Problem with Avid and new Computer
Hey guys, a few weeks ago I put together a new PC for video editing. Right away I began to notice problems with the system however, most notably the MACHINE CHECK EXCEPTION blue screens I get at least once a day. Not to mention the system freezing and restarting multiple times every day; most of the time when trying to play back media files in avid, windows media player, and even winamp. But the error has happened so frequently and in so many different applications, its hard to determine what exactly is causing it. But let me say, it happens most often, and most inconveniently while running Avid Xpress Pro HD. I will note that every time I try to digitize something for 7 minutes, the computer will restart. At school, that avid has digitized things ALOT longer than that without any problems.
Here's what I'm running: AMD ATHLON 4400 64X2 dual core WD SATA RAPTOR HD (74gb,10,000rpm) Radeon x800gt 256mb video card 1gb PNY pc3200 (400-mhz) Biostar NVIDIA GeForce 6100 mATX Motherboard -socket 939 Firewire card +60gb Firewire HD and 250gb Firewire HD 420 watt PSU Now, I read another thread on this website talking about the motherboard having to be a certain chipset for Avid to run correctly which caught my attention but strongly feel the problem is deeper than that. I also followed all the advice in another therad on trying to avoid the MACHINE CHECK EXCEPTION error (making sure enough voltage is getting to the motherboard, putting the system on a "cleaner" power strip, etc.), but to no avail. Now, I disconnected almost every piece of peripheral hardware, swapped the RAM from my other computer into this one, and the errors still happen. I have the ability to swap out my current video card for an older one just to make sure its not what's causing the problems, but I have a stronger hunch it's the motherboard. It is allegedly able to support the ATHLON 64 X2, but I don't know honestly, I've never heard of this company BIOSTAR and found very few opinions on their products. I'm even having doubts in myself purchasing the right memory; since the motherboard supports dual channel memory, does it need dual channel memory or not? I also read that AMD's are not supported by Avid, but then I saw alot of people say they have AMD's and Avid works great, so I don't know what to believe. Before I go and buy a "better" mobo or even swap everything for a XEON setup, I would like to get a second opinion from someone; also, please let me know if I've messed up or neglected anything when building this system. Thanks in advance, Mike ********* I forgot to mention, I installed XP x-64 without knowing Avid doesn't work on it, but the blue screen's still happened on that, so I rolled it back to XP PRO CORP, but the errors happen as frequently. |
Coupla things, firstly, do you have a larger power supply you can test? I'm not positive, but with a dual core processor, 10k rpm internal drive, x800 video card, and two external drives, 420 watts sounds a little low; I could be completely off base, but power requirements have changed since I last built a computer, and I could easily see that being the cause of the problems (I see you checked to make sure enough power's getting to the motherboard, but what about all the other devices?).
More importantly, have you written down the codes on the blue screens you're getting? The "0x000..." things? You should be able to look them up on Microsoft's website, there's usually some info on them in there. Same thing with the Event Viewer: take a look at it, see if there are any interesting details (Start->Run->"eventvwr.msc") under "Application" or "System". Might help point you in the right direction. You may also wish to peruse PC Perspective, an AMD focused enthusiast community, if you haven't already. They have excellent forums, one of which is dedicated to Biostar motherboards. |
Hey, thanks for responding.
Quote:
Quote:
I'm checking the eventvwr.msc right now like you said. here's what actually caused the system to freeze the last time. check it out: Event Type: Warning Event Source: Tcpip Event Category: None Event ID: 4226 Date: 11/26/2005 Time: 11:31:09 AM User: N/A Computer: M-ED38A79541BE4 Description: TCP/IP has reached the security limit imposed on the number of concurrent TCP connect attempts. For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp. Data: 0000: 00 00 00 00 01 00 54 00 ......T. 0008: 00 00 00 00 82 10 00 80 ....‚..€ 0010: 01 00 00 00 00 00 00 00 ........ 0018: 00 00 00 00 00 00 00 00 ........ 0020: 00 00 00 00 00 00 00 00 ........ I was actually in Firefox before as the system crashed... the next time it freezes, if its a TCPIP error again, I will try and install a different ethernet card and not use the onboard one. Quote:
Your advice definitely helped, as I forgot about checking the Event Viewer :) |
You're quite welcome; while you're at it, take a look at the Windows XP Professional Resource Kit. I don't know that there's any useful information in it for this particular situation, but there's all sorts of other stuff to browse. I'd never heard of it 'til a few months ago; it's like finding an extra, unopened birthday present for yourself.
Also, don't forget (I did, because I'm stupid and never make use of it) to check your Dr Watson log file, which contains a whole bunch of potentially useful data. It should be in "Documents and Settings/All Users/Application Data/Microsoft/Dr Watson/drwtsn32.log", just in case you can make some sense of it. I'm no programmer, but it seems like it could shed some light on the problem. And thanks for that blog link; quite an eye opener, I've got to pick up one of those Kill-A-Watt dealies. |
Michael, you can try running the following problems to nail down if it's a hardware problem. If you can get the monitoring utility working, then it'll show your power supply voltages and help you see if the power supply is good.
Prime95 Prime95 is a distributed computing program that tries to find certain prime numbers. Its 'torture test' stresses your computer with mathematical calculations and checks the output against known results. This makes prime95 a good diagnostic for instability problems from your CPU and your RAM. Download it from: http://www.mersenne.org/freesoft.htm Run the “torture test” through options --> torture test --> Ok Prime95 will stop immediately once it detects an error, and continue indefinitely if it does not. Prime95 will log errors in results.txt in the installation directory. The icon will also turn from red to yellow when it detects an error. If your computer fails Prime95, it is probably a CPU error although it could also be the RAM or the power supply. Use memtest86 (memtest86.org) to check if the problem is RAM-related. Use the following program(s) to help narrow things down further. Speedfan or other monitoring utility Speedfan does the same thing as Motherboard Monitor and your motherboard manufacturer's monitoring utility (not all motherboards will have one). These programs can take readings off the motherboard's sensors and tell you if the temperatures are too high or if the voltages are out of spec. You may also wish to run these programs as they are more descriptive. Your motherboard's monitoring utility is the best to use (try looking on your motherboard manufacturer's website). To get Speedfan, download it off of http://www.almico.com/sfdownload.php Look at the "Readings" tab in SpeedFan (this should be the first thing you see). Look at the bottom for the voltage readings. The reading for the +12V, +5V, and +3.3V voltages should be within 5% of 12, 5, and 3.3. Each monitoring utility calls these voltages a little differently, but they will be labelled with the numbers 12, 5, and 3.3. Faulty voltages indicates a power supply problem (or motherboard, if it has leaking capacitors). The temperatures should generally not exceed 70C. High temperatures indicate overheating components. --- |
Quote:
All of the voltages are indeed within 5% of the indicated value. EXCEPT for the -12V which is running at roughly -3.8V ... but I'm assuming it doesn't matter for this voltage. The temperature monitor indicated that a few different devices were running at the following temps ..... Temp1: 127C Temp2: 127C Temp3:33C ... HD:35C ... TEMP1 (APCI) (PCI EXPRESS CARD I'm guessing):40C. Now at first I was like WOAH, 127C is a number which is critical, but the device on TEMP1, 2, and 3 which I think is the CPU all share the same address, so I'm just negating the first two and assuming temperatures are fine. Next, I will run the stress test ASAP. By the way, do you think not using dual channel memory could cause a problem? |
Some memory controllers can run two identical sticks of RAM in dual channel mode. The sticks of RAM do not specifically have to be dual channel, although they do have to be similar enough (mismatched sticks of RAM can't be run dual channel). You don't need to buy RAM specifically marketed as dual channel.
Those sticks of RAM just mean that the pair has been tested as a pair, as opposed to individual sticks. Running RAM in dual channel means tighter tolerances, so theoretically it's possible to get sticks that run fine individually but not dual channel. Prime95 is a good test for your memory, but you may need to wait overnight. memtest86 and microsoft's memory testing programs are both good RAM diagnostics, but in rare cases they don't always catch everything (same with prime95). Quote:
2- If you are getting the same or similar BSOD codes every time, run them through google. It may be software or drivers that are causing your problem. |
Quote:
Application exception occurred: App: C:\WINDOWS\system32\Ati2evxx.exe (pid=908) When: 11/14/2005 @ 19:33:33.312 Exception number: c0000005 (access violation) ...which is the ATI program that runs at startup. Also, I just got a BSOD and I wrote down the code: 0x0000009c (0x00000004, 0x8054d5f0, 0xb2000000, 0x00070f0f). I checked microsofts site and got this troubleshooting page, which I've seen before, but is always worth a read. Tomorrow I'm going to head to a computer show and try and obtain a few things: A more powerful PSU to run a test, and a "Kill-A-Watt type" meter Incidentally, the BSOD said to "turn off memory caching" or "turn off CPU caching" and one of the diagnostic programs said caching was on, but I looked through my BIOS settings 100 times and never saw such an option. Do you think that could be it? |
Interesting...I'll ask the stupid questions, have you got the latest ATI driver? If so, have you tried a previous version? And do you have the Molex power connector plugged into the video card?
I've heard the same thing about caching, but off the top of my head I'm afraid I don't know where that setting would be. |
Quote:
Quote:
Here is the exact report: Code:
*** ERROR: Symbol file could not be found. Defaulted to export symbols for C:\WINDOWS\system32\ADVAPI32.dll - Quote:
And this model ATI doesn't have a power connector, I think it's just powered through the slot on the mainboard. ------------------------------------ Perhaps another silly question for both you guys, would a motherboard with a GeForce chipset have conflicts with an ATI graphics card such as mine??? that shouldn't matter right? |
Do you by chance have Samsung memory chips installed? According to that link you posted, you're using the GeForce 6100-m9 motherboard; I checked Biostar's website, and there is an updated BIOS available, if you feel comfortable flashing it. However, they say it fixes a "minor" compatibility issue with Samsung chips, and this issue doesn't seem very minor to me...still worth a try, though, if you're up for that kind of thing. Worse comes to worst, you could always fill out their support form.
One more thing regarding your drivers, are you using the version with the ATI Control Center or the Catalyst Control Center? And have you tried the opposite? Oh, and I see after a little more research that this board has integrated graphics capabilities (the GeForce 6100 chipset, from NVidia). Is this disabled? Have you tried using it? I know the problem has been fairly well narrowed down to the ATI card, I'm just throwing out ideas right now. Finally, if all else fails, try simply disabling the service in question: Start->Run->"services.msc", find "ATI Hotkey Poller", double-click, and set its "Startup Type" to Disabled. Restart the system, see what happens. This may actually be the easiest method, if you can live without the service. Most people can, from what I understand. |
Quote:
Quote:
I will try to install Catalyst again to see what happens. Quote:
Incidentally, it's interesting to point out that the CD had an install for a certain "AMD 64-bit DRIVER" which I installed every time I formatted and reinstalled windows. Have you ever heard of such a thing and is it necessary? There was no real explanation in the documentation that came with the mobo for it..... I'm going to disable the ATI Hotkey Poller, and see what happens. thanks again for your advice :) |
Not sure about that driver...could be needed to use a 64 bit processor in a 32 bit OS, or maybe only necessary if you have the 64 bit version of XP. I have no experience with that (not yet, anyway; sooner or later I'm gonna get me some new 'puter parts).
I'd say your best bet right now would be disabling the service. If that doesn't work, then you can take things further, but if it does, it could save a whole lot of headaches, you know? Here's hoping all goes well! |
Quote:
Getting the same BSOD code seems to indicate that you might have some sort of software-related problem. |
The article about true power consumption was nice, but looked to me as a wishfull thinking. I bet a lot of manufacturers lable thier PSUs for higher rating than they could safely sustain over prolonged periods of time.
First figure that PSU has only about 70% efficiency so whatever the total consumption of your parts inside would need at least 30% larger PSU. That would put PSU at 100% load wich is very bad. It's like running a car at top RPM all the time. So I would get a 30-40% larger PSU on the top of everything. Also what happens at peack levels the quality of current from PSU will deteriorate - you will get ripples instead of a straight line if you look at it with oscilloscope. Also as it was said you need to take in consideration multiple outputs on the PSU (+5, +12...) if any of them are getting close to maxing out you will have problems. You would have to get all those numbers for you MB, CPU, HDD and whatever you have in there add it up and check against that of a PSU for each of the outputs. Video card could be a major consumer of power. For example XFX 7800 GTX will require a minimum of 500W PSU with at least 26A on a +12 bus - that's what manufacturer recomends. I would imagine x800gt is pretty power hungry too and the fact that it doesn't have a separate power connector doesn't help. Your MB may not be able to supply that current to it. Good luck. I would get a bigger case than mATX, or the one that you can fit a regular sized PSU. Though the PSU may not be a problem or not the only problem in your case. |
A hot-rod computer with a high-end video card will draw about 150W maximum, at load. silentpcreview.com has some figures.
In any case, if you can measure the power supply voltages when the computer is under load you can check if the PSU is good enough. One way is to manually measure power supply voltages with a voltmeter (this is somewhat tricky). The monitoring utilities are better if you can get one to work. An alternative approach is to stress test the computer with prime95. |
Quote:
|
Hi Alexey,
You can run two instances of Prime95 to utilize both cores. See http://www.dfi-street.com/forum/showthread.php?t=16446 |
You gave us detailed information concerning your CPU, MOBO, etc, but you didn't specify your power supply's brand nor model number! My guess is: the problem is not not having a high enough wattage rating, but rather that you don't have a high quality power supply.
|
Quote:
I thank you though cause it made using this computer ALOT more bearable, I can actually work on things long enough to actually get them done... But I am at a loss, cause like I said, the same crashes still happen from time to time... I might try to format again to see if it was something I might have installed that's causing the crashes... Quote:
would you or anyone have any suggestions on a good card for video editing? Considering I'm having tons of problems with ATI, maybe something from someone else ... I would say price is no object, but I'd feel real crappy if I buy a great new card and the problems persist... Quote:
Quote:
|
Michael,
If your computer passes Prime95 for 10 hours then the CPU, power supply, and motherboard are fine. The RAM is very likely fine, although you can double-check with memtest86 if you want (I wouldn't bother yet). It seems like the problem is related to the video card... you could also try downloading the latest ATI drivers and trying those. Download the drivers, uninstall the current ones, REBOOT, and then install the latest drivers, reboot. In rare cases you might also get bizarre behaviour if the video card is not seated properly. You can try removing and re-installing the video card. Watch out for static and ground yourself first of course. |
Thanks Glenn,
I did exactly that, except first, I formatted my drive and started with a completely fresh system... downloaded/installed the latest ATI drivers... the works... now, Avid did freeze twice, so I disabled the ATIEXX service like before and the crashes stopped altogether, used it for a day straight, no crashes, no BSODs... *KNOCK ON WOOD* BUT I do plan on obtaining another video card (non-ati) within a week, just to be sure. I'll keep you guys posted if I have more problems. thanks again!! |
All times are GMT -6. The time now is 12:34 AM. |
DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network