A once stable OC now is not

[thechief73]

Hello,

I have a bit of a conundrum. I had a stable OC of 3.74Ghz on my Intel P4 650 3.4Ghz. I ordered a Megahalems that was on sale for a future build and decided to use it on my current 775 until I order the rest of my parts. So I dissasembled my PC, only to find that on my Micro ATX Mobo (Asus P5LD2-VM) there are two inconveniently placed capacitors blocking the mounting brackets, so time to put everything back as it was.

The only things I did to change from the previous assembly was I cleaned off the dust from all the parts, I changed a few screws because they were not fitting in my stand-offs correctly, and I used Prolimatech TIM instead of OCZ Freeze. About the TIM, I have cooler temps than with the OCZ(So my logic, the temps are not the problem, both TIM's were @ ~60c under any load testing), I think the better tempratures may be because I got a better install with my old cooler(after 5 years some screws had loosened) or Prolimatech has some quality TIM, maybe both.

So to conclude this mess, I usally run OCCT(Overclock Checking Tool) to get a graph of the CPU temps and Vcore for about 30-35min(I know this is not long enough to check for complete OC stability but I have never incountered a problem gaming or everyday use) and I sometimes run Orthos for the same amount of time(then I stop it manually). I could always complete these tests before my rebuild and now OCCT fails in under 15min(Everytime PC restarts on its own) and ORTHOS just failed at 17.5min and halted the test but listed zero errors.

I have tried everything I can think of to find the cause and I am thinking it may be a PSU thing due to the restarts, but I think my Antec NeoPower 500w should be able to handel my modest 5yr old PC.

Please if you can think of anything that may cause this kind of behavior or make my tests that I could once PASS now FAIL let me know I am really bothered by this and want to get it resolved.

Thank You.

 

[wolfram23]

Hmm, I think you're on the right track about the PSU. I haven't experienced PSU problems myself but I know they can degrade over time, plus, it's very possible you somehow caused a surge when taking it apart, possibly from static or else just your procedure.

Alternatively, when you replaced the stand off screws, did you have little paper washers? Those are to seal it off from being conductive, so it's possible you're bleeding energy off/shorting something if you didn't use them.

Other than that, if everything seems ok then I guess try a little more voltage. If it's still doing the same thing, then it would be very likely it's a PSU issue. If possible, try a different one in there.

 

[saint19]

I think that is overheating problems. check if all is installed correctly

 

[d1rtyju1c3]

and seated properly.

 

[thechief73]

Thanks for the ideas everyone.

Hmm, I think you're on the right track about the PSU. I haven't experienced PSU problems myself but I know they can degrade over time, plus, it's very possible you somehow caused a surge when taking it apart, possibly from static or else just your procedure.

Alternatively, when you replaced the stand off screws, did you have little paper washers? Those are to seal it off from being conductive, so it's possible you're bleeding energy off/shorting something if you didn't use them.

Other than that, if everything seems ok then I guess try a little more voltage. If it's still doing the same thing, then it would be very likely it's a PSU issue. If possible, try a different one in there.

First - My PSU is older model but was never used until late August this year so in a way its brand new. I sure hope I didnt cause any static problems that would be a real bummer, all seemed to go OK and normal.

Second - I did use the washers, but I think there is 8 stand-offs and I never had enuff to use on all of them so a few do not have washers, but it was always that way. One thing though the new screws have a slightly larger head on them.

Third - Either my Mobo or CPU causes me to use a very high voltage for my CPU, the highest spec. VID on Intel datasheet is 1.6, Intel Lists 1.2 to 1.4 on the product page, my Mobo options are (Auto / 1.575 / 1.5875 / 1.6) When set to Auto at stock or OC'd CPU-Z reads vcore at 1.65@idle loads. This Mobo only has the very basic of OC options.

I think that is overheating problems. check if all is installed correctly

For a Intel p4 @ 3.74Ghz ~60c max under full loads I thought was quite good, I have read of people saying 68c is top limits for this CPU before throttling kicks in or damage occurs. But myself and they may be wrong. I cannot find a good source for the thermal limit for this CPU if you have sources please let me know. This is a link to my current cooler: ZALMAN CNPS7700-CU

and seated properly.

I never removed the CPU so that should be fine, I checked all 4 RAM Dimms, My GPU and Sound should be fine, they wouldnt install correctly or work if they were not, I also check all the front panel header connections.

I am thinking that the stand-off screws or maybe a bad wire connection possibly are the cause. I cannot understand why my PSU would be the cause even though im getting the restarts with OCCT, because all I have is a WD 160gb HDD, 4x4gb 667 RAM, ATI 4670 (Stock - Will OC when I get this sorted), Sound Blaster Audigy 2, DVD-Burner, in a Antec 300 Illusion w/all stock fans on low.

I am going to run memtest86 here soon and see what happens.

 

[thechief73]

UPDATE: 4x4GB DDR2 667 PASSED Memtest86+, OC'd @ 733MHz

 

[thechief73]

UPDATE: I ran OCCT again with Normal priority instead of High like I used previously(a band-aid solution) for a 35min run and failed at about 27min. Vcore 1.575(Lowest Mobo allows), max temp @ 63c as far as I saw. Im starting to wonder if it is a temprature problem. I may try to run it again with the High priority and see what happens at that setting again. Tomorrow I am going to remove the Prolimatech TIM and replace it with the OCZ.

Also I am begining to wonder if their is a problem with my Bios wanting to set my Vcore so high on this CPU, I may also to to revert to an older version before they added the Pentium D suppport or just re-flash the current version.

Any other suggestions are very much welcomed.

 

[thechief73]

UPDATE: I have now installed insulators on all my standoffs, someone dropped the ball when manufacturing the screws and/or stand-offs, because very few of them would tighten down for me out of 20+ screws.

I have also reinstalled the TIM and hopefully I have done a better job of it this time and drop my temps a few C.

One question though, I do not understand why you would put insulators on the top of the motherboard and not the bottom? Seems like it would be wise to do both rather than one side, because the bottom of the motherboard contacts the stand-offs.

Please, if you have any other suggestions for me to diagnose this problem share them with me. Thanks.

I do not have time right now to run OCCT again, I will be doing that in a few hours.

 

[thechief73]

UPDATE: I am still failing OCCT tests, but I think I may have found the source of my problem: A failing CPU temprature sensor. During my last test run the temp went to 128C and stayed there, the test halted but the CPU was cool to the touch and the PC ran fine just as normal. A restart put it back to normal operation, but I think it is still out of wack, any tool I use to report CPU temp is of by 5c to 10c! from what I read directly from the BIOS at startup.

One other issue I noticed about a year back is also bothering me and I thought I would mention it because I reflashed my BIOS and it is still happening. My BIOS and any tool under the sun you can think of reports my CPU as being the Extream Edition which it is not I have a P4 650, and I cannot figure out why this is happening or how to fix it either.

 

[AllanCameron]

When all else defies explanation, I blame the PSU.

 

[thechief73]

UPDATE: I have passed a OCCT test!!! I re-flashed another BIOS version (3 down from the newest, and two down from the other I had tried, I also tried the very first but it cause my PC to be very slow and lack any of the later added BIOS refinments and features so that was not a viable option). I have also, discovered that the CPU voltage may have been set too low in an effort to minimize heat(as I had said earlier my CPU requires an unsually high amount of voltage to run to begin with), so I have returned this to an auto setting.

Thanks everyone for your suggestions, but I belive I have everything sorted out now. I guess only time and more tests will tell.

 

[thechief73]

FINAL UPDATE: I now have my CPU OC'd - 10% @ 3.74GHz and RAM @ 733MHz running smoothly without errors. Also I have since flashed my GPU Bios to one I custom edited with RBE(Radeon BIOS editor) with a slight OC and better fan profile. Everything is good, now just to get those tempsa little lower all around

 

[wolfram23]

Nice. I used RBE as well, it's great. Although, for me and I know from googling many others, the custom profiles don't work once windows boots up. Some wierd thing with the registry or something rather. But anyway, Afterburner does work for custom profiles so if you notice that they're not doing what you set, IMO just go back to Afterburner.

And yeah, temps you can only do so much. I mean, try to lower voltage but as you already have experienced, it can cause instabilities.

 

[thechief73]

Oh really, mine have been working perfectly. I used GPU-Z and Speedfan to create a log file while I am gaming and all seems to be right on the money. I can hear the fan ramp up with the increase in temp. I have downloaded Afterburner but have not yet tried it out, since everything appears to be working. Thanks for the tip though.

About the temps and voltage, I only have three CPU setting in my BIOS(Auto / 1.575 / 1.5875 / 1.6), I just left it on Auto to let the Mobo do the work for now and all is stable. As for the GPU, I really didnt want to mess with that as much, it seemed more risky. I dont want to brick my GPU with a bad flash and not be able to recover it or "blind flash" it back to normal. As for the GPU temps I am around ~55-62C max while gaming with high settings with 1080p res. Not bad, but I have read of others getting lower temps with this GPU, so always room for improvement to keep me busy. :lol:

 

[thechief73]

Best answer selected by thechief73.

 

[wolfram23]

Too bad you don't have a better mobo BIOS, but at least it's all working fine. Congrats