1.2 Ghz T-Bird Unstable at 1.1 Mhz & above :-(

[collindv]

OK guys, I have a situation for you. I had an 800 Mghz duron (KT7A Motherboard) overclocked to 888 (111*8) and decided to upgrade to a 1.2 Ghz T-bird. I got the chip installed fine (used arctic silver & pencil trick) and booted up. The system booted to Windows fine, but locked up after a few minutes. I dropped back to 900 Mhz and kept increasing until the system became unstable, which occurs around 1100 (11*100). I assume I am overheating. My question is this, I am currently running a Global Win FOP32 HSF, shouldn't this be enough to properly cool to at least 1.2 Ghz? I don't know my chip temp because I'm running Windows XP beta, which doesn't like my Via Hardware monitor software :-( Anyway, if this HSF is not adequate can someone recommend a better one (not peltier or water)? Thank you.

[VecchK]

A FOP32 may not be good enough. Even if it is though, you might be happier with a newer heatsink so you can overclock. I'm currently using a ThermoSonic ThermoEngine w/ a 7200 RPM fan. Works really well, but you can do better. Go to http://www.hardocp.com. They do a heatsink roundup every few months. In may the ThermoEngine won (schooled the FOPs and barely beat the copper one) but I think in June there were two newer ones that were better. Never use an orb. They crack chips and don't do that great of a job anyway. The best one there doesn't even use the socket clips. It uses the 4 holes surrounding the socket (yes, they're there!). Pretty neat, eh? Happy cooling.

[el]

okay you have the kt133a chipset so go for 133x9! I think the fop32 can handle that at default.

[joey_rjm5]

Is the one that uses the 4 surrounding holes the zalman radial fin hsf. I saw that and it looks awesome but it's really expensive. I don't know what's with your chip. Maybe you should try resetting the hsf. it might help. Also, you might want to try running it without the pencil trick. Some say a pencil eraser cleans it good. Try booting and going into the bios and watch the temp to get an idea of what your temps are. Good luck

[VecchK]

Yeah, that was it. The Zalman Radial fin copper heatsink. It's the best Socket A heatsink on the market right now. Didn't even think to look at how much it cost though.

[collindv]

OK, checked my temps from the Bios (since I can't do it from windows :-( ) Running at 1.2 Ghz (9*133) the temps are 44 C. Is this too high, I would think so considering that the cpu is under very little load when checking the temps from the Bios? Thinking about replacing the Globalwin with a Zalman Radial. Any comments welcome.

[VecchK]

It will stay toasty warm in the winter time but it's not too high. I doubt a 1.2 goes unstable at 1.2 with a temp of 44.

[collindv]

Yeah, but that temp is from the bios. Isn't the temp alot higher once windows boots? Also, I took it down to 1064 (8*133) and the bios temp drops to 36 C. Seems to run rock solid at that speed.

[WillysNut]

The Zalman radial fin cooler is "nice looking" but you really could do a lot better. Something Copper like the SK6 or the Glaciator will eat the heat off that chip and allow you to at least run stock speeds...not to mention much higher than stock :)

It's obvious that if you have to under clock to 1064 on a 1.2 chip to get 36c idle...which is pretty good, it's time to get some cooling on that bad boy. Good luck!

[AmIdYfReAk]

i heard that if your fan on your heatsynk dosent meet a Predetermand rpm setting... that your mouther board would lock up and be VERY unstable...

it was a higher rpm.. i forget what....

[joey_rjm5]

I think you should try to get another one. It just doesn't seem right that it is unstable. They are supposed to work out of the box with a stock hsf. Something is up. Take it back.

If you want to see your temps in windows, you could just install a previous version. It wouldn't take too long.

[joey_rjm5]

joey_rjm5 (Jul 19, 2001 12:41 a.m.):
-update- sorry guys I meant to modify the previos post.

The zalman isn't much more expensive than some of the other top of the line hsf combos. It's just too much for me.

I think you should try to get another one. It just doesn't seem right that it is unstable. They are supposed to work out of the box with a stock hsf. Something is up. Take it back.

If you really want to see your temps in windows you could just install a previous version. It wouldn't take too long. That doesn't really matter though if you have a problem like the cpu is not running at spec.

[dozier768]

you also might try using the defogger paint or circut pen for the bridges sometimes there is just too much resistance with a pencil and it could cause stability problems

[collindv]

Let me ask this, Would my temps be too high if I used too much arctic silver? This is only the second processor I've installed and I think I may have layed it on too thick (probably about the thickness of your thumbnail). I might try to go ahead and reinstall it, but how do you get the silver paste that is already applied off of the chip in order to start over? Also, what is everyone's opinion on copper shims? I had one in my duron, but didn't put it on the athlon, maybe I should try it. I appreciate the help all.

[Paiynn]

Have you considered your memory or PSU?

My 1.2 had a TON of lockups at 1.2 till I switched out the PSU.
After the new PSU I still locked up maybe once in two days and couldnt post at 1.4. After changing the memory, I was rock solid at 1.2 and clocked to 1.5 with no problems.

[Paiynn]

In answer to your other question, I use acetone to remove thermal compunds, works about 10 times better than rubbing alcohol.

There is also a how-to on Arctics site for applying AS2. As thick as your thumbnail is alot more than they say to apply. An amount the size of 1/2 of a grain of white rice is what they say to cover the whole core.

[collindv]

PSU is nearly new and should not be causing a problem. Memory is 256 MB of Crucial 133, also should not be causing the problem.

I had to use WD40 to remove the pad on the bottom of the heatsink before I installed it, would that be causing a problem (I followed it up with an Isopropyl alcohol scrub)? I know the people at arctic silver say not to use WD40 on the base of the heatsink because it doesn't evaporate. I don't know, I'm trying to avoid buying a new heatsink if possible, just looking for ways I might have flubbed the HSF installation.

[Paiynn]

Not to beat the issue, but is your PSU at least 300watt?
At least 300w for a good brand name or 350+ for a generic.

Only thing I can think of, rest of your stuff sounds like it should be dandy-o

[collindv]

300W Power supply came with the case I bought back in April. Case is made by Antec.

[collindv]

OK, I reinstalled the HSF with the copper shim and less artic silver and am getting the exact same bios temps that I was before (41 C at 8*133=1064). Short of buying a new HSF I'm at a loss.

[collindv]

OK, tried to boot at 1120, bios temps say 42 c. Same as when I try to boot any higher, it boots to windows, lasts about 5 seconds and crashes. Did I get a bum chip and is there anything I can do about it? Or should I bite the bullet and drop $50 on an sk6? Is there a way to see if the problem is in my power supply, short of replacing it with a higher wattage?

[dozier768]

i dont think its the chip when they are bad they usually just dont work. have you checed to see if you have ang irq sharing or conflicts going on? my sound and video card shared an irq and it caused all sorts of stability problems and crashed all the time. i just moved the card down one slot and it fixed it right up

[collindv]

Nope, IRQ's are fine, no conflicts at all. Plus, it's rock solid at 1064.

[!-=sky=-!]

did you try to pull out all the cards then boot the machine?

that might work.......

also, u can try to lower your memory settings

[collindv]

I've tried it with the memory settings as low as possible. Removing the cards is one thing I have not tried. Only have a SoundBlaster ISA sound card and an ethernet card though. Might be worth a shot to see if the problem is in the power supply. Stuck at work unfortunately, will have to wait until 5!

[collindv]

Also, here are the markings on the chip, anyone have any experience with one similar. Is anyone running a FOP32 with a 1.2 athlon? If so, what are your temps? Just curious. Thanks.

A1200 AMS3C
AVIA0113BPMW
Y73507002074
1999 AMD

[collindv]

OK, I removed all my cards and disconnected all my case fans (it's open anyway). It started up and ran fine for about 3 minutes, which is as long as I have gotten it to run at 1200 so far, but then it froze. I went ahead and put the cards back in and restarted it at the same speed (1200) and it again ran for a few minutes before locking up. Seemed to be a fruitless experiment. I think I'm going to try uninstalling windows XP until I get these problems straightened out, at least I will get a better idea of the temps I'm running at. Who knows, maybe it's XP that is causing the problem.

[Fiz]

I read about a guy who tried to OC in windows XP and had all sorts of problems wit it. No matter what he did it just kept locking up for no apparent reason, I would try a different version of windows and see if that helps.

Fiz

[diehrd]

If you are hitting 42c in bios you have some head room,HAVE you cleaed the CMOS ??

I am guessing this is a fresh build if so always clear the cmos before first boot up and remember to disconnect power completly before you clear cmos

[AZZKICKER]

what size power supply you running?????????

if it aint a 400 or better than that chip aint going to run over 1100 correctly


take it from the man that tride and trued this theroy from the beggining

when i got my 950 t bird 1100 was my wall

i connected 2 300 watt power supplys together at the ATX connector and i broke 1357 with my 950 ( that was air cooled but case temp was 20F because it was in the window and it was 9F outside...heeheee

this was in november and december of last year

trust me........if you aint got enough power then your not going to run right at all

AZZ........electronic engineer

[diehrd]

This may be a bios problem,If I were you I would go to the M/b home page and look for updated bios,And read to see if in fact there is a known conflict in the area of processors.

Abit I think had an issue with a 1.33 it would default to a weird setting and cause hangs

[collindv]

To answer the questions, I did not clear the CMOS before first boot and my power supply is an Antec 300W.

I got Windows XP uninstalled and am having the same problems. I can see my temps now at least. Running idle in Windows I'm running about 34 C at 1115 Mhz. It's looking more and more like my power supply just isn't going to handle anything higher. I guess I will bite the bullet and try ordering a new power supply. Anyone have any reccomendations on a good, cheap 400W PSU, or any other suggestions on things to try?

[AZZKICKER]

when you run AMD you have to run something better than cheep

stay intel if you want a cheep system......

buy only a antec 400 watt or a entermax 430 watt

[collindv]

Out of curiosity, is anyone else running above 1100 MHz with a 300W power supply? I hate to drop $75 on a new 400W Antec if that's not the problem.

[joey_rjm5]

I think it probably is the power supply. If you have another power supply you could connect the two together and see if that works. It would save you some cash.

[collindv]

The only other power supply I have is off an old 486, would this work? How would I wire them together?

I actually got it to run at 1200 (9*133, 1.8 V) for about 20 minutes last night. Temps were 34 C idle and 53 C full load with an ambient of 31 C.

When I shut it down and restarted it went back to the same old thing, freezing after a few minutes.

[collindv]

OK, here is the latest and greatest on the situation. I finally bit the bullet and bought a 400W Antec power supply last night. I got it installed and attempted to boot at 1200 MHz (9*133). Surprise surprise surprise, it froze up 30 seconds into Windows, just as it had been doing before. Well, being pretty much pissed off that this didn’t solve the problem, I decided to OC it just to see what would happen. I started it up at 1333 MHz (10*133). To my surprise, the system did not lock up after 30 seconds. It actually ran for a long time. I was able to check my temps, which turned out to be exactly the same as they were at 1064 MHz (33 C idle, 53 C full load). Well, I thought the problem was solved, that maybe there was just an issue with the 9X multiplier setting I had been using. Turns out I was wrong yet again, the system ran fine for about 2 hours then did the same thing, suddenly froze. I rebooted it at the same speed and it lasted another half an hour before freezing again. For some reason though it is more stable at 1333 than it is at the rated speed of 1200, even with the same FSB!

I’m pretty confused as to the problem at this point. The only thing I can narrow it down to is the pencil connecting the bridges. Maybe they just aren’t adequate. I’m going to go ahead and erase the bridge tonight, and I have ordered a conductive pen. Hopefully this will take care of it.

Anyone else have any ideas, now that the power supply has been eliminated as the possible cause of the problem?

[mojo]

It's probably the heat. I think your CPU is getting too hot. I have the same CPU (1.2 Ghz AXIA retail) and KT7A mobo running at 1200 Mhz. I installed the VIA hardware monitor software. I have Globalwin CAK38 (that mutha is f***ing load) and Arctic Silver 2. At idle temp is 34 C. I fired up SETI and watch the temp rise to 48 C, then 49, then it hangs. I shut it down for 10 minutes boot up again. Ran same thing. Again exact same result. I'm guessing the VIA hardware monitor software is about 2 degrees off. When it hangs at 49 C, I'm guessing it's actually 51 C. Somehow my CPU does not like running that high a temp. I lowered the CPU clock to 1050 Mhz. Ran fine. I ran SETI for 1 hour and temp never get above 45 C. This pisses me off real bad. My f***ing retail AXIA cannot run at it's spec speed, even with a good hsf and Arctic Silver. My case is well ventalated (3 fans including PS). This sucks. I'm going back to Intel, at least my 700 PIII running at 980 is rock stable. Too bad I F***ing sold the system get this mutha of ****.

[mojo]

I'm begginng to think I may need a better PS also. I am currently using a 300 watt Enlight PS. Anyone have any recommendations? Sorry for the profanity in my previous post. I am just a little annoyed with all the problems I'm running into with VIA chipset and AMD processors in general. I never have any such problems with Intel chipsets with Intel CPUs. I know the Athlon is superior and all, performance wise. But good does a powerfull CPU does if you system is flaky and can puke at any moment. Thanks.

[PDL]

Could it be the Vcore voltage???
Try going a bit higher and see if it is stable at 1.4.

Just a thought!!

[rezo]

i think its the heat... 53 degrees C is pretty hot.. almost 130 degrees F... the Thermalright SK6 is a really good heatsink and it doesnt cost too much either... you can get one with a 7k rpm delta fan @ kdcomputers.com for around $45... the fan is really noisy but it keeps the temps down... if you dont want the noise but you still want the cool you should move to h2o...

[Greybeard49]

I have had a similar problem, so here is what happened to me. An Asus A7V133 and a 1.2 Gig Athlon. Note all temps below are at idle. When I first got it, it would lock up if I put the case cover on. With Asus probe MB=35, CPU=55. Thereafter the case side was always open.
Some said ASUS probe read 10C high so I got MBM5. Then CPU read 46.
Figured it was the tape on the unit so I applied HS compound and gained about 3C, now at 43. However on days it was 85 to 90F out, with NO AC in my house still crashed.
Also note that many times if I restarted the computer and the fan stopped during reboot, I would have to wait 10 to 20 minutes before it would start again. Heat build up when fan wasnt running even if processor was off. This is not unusual cause once the fan is off heat disipation drops. Only that it locked up is unusual.
Tried with side of case open with box fan. MB dropped about 6 to 8C but CPU only dropped 2C.
Now I was sure heat wasn't getting to heatsink. Took the coolmaster off and put back on 3 times, same results. Tried a GoldenGate heatsink, same results. Everytime I took off the heatsink it looked as though the connection heatsink to chip was perfect. Would squeeze out almost all the compound. At this point I concluded the CPU was deficient internally cause it could not get heat to top of cpu. And decided to try to get it replaced.
But before that one night it died while running showing only 46C, but restarted after the 15 minute wait. Next day did same but was now dead permanently.
Hopefully I will get replacement tomorrow. OH forgot, when it finally died it was 90F out, but I had the bus speed down to 100, so the chip obviously could not get the heat to the heatsink. Hope I don't have to get aggravated by the dealer I bought it from.
Please try the box fan thing. I am really interested in your results. Personally, from reading here I think some of the chips just aren't right with their internally cooling capability. And I think more than a few crash when the CPU temp is in the high 40's.

[collindv]

You just described my situation exactly. I took off the HSF to remove the pencil traces, reinstalled it perfectly and the chip refused to post. Looks like the chip burnt out. I sent it back on Friday and am hoping (but not expecting) that they will send me a new one. Also ordered a new SK6, so hopefully when they get here the problem will be eliminated.

[VecchK]

mojo (Jul 27, 2001 09:35 a.m.):
It's probably the heat. I think your CPU is getting too hot. I have the same CPU (1.2 Ghz AXIA retail) and KT7A mobo running at 1200 Mhz. I installed the VIA hardware monitor software. I have Globalwin CAK38 (that mutha is f***ing load) and Arctic Silver 2. At idle temp is 34 C. I fired up SETI and watch the temp rise to 48 C, then 49, then it hangs. I shut it down for 10 minutes boot up again. Ran same thing. Again exact same result. I'm guessing the VIA hardware monitor software is about 2 degrees off. When it hangs at 49 C, I'm guessing it's actually 51 C. Somehow my CPU does not like running that high a temp. I lowered the CPU clock to 1050 Mhz. Ran fine. I ran SETI for 1 hour and temp never get above 45 C. This pisses me off real bad. My f***ing retail AXIA cannot run at it's spec speed, even with a good hsf and Arctic Silver. My case is well ventalated (3 fans including PS). This sucks. I'm going back to Intel, at least my 700 PIII running at 980 is rock stable. Too bad I F***ing sold the system get this mutha of ****.


Hey Mojo, if you're going back to Intel and selling that 1.2 GHz AXIA, let me know. :-p I didn't order my CPU by code and they gave me this POS 1.2 that won't go above 1.33.

[VecchK]

.. and sorry for the kissy face. I was trying to make the one with its tongue sticking out.

[dimitri]

Acck i wish i'd read this sooner... This is most likely just a bad pencil job. When the connections arent perfect different multi's will work while others dont. IMO the safest way is the magnifying glass/defogger paint technique. I'm running at 1.4 with a a decent 300W ps. If your chip had the 1.2 code and you bought it new, 99% of the time it'll run as rated. I'm pretty sure teh pencil job caused your lockups on certain multipliers, i've seen it on 4 comps now, it's frustrating because it's apparently random and you try and find a pattern.
As far as you burning your chip, how did that happen?? Did you not have the heatsink properly seated ? Also, if the chip was initially locked, did u try booting at 9x133 b4 the pencil job? As i said earlier, i wish i'd noticed this post earlier b4 u spent money on a new PS and hsf. The FOP-38 is o.k. and those temps u had were fine for 1.2 , oh well, thats how we learn..
GL

[collindv]

Ok, will answer best I can. I didn't try the new chip before unlocking it, I know that was silly, but I was in a hurry :-( Very stupid and at least I learned my lesson from it. As far as the power supply, I really needed a 400W anyway, just to be on the safe side (I need to add about 2 more case fans). Also, I was running the FOP32, not FOP38. As near as I can tell from the reviews I have read, the sk6 should beat the temps posted by the FOP32 by anywhere from 5 to 10 C, so it was probably a worthwhile investment anyway.

I do believe it was the pencil trick which was messing my chip up now. I'm not sure how it got burned up, I'm assuming the HSF was not fully seated. I don't have a removable motherboard, so it's kind of cross your fingers and hope it's seated well or take 2 hours to disconnect everything and pull out the Mobo, which of course has it's own risks. All in all it's been a definite learning experience. A word to the wise all, test the chip before unlocking.

Anyway, thanks to all those who helped, I figure I should be hearing from the company I bought the chip from tomorrow to see if I will get a replacement. If not I guess I'll run the SK6 on my duron 800 chip. I will keep you all updated. Thanks again everyone.