AG Knot crashing

[Cris001]

I've been using AG knot since it's been released, and never had any problems with it. Just upgraded it to the latest version and since then it started crashing my pc during the compressibility test.

I've tried to find a solution online, read loads of stuff about it, and it seems that the most commonly problems for that would be memory, HD space and CPU heating. But... why it never happened before then? I dont get it.

I tried getting back to newer versions, but now nothing seems to solve the problem.

Im using an AMD Athlon XP 2.600 - 1.92 GHz 500 MB RAM
120 GB HD. usually using 12GB free when ripping.

Can anyone help me please? What I dont understand is why it used to work fine and now it wont anymore.

Thanks a lot in advance!

:thanks:

[jggimi]

Hello, and welcome to the forum.

1) "crashing my pc"

is it the same "crash" every time, or different?

If its the exact same problem that occurs at the exact same point ... I would suspect software. If the "crash" occurs at different points in the encoding (different frame numbers for example), the I would suspsect hardware.

Windows NT-based operating systems (WNT, W2K, W2K3, WXP) have an Event Log that you should look through, also, in case the "crashes" produce log information (system or application or both).

What is the actual problem that occurs? "crash" doesn't mean much, without additional info.

2) "...why it never happened before then? I dont get it. "

If software, you may have installed something that doesn't interoperate well with the existing environment, or a configuration change occured -- with or without your knowledge -- which also may cause troubles.

If hardware ... and if heat is the problem, airflow is a likely culprit. Look for vents that may be blocked by dust or by things near them. Look for failed fans (non running or running at low rpm).

If non-heat related and hardware .... hardware fails.

Have you run any of the hardware stress tests you've seen discussed when searching? I recommend it, unless you're sure your problem is software.

[Cris001]

Hello jggimi :)

That was a quick reply. Thank you so much!

When I say 'crash' I meant that the pc reboots itself. It always happens at the same point. And it's always the same. It goes off. I've tried it several times. And I mean several. And it's always during the compressibility test.

The new version doesnt give me a log. It gives me the avisynth script.

As for heat, I've got 2 main fans, which are working properly. I also got my machine cleaned and left it open during the encoding just to check if it was heating, but that didnt seem to be the problem.

I've also unistalled everything. Divx, xvid, ac3 filter....all the codecs, and went for a new installation. Also uninstalled AG knot and got it installed again. I dont use codecs pack. I usually get them sepately as they are in the site (doom.org).

As for space and memory....Isnt 12 gig enough? To rip a dvd into a 700mb file? converting the audio from AC3 into mp3 128? And what about 512 MB RAM...Isnt that enough memory?

As I said, I've always used AG Knot...and never had any problems, so I dont know if there's any conflict with another software... or what on earth is happening.

I havent tried any hardware stress test though. Maybe I should do that?

Thanks again!

[jggimi]

Reboot tends to be hardware. That doesn't mean it can't be software, instead.

Take a look and see if ANYTHING is in your Event Log at the time of the reboot, as I mentioned above.

Yes, stress testing your hardware would be my recommendation.

In addition, most motherboards have sensors for fan speeds and temperature. I'd install one of the many speed/temp reporting tools that are readily available.

For more on confirming or eliminating hardware as the culprit, see http://forum.doom9.org/showthread.php?t=54130

[roflcoptr]

I've got a similar problem.

I recently reformatted my copmuter and reinstalled everything. I also decided that I would back up all of my DVDs as Xvid avi files and after some searching on videohelp.com, I found Auto GK. This worked flawlessly on ~40 jobs and I was very happy with the results. I would typically set up 4-5 jobs and select "shutdown when complete" and then walk away. I came home one day to find my computer shut off, which surprised me as I'd set up a bunch of jobs to run that morning and I thought that there was no way they'd all be finished by the time I got home. After booting up my computer, I found that one of the jobs I'd run hadn't completed properly and none of the other jobs had started at all.

Since then, every time I try to run Auto GK, it will get through the demuxing and indexing and the compressibility test, then as soon as the conversion thing starts, my whole computer shuts down. Not rebooting, just power off and doing nothing.

From what I've read on this forum and other sites, most people seem to suggest that it's a hardware problem. I'm a bit suss about this, because I run my computer a lot and don't seem to have these problems elsewhere. Also, my computer doesn't feel overly warm after shutdown and I was able to run this software just last week without any problems. FYI - the computer I'm running AutoGK on is offline, and I haven't installed any new software or upgrades, since I started to run AutoGK. But I'm deferring to the knowledge of the experts on this one and if it is hardware, then it's possibly one of the following three things:
- an insufficient power supply and I can't draw enough watts
- my cooling fan has started to die and it's overheating
- I've got a [Profanity-edited] bad motherboard.

I've downloaded the Ultimate Boot CD as recommended on the thread jggimi linked to above, but I'm a bit of a n00b and I'm not sure what I should run to check my hardware.

I'd really appreciate any coments or suggestions on what people might think the problem is. Also if someone could hold my hand a bit to help me figure out how to do a diagnostic check of my hardware, then that'd be great too.

Many thanks

[roflcoptr]

Oh yeah - and how would I go about finding an event log from when it shuts down?
Also, how do I access the BIOS to figure out speed and temp? Or is there a speed/temp monitoring tool that you would recommend?

[communist]

Try speedfan for temparature monitoring.

[jggimi]

IF the shutdown was managed by software, there will be an entry in your system event log, if your OS is WNT-based (NT, 2000, XP, etc.). W9X-based OSes don't have one.

You can find the Event Viewer in your Administrative Tools.

If the shutdown was managed by your hardware without informing your OS, there will not be an entry in your log.

[Cheech]

Turn off automatic restart: Right Click my computer > properties > advanced tab > start-up and recovery > uncheck automatically restart.

That should give you the blue screen of death at the point where it used to crash. I suspect a memory error so run Memtest86+ http://www.memtest86.com/ and find out for sure.

[roflcoptr]

Firstly, let me just say thanks heaps for the help so far.

I've installed speedfan and turned off automatic restart as suggested.

When I first turn my computer on and it's idle, the readings from speedfan are
temp1 33
temp2 77
temp3 25
HD0 33
Could anyone help me with what the different temps are?

Temp 1, 3 and HD0 don't change more than +/-1 at any stage, but Temp 2 increases a bit.
I start encoding and then while it's demuxing and indexing the temp goes to 86. Processing file and then decoding audio (temp goes to 96), normalizing audio, encoding audio (temp goes to 101), running compressibility test (103) then my computer switches off.

It didn't go to the blue screen of death, it just did the same thing it's always done and turned off and done nothing.

My computer powers off at the same point - just after the compressibility test - every time I run AutoGK. I've checked the system and application event logs and there's nothing in there. So I guess this may rule out the software (?), but I can't understand how it's a hardware problem when this seems like such a clear trigger.

I've got Memtest86+ and I ran it about a month ago for ~24 hours and it didn't find any errors. I'm going to try it again, but is there anything else I could try or are there any suggestion as to what may be going on here?

This problem is driving me nuts so thanks again for any advice or suggestions.

[jggimi]

It will depend upon your motherboard what each sensor is. But...it seems fairly obvious to me that Temp2 is your CPU temp.

I poked around at AMD's website, and it seems to me that Athlon XP maximum operating temperature is 85C or 90C, depending on the specific model.

You are exceeding this maximum temperature. In addition, the thermocouple sensor is mounted below the CPU, so the reported temperature is often several degrees lower than actual. Your hardware is shutting down in an attempt to protect your processor from physical damage.

I would look first to your CPU fan -- is the CPU fan plugged in and can you watch it spin when power is applied? (I've had ribbon cables curl -- or uncurl -- and jam themselves in between the blades ... that will cause this sort of problem. If there's something blocking the blades from spinning, yank it out of there.) If the fan is freely spinning under power .... what RPM range is shown by the software? If the software doesn't report CPU fan speed, sometimes the BIOS can do so. You asked above how to determine that; you'll need to read your motherboard manual (or your computer hardware manual if you purchased a complete computer rather than assembling your own from components).

--------

EDIT: Oops. It was Cris001 who had the Athlon XP, roflcoptr hasn't mentioned a specific CPU. But most PC CPUs have similar temperature limits. Roflcoptr, if you're not using an Athlon XP, just look yours up at your manufacturer's website. But it will likely be 80-90C. :)

[roflcoptr]

uh oh.... I think I've killed it.

I checked the cooling fan and all seemed to be working okay, but it was incredibly dusty, so i switched everything off, unplugged the cpu fan from the power on the motherboard and took off the fan. Now here's the bit where I admit doing something really stupid. I didn't realise that the cpu fan would just unclip and detach, so I went and unscrewed the whole thing from the motherboard to remove it. This didn't seem too stupid until I went to put it back in. Crap! The casing with the screw sockets underneath the motherboard had shifted and I couldn't get it back in. Oh well, after an hour or so of tipping and tilting, nudging and sliding the thing around, I finally manouevered it so that I could get the base of the cooling fan housing screwed back in. (I probably could have been a bit gentler, but what's done is done)

While the fan is off, I've had a look at my chip and I can now tell you that it's an SL7PM Intel Pentium 4, 3.00GHz chip and the recommended maximum temperature is only 69.1C. So I guess the 103C that Speedfan was reporting earlier is a little bit higher than it should be getting. :eek:

So I've put the fan back on, reconnected everything to the way I think it was before. Except now when I turn my computer on, I get the screen with the motherboard logo (which used to flash up for about 1/2 a second before my computer booted), and it just stays there. The fan is still running fine, but nothing seems to be whirring or clicking or making "loading up" noises. No idea what I've done or why this is happening??

So I pull the fan off again (this time, just unclipping the housing) and check it out. Everything looks okay. I connect it back up again and now when I try to start my computer, it still gets nowhere, but it only loads up for about 20 seconds before doing a complete shutdown.

I think we've managed to establish quite conclusively that I've got a hardware problem, so if this is no longer the place to chat about this sort of stuff, then I'll run away and try to get one of my local tech mates to have a look at it. But before I go, (or if anyone is still feeling generous with their time and advice) can I ask if I should be alright if I just replace the cooling fan? Also, if I have fried the chip or the board or something, is there still a chance I'll be able to recover what's on my hard drive? I've got 99% of it backed up (thank god), but I'll sleep a bit easier if someone can tell me that I might have something salvageable.

Thanks again for the help.

[jggimi]

My guess -- and its only a guess -- is that your removal and reinsertion have done damage. Static discharge, or physical stress on the cpu underneath.

At least now you know what the problem .... was.

Good luck, do let us know how things work out for you!

[Cheech]

I hope you used some sort of thermal grease on your CPU when you reinstalled the heatsink. If not you can perty much guarantee the CPU is still overheating when booting. CPU's only take seconds to get blistering hot, if you want to test this, take off the heatsink, firmly press your finger on the CPU, turn the power on, and see if you can count to 3 before you get a 2nd degree burn :)

Replacing the Fan/heatsink with proper thermal grease is a 50/50 shot of it working, I can tell the CPU is still somewhat working because it would not get the the MB logo unless it was, but parts on the CPU may verywell be toast.

As for your data, it should be 100% fine, I can cite some ways it could have corrupted it self in your situation, but I think you will be fine.

[roflcoptr]

aaaahhhh, thermal grease. That's what that metallicy gooey stuff I scraped off was. I had to scrape it of with my thumbnail to read what sort of chip it was. There's still some residual grease left on the bottom of my cpu fan, but probably not enough. Hmmm, well I'll have to try to get some of that. A regular computer store would sell it right?

I was using was an Intel fan and heatsink (Part No: C91249-002) and other than it being dusty as hell, is there any reason that this wouldn't have been cooling my cpu sufficiently? Is there a super powerful cooling fan I can get that would actualy get my chip temp down below 70C?

Or is there a chance it could be something else that's causing my computer to run at such high temps?

I have to admit, as much as this has been frustrating at times, I'm actually having fun trying to figure this stuff out and pulling the box apart. To both jggimi and cheech - thanks heaps for your help. I'll buy a new cooling fan and heatsink and see if I can start it up again. I'll keep you posted.

[CWR03]

There are many, many different types of processor heat sinks available, almost all of which will do a better job of cooling than your stock set-up. Two things to look for, at least in my opinion, is whether the contact area is all copper which aids in heat transfer, and the noise level of the fan. I use a Coolermaster setup which is very lightweight and all copper, plus it's almost silent (I believe it's only 18dB). It's made from very thin strips of copper which are sandwiched in the center and spread out to form almost a flower shape, with the fan mounted in the center.

[jggimi]

I'm gonna take another wild guess, and assume the intel CPU fan speed was low, even though it was spinning. I say that only because my son told me that was a problem he was having with his dekstop recently.

Exceeding the maximum operating temp is likely to cause damage, but we can't know the extent of any damage unless you run diagnostics. And any damage may just shorten component lifespan without showing any indications of problems now.

This current machine I'm sitting at is an AMD Sempron 2600+. Effectively the last of the Socket A CPUs, maximum temp is 90C. Currently my temp stays at about 1/2 maximum:jggimi-> sysctl hw.sensors
hw.sensors.0=lm0, VCore, 1.63 V DC
hw.sensors.1=lm0, +3.3V, 3.31 V DC
hw.sensors.2=lm0, +5V, 4.95 V DC
hw.sensors.3=lm0, +12V, 11.31 V DC
hw.sensors.4=lm0, -12V, 0.64 V DC
hw.sensors.5=lm0, -5V, 5.11 V DC
hw.sensors.6=lm0, 5VSB, 4.92 V DC
hw.sensors.7=lm0, VBAT, 3.25 V DC
hw.sensors.8=lm0, Temp1, 46.00 degC
hw.sensors.9=lm0, Temp2, 49.50 degC
hw.sensors.10=lm0, Fan1, 2909 RPM
hw.sensors.11=lm0, Fan2, 3375 RPM

jggimi->
An aftermarket cooling tower/fan assembly usually comes with thermal grease included; you shouldn't need to purchase it separately.

Once again, I wish you luck, and that you can still use your CPU and your motherboard when repairs are completed.

EDIT: The thermal grease may be a rectangle already in place with a peel-off cover, or it may be a tube of stuff you smear into place.

[roflcoptr]

After speaking to some tech people, I've been thinking that even if my heat sink and fan were a bit dodgy, the temperatures were a bit high for that to be the sole explanation. So I've dropped it in to someone who knows how to check the components properly. Hopefully, the CPU is still okay and I'll get back a working computer.

Thanks again to everyone for the advice and once it's all over I'll let you guys know the diagnosis, solution and if Auto GK worked again. Cheers.

[roflcoptr]

Well that was a drama. I won't go into the details, but after a month of being without my computer, it's now up and running again with a working CPU (didn't need a new one) and a restored hard drive (that is now 100% backed up).

Anyway, I just tried running AutoGK again, and this time instead of switching off, the computer just froze and refused to do anything. It seems like it happened at the same point as before.

I'm running speed fan and the temp got up to ~75, (it's at 57C when idle) but nothing like the 103 it was hitting before when it shut down. So now I'm thoroughly confused. It may still be a hardware thing, but it doesn't seem to be the cooling fan. Any ideas?

[BigDid]

... I'm running speed fan and the temp got up to ~75, (it's at 57C when idel) but nothing like the 103 it was hitting before when it shut down. So now I'm thoroughly confused. It may still be a hardware thing, but it doesn't seem to be the cooling fan. Any ideas?
Hi,

75 Farenheit or celsisus?
If celsius, still too hot for an Athlon. You need to get it down at least under 70!
Quick and easy, try removing the side panel of your case. If not enough you'll have to blow a home fan from 20/30cm aiming at your heatsink.
If not enough you'll have to get a better heatsink/fan; a good zalman or thermalright can do it...
If not enough, report back :scared:

Did

[roflcoptr]

I was talking celsius, so yeah, it's still pretty warm. I'll try it with the side of the case removed and the room fan trick tonight.

Apparently, some of the first gen Intel P4 3.0gig chips (what I've got) ran very hot, but it didn't affect their performance. I'll look into other cooling options as well though.

[Gehenna]

Even @ idle ,those temps seem too high for me,especially considering your CPU is rated temp is 69.1 Max (http://processorfinder.intel.com/Details.aspx?sSpec=SL7PM#).

What do the temps show in your BIOS ? (usually pressing DEL while machine is in POST should take you into bios)

Also what kind of vcore reading are you seeing (Your CPU should be between 1.287V-1.4v).You should see this in the bios

For the moment ,imho you are probably better off using sp2004 or prime95 to test the thermal problems you are currently having,instead of using AutoGK.

just google sp2004 or prime95