Help! Stable DAW now freezing

M

mikesong

New member
Hi guys, I really need some help here...Does anyone know what causes a Win2K system to just freeze up all the time? Read my scenario for the possible causes and the possible causes I've already eliminated.

I had a rock-solid DAW that I built myself and worked with for 6 months without any problems. I then packed it up and drove it across country, and since I got back to New York it's been nothing but problems. The first thing I did after I unpacked it was take off the heatsink and install a larger and quieter fan on it. However, when I booted up the computer I got a blank screen...I took it into a repair shop and they said the CPU was dead, despite not having any visible signs of damage on it. I'm not sure if I caused this with the heatsink/fan replacement or if it was caused by being bumped on the road although I did pack it fairly well. In either case, I had the store put in a new CPU as well as put the heatsink on it to make sure it was done right. Voila, now the computer would work, however it freezes all the time, no matter what program I'm in. Even if I'm not in a program, just browsing through a folder it will sometimes do it.

I did install the latest drivers for my Delta, Midisport, etc. as soon as I got the computer working again, but I doubt those could have caused these problems. I've reformatted my OS hard drive and reinstalled Windows because I thought it might be a software/driver issue, but even with no software on my computer but Windows it still freezes. I've removed ACPI and done all the tweaks I could think of but none worked, it really seems like a hardware problem. It does not actually crash (no blue screen, maybe that's because of Win2K) but just freezes there - no mouse movement, etc. Could this be caused by the new CPU? Maybe the video card or mobo was also damaged in transit? My question is what piece of hardware could cause a computer to work, but then freeze up after a few minutes. I have noticed that certain things such as playing back Cubase projects or working in Wavelab would cause it to freeze a little quicker than if I'm just browsing through a help file or something, however even help files will make it freeze. Anyway, here's my setup, if anyone even has any guesses that would be appreciated:

Asus A7M266 mobo, AMD XP 1900, Windows 2000, 512 MB Micron DDR RAM, 2 40 gig 7200 RPM Maxtor HD's, Maxtor G450 vid, Plextor CDRW, Delta 66/Omni Studio, Midisport 2x2, Enermax 350W PSU.

By the way, the new heatsink/fan is cooling adequately (highest temps go to 59 C, however my mobo supposedly reads 10 C too high) so I don't think that is causing the problem, in case any astute readers might have suggested that. Thanks again for any help!

Michael
 
It is hard for me to believe that the CPU was defected. I think that the cause for your first problem (blank screen) was your display card actually. It happens a lot when you move your PC as the display card might move in the AGP slot and this cause the system not to boot.

For the freezing problem -
I assume that the new CPU (if really was replaced - did you check that it was a new CPU?) may not be configued well into the mobo - that is to say the mobo was not jumpered properly to fit your CPU specification. This could be also an improper type of memory that casues system unstablilty, or simply a defected display card (you can start by just replacing the display card and see if this happens again). CPU overheating may cause the same problem. Is the fan you installed strong enough for the CPU ? if not, THIS can cause your CPU to go defected.


Take it back to the store so they check and fix it up !
This is not a software problem so you're just wasting your time formatting your drive again and again !!
 
matrox tweaks

Sweet! I fixed it, I think. I've actually had it running for more than an hour with no freezes, we'll see if it keeps running fine. I'll have to record a couple-dozen tracks over several hours, see if it can still stay stable like it used to. What I did was make a few changes in the BIOS, which had been updated to the 1007 BIOS along with the new CPU, and I forgot to make all the necessary tweaks to the new BIOS. I disabled the following options (don't quote me on exact wordings):

AGP Fast Write
Video ROM Shadow Memory

I HIGHLY RECOMMEND disabling these options to any Matrox owners (I have a G450 but I think it applies to all their cards since according to the forum at their website, these options aren't supported by their cards. There are other possible tweaks if you're having problems like lowering the AGP Rate from 4x to 1x and running a hardware diagnostic program but I haven't needed to do those yet because it's finally stable again, at least so far. If any Matrox owners haven't checked these options out in your BIOS I encourage you to do so.

Thanks a lot for your post gilwe, you definitely led me in the right direction. Making sure the card was fully plugged in was just about the first thing I had checked, so I knew it wasn't that...and I'm pretty sure the old CPU really was fried because I watched the guy at the store fail to get a screen like I had with it and then replace it, causing it to work. I do still have the old CPU, I guess I could check it myself or bring it to a different store to check it, but I believe them that it was fried. And yes, the new one was actually new, and came with the 3-year retail warranty and all, since I learned my lesson having gotten screwed by an OEM warranty on the last one. But you focused me in exactly the right direction in terms of looking into how the new CPU had been config with the MOBO, and also that this problem was logically being caused by something to do with the display card. Anyway, thanks...
 
You're welcome. I still suspect that you old CPU is OK. They don't just get fried... If it worked OK before you moved the PC, it should have still worked fine. Sometimes when you move the PC from one place to another, parts can move in their slots, causing the system not to boot up. It can be a display card, memory and also CPU. You generally solve it just by unplugging these parts and putting them in back again.
 
Arghh!! Well, it turns out once I worked it a little harder that my system isn't stable at all. I went ahead with a bunch more video card tweaks which I got off the Matrox tech support forums, such as hardware diagnostic tools, driver reinstillations, etc. and I'm pretty sure now that my freezes aren't being caused by the display card. The one modification I haven't been able to make because the card won't let me is reverting to AGP 1x mode, not sure if this would help me.

However, now I think the CPU is the culprit. The one thing that does seem to help is running the bus at 100 MHz instead of 133 which makes my 1.6 gig Athlon a 1.2 GHz. This seems to make the computer much more stable but I have to do more testing. Does this mean the problem is definitely related to my CPU not being able to run well at full speed? Or could it be something else in my system not being able to run at that bus speed? At this point I'm inclined to think it's the new cpu that the store put in there, so I'm going to bring it back to them and have them put in a different one, see if that works. The only other thing I could think of is that the new heatsink/fan isn't cooling adequately, causing lockups...my highest temps are just under 60C, can any other users of ASUS A7M266 boards which supposedly read high tell me if this is too high a reading for an AMD XP 1900?

Oh, and gilwe, the last problem really was the CPU. As I said I watched them replace it myself, which then caused it to work. The display card had not been touched. And as I already said, checking to make sure all cards were plugged in was just about the first thing I had done myself so it was definitely not caused by that, although that was a good guess since I had just moved the computer.
 
"The first thing I did after I unpacked it was take off the heatsink and install a larger and quieter fan on it"

There's a chance that the new fan you put on is not strong or comatible with the CPU and this has caused it to get fried ... CPUs get fried when the overheat. Improper jumpering can also cause it to get fried !

Anyway,

changing the bios setting should not solve your problem, you should use BIOS default for diagnosis.

Additionally, I need to know further specification such as for what speed the CPU is jumpered to and what type and speed of memory you use.

Generally, every part should be run in it's natural speed. When you overclock or downclock you may have any type of problems.

Except that, freezing has been always related to a defected display card (change the display card and see if the problem is gone) or CPU overheating or defect that is generally caused by an improper setting or cooling. It also can happen if you don't jumper the board well or when parts doesn't match well or parts' speed are not right configured.


The link below will be usefull for you:

http://www.hardwarecentral.com
 
I've removed the new fan and reverted to my old noisy Thermaltake Volcano 6Cu+. The temps have dropped about 10 degrees, to about 50 Celsius. This is also an AMD approved fan. Yet this has had NO effect on my freezing problem, so cooling is probably not the issue. As for the other stuff you asked, I've returned my FSB to 133 MHz, the speed it's supposed to run at, which gives a processor speed of 1.6 GHz on my Athlon XP 1900+. Still freezing at that speed. The reason I made the BIOS changes is because my MOBO is config'd for Jumper-free speed mode, controlled by the BIOS which is what worked for me before. I did find that the voltage jumpers on my BIOS were one higher than the default setting, so I returned them to default and it's still freezing. My memory is PC 2100 DDR RAM from Micron/Crucial (can't remember which one, oh and there's 512MB if that matters.) I've returned the AGP bus to 4x, so now everything is running at its normal speed like you said. I have a feeling if I could get the video card to run at AGP 1x and slow my CPU bus down to 100 MHz that might solve the problem, but I shouldn't have to do that to get a computer to run stable, should I? As I've gone through on these previous posts, the video card has passed the hardware diagnostic test, and I have tried almost every tweak suggested by Matrox so I am leaning towards it not being the cause of the freezes, particularly since it wasn't one of the components changed. Right now it's back to the store to have them swap out parts to see if they can identify the problem hardware...I'll let you know if they solve anything. I'll also get that old CPU tested again to check your hypothesis.
 
Late today the store said it was ready, and I'm picking it up tomorrow to see if it really is stable. They claim that by doing a "Windows Update" from Microsoft's website before installing the video drivers, they've gotten rid of the freezes. That was one thing I wasn't able to try since I don't have any internet stuff on my computer, but I had been under the impression that Windows 2000 Service Pack 2 didn't need any updates...it certainly didn't in order to work for me before this problem. We'll see how real their fix is, since I've become convinced it's a hardware, not software, issue. I guess what I really need now is a strategy for getting my computer back and testing it myself BEFORE I pay them, since I can't afford to pay any more money for half-solutions. I'll post tomorrow with the results, if there are any.
 
What they told you doesn't sound to me like the source of the problem either .... 100% of al PC freezing I cam across were
hardware problems. Let's see if they're right ...
 
As you said, they were wrong, and I got 90% of my money back but don't yet have a stable computer. As we both said, it's hardware, not software...they gave me back the computer supposedly working and it crashed within 15 minutes of turning it on. I then brought it back for a refund of all my money, because I had originally asked for each piece of hardware to be swapped out and checked, and they had wasted 2 day's time reformatting drives, reinstalling Windows, drivers, etc., which I knew wouldn't work since I had already tried it myself and I didn't want to pay for them to do it. I also made them take the new cpu back and refund my money, in case they gave me a bad one that was the weak link causing the freezing. However, I tested my original cpu and they were right about it being dead, so I let them keep the original diagnostic money I had paid them. I've since purchased an Athlon XP 2000+ at a local computer show for less than I paid for a 1900 from the store. As soon as I decide what heatsink to use and install it myself, I'll let you know what the story is.
 
"because I had originally asked for each piece of hardware to be swapped out and checked, and they had wasted 2 day's "

Did they do this?

It's troubleshooting 101 and the best way to narrow down the source of the problem. I doubt changing the processor again will solve your problems, but ya never know. Is it possible you stressed the mainboard switching fans around, scratched a trace, put too much pressure on it?

I'd like to know if they tried different memory. And if they did and it didn't correct the problem - did they try your chip on a different motherboard?
 
No, they didn't do anything to my hardware. Troubleshooting 101 and they failed, which is why they had no choice but to give me my money back. Another reason I gave them back the new CPU was because their unwillingless to simply swap out each device like I asked and their insistence that it was a software issue made me think they might have been trying to hide something.

That said, it is quite possible that I "stressed the mainboard switching fans around, scratched a trace, put too much pressure on it?" Since I've been concerned that cooling is my problem, I've been switching around a few heatsinks on a few chips and I've tried to be delicate (I don't think I put too much pressure on it) but my screwdriver did slip a few times while removing my tight volcano 6Cu+ and it landed on the mainboard. The mobo has a thin plastic sheet protecting some of that area and I don't see any visible scratches, but who knows what's going on microscopically? To make matters worse, I ran the computer for about 3 seconds (literally) with no heatsink on the cpu to test which way the hsf was blowing. I had read that less than 10 seconds of this wasn't going to cause any damage but maybe this was the cause of my cpu going dead. Probably more likely than the original reason I assumed which was that it had been damaged while the computer sat in its cardboard case in the trunk of my car while I drove home 3000 miles. But could this have damaged the mobo as well in such a way that the computer would work temporarily and then freeze. You would think all the errors I've described above would either knock out the mobo or not.
 
Mike,

Number one cause of freezing systems is CPU-cooling problems. Had one system in my office with this problem and have seen it happen with machines of two of my clients. Every time the main problem was freezing system performance (and a lot of blue screens) and everytime it got fixed with a simple new CPU cooler.

Peter.
 
Do you think it's possible mine could have been freezing due to cooling if the temps never went above 60C? I have a Volcano 6Cu+ which comes with a loud fan. I replaced this fan with an admittedly low-powered (but strong compared to a 60mm) 92mm fan attached with a 20 dollar aluminum adapter. This had my temps under 60 but produced freezing. I then switched back to stock volcano 6cu+ fan, 38 CFM, 39 dBA, which now kept the temperature under 50C, however even at that temp it was STILL FREEZING. Note two important things about these temperature readings: I have an Asus a7m266 board which is known to report cpu temps 10 degrees higher than non-Asus boards (the temps above have not been adjusted to take this into account), however, the temp readings come from the BIOS taken a few seconds after the freeze because I am forced to restart which may give it time to cool down a little, I don't know.
 
"The first thing I did after I unpacked it was take off the heatsink and install a larger and quieter fan on it. "

That bit of information, in your first post might mean something.

The problems didn't begin after the 3000 miles trip, they began after you switched the heatsink around?

Here is what I would do, your not afraid of messing around with hardware, which is good.

- Remove the motherboard from the case.

I know you built it, but maybe there is some random grounding issue. Screw rolled under the motherboard... heh ... Not likely but it's easiest to start from ground zero.

- Run the machine with just memory, video and CPU. Place the motherboard on something with a low static potential like cardboard, or better yet an anti-static mat.

If it still locks up, you've ruled out grounding problems, at least from the chassis standpoint.

Enermax are good P/S and the QA is great on them, so I doubt that is a variable - but I wouldn't rule it out.

I would do this experiment with your new processor. Although it's very unlikely the CPU is the problem. In my experience they either work, or don't.

Unlike motherboards and RAM which can work, or be flakey.

Without swapping components around it's all speculation. Lockups can be caused by almost any component. Hard Drives can cause lockups, memory, cpu's, improper cooling, a CD-ROM in it's dying days, a flakey motherboard...

Try the barebones setup, outside the case. You can rebuild it, faster... stronger...
 
I don't believe your problem is cooling related - provided you had a properly spec'd fan and HS, which you did. And provided that it was clipped onto the mainboard properly and using either thermal tape provided with the HS, or thermal compound that you applied, paper-thin - to the HS.
 
oh yeah, forgot to mention that my volcano 6 has some visible scratching, although not to the touch. i'm thinking of getting an ax-7 heatsink along with an enermax manually adjustable fan that tops out at 39 CFM, 32 dBA. Do you think this is worth a $40 upgrade from a 38 CFM, 39dBA Volcano 6Cu+? The Ax-7 also has an 80mm fan rather than a 60. In my situation I may have no choice.
 
Emeric, it was attached using poorly applied Arctic Silver 3. I say poorly applied because the coat was far from paper thin and I didn't even remove all of the old non-Arctic thermal compound. However, like you were saying, with those temps, could cooling really be the problem?
 
The XP has a maximum processor die rating of 90 Celcius. Of course you don't want to run it at 89 Degrees, but that gives some idea of what the window is - in the short term. Around 50C is typical for the XP's, your using the ASUS bios hardware monitoring, which is like you say - inaccurate. So it was more likely running around 50C.

Heat reduces the lifespan of processors. Measured in seconds - or decades.
 
You must log in or register to reply here.
Back
Top