Tech Support Forum banner

P4c800 Dx + P4 3.4 Freeze Problem

2221 Views 24 Replies 10 Participants Last post by  whodat
Hi. I just upgraded my prior P4 1.8 with a new P4 HT 3.4 yesterday. After the upgrade something very strange is happening - after using the computer for some minutes it freezes for around 5 to 10 seconds then it continues working as if nothing has happened (no reboot).

I was running bios version 1019 on the P4C800DX motherboards - I upgraded today to version 1021.004 , which looks like improved things a little , but the problem was not solved yet.

Ayone could give me a hint on what is causing this problem and how to solve it ?

Follow below my system specs:

Motherboard: P4C800 DX BIOS 1021.004
Memory: 2 x 256 Mb Samsung DDR 400
Video: Radeon 8500 AIW
Audio: Augigy Platinum
+: Realtek NIC // USR Modem Card
Power Supply: Seventeen Professional 350 W (true watts)
HDs: 1 Maxtor Y120L0, 1 Seagate 312002, 1 Seagate 312082, 1 Seagate 380021A and a 1 WD WD200AA
Running Windows XP SP2 Prof

Thanks in advance for the help !

Regaerds,
Raul
Status
Not open for further replies.
1 - 20 of 25 Posts
Hi,

If it were mine, the first place I would look is yoiur power supply. First of all, 350 watts of any supply is not enough for this unit. Second, although you need at least 480 to 500 watts as a bare minimum, I would not put anything in it except an Antec or Enermax power supply.

Why not boogie on down to the local BEst Buy and try one of the high quality Antec or Enermax supplies of a bare minimum of 480 watts (more is better), hook it up and see if it solves your problem. If so, then keep it, but if not, then return it and it costs you nothing. I would put my money on that issue.
add it up here

add a minimum of 20% to the top
@ rzotti
Welcome :smile: . What kind of temperature readings are you getting before or during these freezes?
I live in Brazil - here once you make a purchase you can usually only return it if the product is defective - so I would need to be sure the problem is caused by the power suppky before purchasing one (good psu's are expensive here).

In general, I am getting very high temperatures on the CPU (i'm using asus probe to measure it). I am obtaining a low of 56 celsius and a high of 74 (!) with high processing, and my first though was that this would be the causes for these momentary freezes I am finding. I've made some tests and the computer has not freezed at all when working at the top 74o temperature for 10 minutes, but then it froze later for 10 seconds when the temperature was at 58 for some time already - so i'm almost ruling out the temperature as the cause for these (although I am aware that I need to deal urgently with this high temperature to do not loose performance or risk burning the processor).

What I find strange is that the freezes are only for some seconds (up to 20 sec until now) . I would expect that once the system froze, I would need a reset for it to come back. Although I am used to computers for years, this is the first time I've seen something like it.

I was thinking to take out the modem, network and audio boards I have to see if it is some kind of conflict (or driver issue) that is causing these freezes. WOuld it be worth the try:

Any help or ideas you can provide on this issue will be very welcomed.
See less See more
@ rzotti
Your temps ARE very high. That is something that you should address as soon as possible. It may be why you are having problems. It might be that excessive internal temps are overheating your PSU, as Tumbleweed36 has suggested. Try to reseat your heatsink and fan. Use Arctic Silver 5 if possible...it is the best thermal paste out there.

I would think the system would freeze and need a reboot,like you said. It is a strange type of momentary freeze. I've never seen or heard of it before like that, either.

Yes, trying to remove and reseat anything would be worth the effort, if you are still having problems after fixing your overheat problems.

Let us know how it goes
Hi,

I still think your power supply is suspect, however, after seeing your temperatures, you may have more than one problem. Those temps you listed are simply way too high.

For now, the heat thing is a real immediate need to get fixed. Put the thermal paste in there, then try it. If the temps are at any time even under-stress above about 63c, then take the side off the case and blow a regular fan in there while working and see if that helps. You must get that temp down because heat kills. There are other things we can do to help you if that is a real issue.

Let us know what works and what does not and one of us might be able to help you with possible solutions.
I just spent several hours trying to identify the reason of the problem . First I re-installed the processor , took off the original thermal paste on it, applied a different one I gained when purchasing the processor (no brand, white colored). No difference in temps. Re-arranged fans in the case (I have 3), Got a one / two degree improvement. Suspecting bad flow of air inside the case, as suggested , I opened the case and started working with the case opened. Opening the case, I got a low temp of 55 and a high of 69 -still way too high and random freezes continued.

To see if I was having a power shortage, I took out all boards but the video (one by one) , plugged out all drives but the one with the os. Temps have not lowered, ramdon freezes continued.

Then I begun turining off all optional on-board devices in the bios . No improvements.

Then the last change I made apparently eliminated the freezes - I turned off hyper-threading.

Well at least it is identified that HT was the cause of the freezes. Now I am in doubt if it is the combination of HT + high temps that caused them, it is any kind of incompatiblity / corruption from the OS /drivers with HT, I just purchased a faulty processor or the gods of technology are just punising me for not going on with an Athlon 64-FX. :smile:

..... Any suggestions ... ?
See less See more
Just FYI I am running the same board with the 1021.004 Bios with no problems. Specs in my signature.

Now that you have isolated the problem to some extent there remains the possibility that some running software, virus or driver may be having issues with hyperthreading. I would not rule out the Bios either, I have had strange things happen with the P4C bios.

It could be the CPU but I dont think we can be absolutely certain of that yet. The same procedure you used to find that Hyperthreading is related to your problem could now be applied to the software and drivers you presently have running on your system.

The symptom of freezing without a crash may be a good thing in that you can check your event log and see if there is anything being logged relating to this failure. I would try that next.
@ rzotti
IMHO, I think turning off your HyperThreading proves that your CPU and HeatSinkFan connection is poor. Did you thoroughly clean off the old paste using denatured alcohol? Did you apply just a thin consistent layer of the new paste to the CPU? It is possible that the interface between the HSF and CPU is defective and unable to connect properly. I had one happen recently to me. :sad:

The poor-cpu/hsf-connection scenario is even more apparent to me when you have your case side open and you are still seeing extremely high CPU temps. Turning off HT has the effect of using only half your processor, therefore, when you stress the processor with no HT, it is only developing 50-75% of the heat it would under HT conditions. I think that is why your system works w/o HT. No PSU problem, no virus problem. I bet your system still runs quite hot w/o HT :3-sqhot: . The P4 3.4C processor is a very hot one. The standard HSF that comes with it is really not up to the task, but you should not see the high temps you are seeing even with the poor stock cooler. You might want to try an aftermarket HSF, like the Zalman 7000 or 7700. The ArcticCooling Freezer 4, Scythe SCNJ-1000 Ninja, Thermalright XP-90, XP-120 and SP-94 are also excellent choices. :cool:

After you resolve your heat issues, you may also have PSU or software/driver issues. Have you run a spyware/antivirus program to make sure you don't have a virus that takes over your machine, as Barry_R has suggested? You could open 'task manager' (ctrl-alt-del) and monitor your CPU usage at idle. It should be 0% to 5%. If there is a virus running, your CPU Usage may be running 100%! Maybe that is what is happening during your 10-20 second 'freezes', your CPU is at 100% usage...I have seen THAT before. :3-nuts:

Still, even with a virus, you should not see the CPU temps you are seeing. With a proper aftermarket HSF, stressing your CPU for 24 hours using 2 instances(for HT) of CPUBurn, you should see maximum CPU temps of 45C to 50C temps and 55C to 60+C with the stock cooler.

Thanks for your feedback so far. Keep us posted on your progress :sayyes:
See less See more
@ please

I totally agree with your comments. I just spent several hours again trying to cool things a little here. Took out the processor , cleaned all the paste again (exaustivelly), applied paste again, tried my old p4 cooler. No change in temps. Tried everything again (with the cooler that came with the new p4), this time making the paste applied a work of art, looking like the surface of a frozen lake :). Got a low of 52, high of 70, ht disabled, case opened.

Regarding virus and spyware, I am pretty confident I am clean. I do not doubt that I have a driver conflivt with HT , I heard nasty things already about audigy drivers and dual processors . I will try to purchase a new cooler to see if things improve, but to tell you the truth, I am beginning to lose hope.

Oh, and by the way the freezes continue, even with HT disabled. They are just alot more sporadic than before (my wife confirmed three freezes during the day while I was working). :4-dontkno
Make sure you have Q-fan disabled in the bios. A 3.4C cpu should run about 29C-34C at idle and about 40C-48C at idle (unless your ambient temps are way up there) Your CPU is a Northwood, which will throttle about 50% if the temps get too hot (provided you enabled that feature in the bios) or shut down all together at about 70C. Your right however, in Brazil the stock P4 cooler would be inadequate, I suggest the Zalman cpns7000 Cu, that should help bring those temps down, I would also suggest you make sure your case fans are positioned correctly...ie front fans set to suck into the case and rear fans set to blow out....there should be arrows on the fans indicationg direction of airflow. I also concure with Tumblewwed a 350W unit is inadequate for your set up, a unit of 450W or better from a reputable company like Antec or Enermax would be much better. And out of curiosity why are you using a NIC card....the mobo has LAN built in
After a long search, got a Zalman 7700 today. That`s what I call a cooler. After installing it, temps got to a low of 50 high of 64 with Ht on and case closed. Still high. Momentary freezes for some seconds are still happening often with HT on.

I am thinking that the cpu is defective . I was thinking in trading this cpu for a new dual core one (3.0ghz), and purchase a new asus board for socket 775 (p5gd2). WOuld thar be a good option ?
@twajetmech

Thanks for the tips. When I was playing with the bios options to see if something decreased the problems I was having, I left Q-Fan enabled. I disabled it now - but I have not seen improvemtens on temps and freeze yet.

I used the on-board lan to connect to the internet and the other one to connect to my home network. Currently I am testing connecting the cable mdoem though USB, so probably I will take this aditional nic card out in the future.

I have never seen the psu you mentioned for sale here in BrazIl, although I know they are all very good psus - the best one they sell here usually is the seventeen brand, btw the one I purchased. I agree that a more powerfull psu would be good, and I plan to upgrade it soon. I made a test with my system running only the processor, video card and one HD and the problems were still there so I think that at least the problems I amn having are not linked to the psu. Would you agree to this conclusion ?
not necessarily, a psu with fluctuating/low/high voltages will still cause the problem. and you have the latest bios?
look for a p/s that puts out a min of 20amps on the 12v line,you are sure to have a local brand that puts out a quality supply,probably the same as here costs an arm and a leg
@whosdat: latest bios, 1021.004. I checked the voltages from the PSU, 12v oscilating between 11.5 and 11.8, all others with almost no variance to their reference values. (+/-1%).

Good news. Looks like I solved the freezes I was having I took out the realtek , modem board and the audigy sound cards AND unistalled the audigy drivers. I am just completing my first hour without a single freeze.

Temps are still high, I got to 67 this afternoon. but on low usage the cpu is usually at 53. I still can try to work with the case open, or otimize internal air flow - but in my experience it would improve my temp on a couple of degrees only. I am out of ideas regarding decreasing the temp now .... hmmm .. may be a liquid nitrogen cooling system :) ?

BUt not everything is good news ... My wife commented to me as soon as I arrived home that the computer was turning off alone almost everytime she was working ... hmm .. playing ( :) )on it. At that time the computer was in horizontal poistion. Could it be the the weight of the new cpu cooler ? After I put it in vertical position up to now no shutdowns happened.

Anyways, thanks for the help all of you provided. You guys created a great community here - please be proud of that !

Regards,

Raul
See less See more
I vote power supply. 11.5v is low for the 12v supply. If the other voltages are equally low, this could be the cause of your overheat. Check the other voltages as well.
I agree, the psu fluctuating like it is is likely the cause of much of your woes, along with the temps you are dealing with. You do need sound however, and I suggest that you use the audigy soundcard rather than the realtek or onboard AC97....be sure you diable the onboard sound and clear out any vestiges of the other drivers. Assuming your ambient temps are in the mid to upper 30C you should expect to run the cpu at about 10 degrees higher than ambient temps at idle providing you have good airflow through the case. Be sure that Spread spectrum is disabled in the bios, legacy usb is disabled in the bios and Turbo is set to Auto, this should help in stability. Also make sure your cpu voltage is set to stock (1.55) and memory is set to stock as well (2.65-2.75V) If finding a decent psu is difficult, check out some of the online stores like newegg.com, tigerdirect.com or zipzoomfly.com, they should have what you want at a reasonable price.
I agree 100% / get a bigger and BETTER PSU

Disabling the HT is the equilivant of removing 100lbs of a horse thats about to fall down !!

Not enough voltage (low voltage resisitance) builds heat / the sporadic mini freezes you are seeing is the CPU throttling down to save itself !!

If you continue to run the system in the current configuration you WILL burn something then that PSU replacement is going to be worth just alittle more !!

I suggest you re-install the P 1.8 until you get a better PSU or your going to smoke that cpu.


regards

joe
See less See more
1 - 20 of 25 Posts
Status
Not open for further replies.
Top