I should have a copy on my Ubuntu Live cd. I'll be downloading prime95 and running that now.
 
Soz I wasnt being serious, I was ribbing you for using "totally" in your title.
Bill & Ted :)

The load temps would be mighty helpful in determining/eliminating where to look for the fault.
68C (155F) is getting near high but is ok if the CPU is fully maxed.
Sustained high temp is required to get a decent measurement, say 1hr+
 
Ah ok!

I am looking for a program to monitor temps/voltages in XP. If you can recommend one, that'd be great. Right now I'm getting prime95 and plan on letting that run for a few hours while monitoring the temps. I have also disabled the reboot on error, so if it's BSOD'ing, I'll get a chance to see that. I also checked the memory dump folder (C:\Windows\MiniDump) and only found one dump from about three months ago.
 
Alright, I've been running prime95 under the stress test with 2 worker threads for the last 20 minutes. I checked my idle temps at about 92F case and 113F cpu. So far the highest temps I have seen are about 97F case and 142F cpu. All disks are at about 80F. As far as the voltages go, I am not getting any readings for my 12V rail in this program, however they do appear in my BIOS. Right now on the +5V I have only gone up to +5.16V, the -5V reading has been stuck at -5.20V the entire time. VCore is at 1.3V, max being 1.4V. +3.3V is reading +3.3V with the max at +3.5V.

So far everything seems in line as far as I can tell. The sytem has been stable with no reboots or any signs of my issue. I'm about to go out to eat and I'll leave prime95 up. When I get back if everything is looking good I'll start checking into a problem with MixMeister.

I just have no idea how MixMeister could be causing the problem, but hopefully I can figure this out. Not being able to work on my music is driving me crazy :-/
 
Can you check your system and application event logs, they may show something leading up to the BSOD.

Everything looks good with temps.
If it is still rebooting, try blowing a housefan at the inside of the PC.
If something else is getting too hot that hasnt got a temp sensor this will work.

My feeling is that there is a problem with MixMeister or its supporting code.
Maybe a part of the program or a Windows dll is now corrupt, especially after a lot of crashes.
Try a program re-install.

The hardware you are using with the MixMeister will have a driver, perhaps this is the problem.
Look for a newer update or just re-install it.
 
I've already checked all the system logs with nothing even close to the times the error is happening. The only errors are TCP/IP or DNS related from months ago. Something is telling me that it is a problem with MixMeister. I do have the Numark MixMeister Control connected via USB, so I think I'll start by disconnecting that and clearing out all driver traces. From there I'll reinstall MixMeister and the hardware drivers.
 
Ok, let us know how you get on.

Something I just found that you might like.
The graphing tools in RivaTuner are excellent but havent been that good for showing CPU temps as the dll used isnt compatible.
You can use the dll from Realtemp which gives all you need to monitor the CPU on the graph.

They show for each core:
Load
Distance to TjMax (keep well above 20C)
Core Temp
Frequency

You can also monitor hard drive temps etc with Rivatuner, its a great tool.
If you need help setting it up, I will assist.
Get it here:
http://www.guru3d.com/index.php?page=rivatuner
The download link isnt highlighted, look for "Download latest public build of RivaTuner"
 
I have actually found a very good program for monitoring temps. It provides everything I needed http://www.hmonitor.net.

I have since reinstalled MixMeister along with drivers, been working on music with no problems. I'll keep this thread posted but so far I think we've figured it out. Seems like some sort of corruption or problem with the software. Thanks very much for all your help everyone! If I have any further issue I will update this thread. Once again thank you for your time & help.
 
Ugh.. I spoke to soon. No more than 10 minutes after posting, she went down in the exact same way as before. There was no error on screen and no memory dump left behind. This is starting to sound more like a power issue to me than anything else. However, I have never experienced this problem playing any games. Only in the MixMeister software, or sometime after closing it down. Somehow the uptime seems to shorten with each crash. Example, last night I was working on some music for maybe 15 minutes before the crash, reboot and go straight back into it, with maybe 5 minutes until the crash. The times kept getting smaller the more it crashed.

However, I just did about 20 minutes of work, then crashed. I rebooted and opened MixMeister, right now my work has been exporting for maybe 15 minutes with no crash. Almost seems that the crashing only occurs while I am actually working in MixMeister, or sometime thereafter.

So I've narrowed it down to the power supply or maybe even the sound card. Can anyone provide me with some guidance on testing these to determine if they are causing the problem?
 
Sounds like a heat issue.
It could be anything: inside the PSU, chipset, a faulty chip anywhere drawing a lot of current, cooling failure (a possibility).

Try blowing a housefan at the PC's insides and see if it improves any.
Also try pulling out any cards, clean the contacts and dust off the slots, same for the memory.

edit: forgot to ask, are you overclocked?

edit2: can you find the +12V voltage reported in the BIOS?
 
Last edited:
It sounds like heat to me, too...but it could be power related. Danny's right on. I have a power supply tester that checks each rail, and I couldn't live without it. It's found the answer to a lot of really odd problems.
 
Back
Top