• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Memtest 86+ question

paperfist

paperfist

Diamond Member
I have a Dell Dimension C521.

It's been locking (hard lock/must power down) up after general usage all the time.

All 4 banks are filled with RAM, 2 stock and 2 Kingston 1GB sticks for a total of 4GBs.

I took out the 2 Kingston sticks and left the stock 2 and it greatly improved the computer's operation with it not locking up at all. I put the other 2 sticks back in and it locks.

So running memtest 86+ produces errors with any combination of RAM (2 sticks or more). If I put any 1 stick in any bank it produces 0 errors. Anytime I add a 2nd stick it's errors galore.

Does this mean the motherboard is shot? There was a problem with the GPU cable being loose to the point it would (screws were too short to make a solid connection from cable to card) often boot with no video. I don't know if my sister shorted something out because of this or it's just a RAM/MB issue.

Thanks for the help.
 
Got a spare PSU to test with? If 1 stick works in any slot, I would put my money on a power issue or mobo failing.
 
ch33zw1z said:
Got a spare PSU to test with? If 1 stick works in any slot, I would put my money on a power issue or mobo failing.
Click to expand...

Interesting take. Yes, I do have a spare PSU or 2 but not sure if they'll hook up with a Dell setup. I will give it a whirl though, thanks.
 
paperfist said:
I have a Dell Dimension C521.

It's been locking (hard lock/must power down) up after general usage all the time.

All 4 banks are filled with RAM, 2 stock and 2 Kingston 1GB sticks for a total of 4GBs.

I took out the 2 Kingston sticks and left the stock 2 and it greatly improved the computer's operation with it not locking up at all. I put the other 2 sticks back in and it locks.

So running memtest 86+ produces errors with any combination of RAM (2 sticks or more). If I put any 1 stick in any bank it produces 0 errors. Anytime I add a 2nd stick it's errors galore.

Does this mean the motherboard is shot? There was a problem with the GPU cable being loose to the point it would (screws were too short to make a solid connection from cable to card) often boot with no video. I don't know if my sister shorted something out because of this or it's just a RAM/MB issue.

Thanks for the help.
Click to expand...
Highly doubt this is a PSU issue. This is most likely a case of the motherboard not liking the RAM.
This is also why most of the time, they tell you to use the exact same kind of RAM in all slots, since mixing DIMMs can cause the issue that you discribe.
If 1 stick (no matter the brand) always works (no matter which slot you use), and the matched pair of that RAM causes errors, then it could be bad RAM, or it could be a short on the motherboard, or could be a bad slot, or bad timings or voltage.
 
Elixer said:
Highly doubt this is a PSU issue. This is most likely a case of the motherboard not liking the RAM.
This is also why most of the time, they tell you to use the exact same kind of RAM in all slots, since mixing DIMMs can cause the issue that you discribe.
If 1 stick (no matter the brand) always works (no matter which slot you use), and the matched pair of that RAM causes errors, then it could be bad RAM, or it could be a short on the motherboard, or could be a bad slot, or bad timings or voltage.
Click to expand...

The 2 sticks that came with the machine give the same errors. 1 stick in any slot works fine, 2 sticks produce errors. The other set is a matched pair of the same specs, but a different manufacturer.

I'm going to try out the PSU today. If that doesn't work I'll have to assume it's the MB and since the machine is around 5 years old now I doubt I can find a replacement.
 
Just a few questions here:

How is the dust build up, have you blown out the system with compressed air lately ?
Is the system overclocked in any way ?
what are your memory timings set to ?

The first question is getting at a possible overheating situation
The second could also be overheating, or various other factors.
The third and most likely issue would be the mem timings are too tight, and/or more mem voltage is needed.
 
Last edited:
Soulkeeper said:
Just a few questions here:

How is the dust build up, have you blown out the system with compressed air lately ?
Is the system overclocked in any way ?
what are your memory timings set to ?

The first question is getting at a possible overheating situation
The second could also be overheating, or various other factors.
The third and most likely issue would be the mem timings are too tight, and/or more mem voltage is needed.
Click to expand...

The insides are virtually dust free.
No overclocking, it's a bone stock Dell.
hmm haven't checked the temps, but during testing the side panel has been off and the ambient room temp is 65.
Haven't checked the memory settings as again it's a bone stock Dell and none of the settings have been changed.

I will check the temps and timings though. Thanks for the suggestions.
 
If you havn't checked the latencies you definitely should

loosen them one notch and there is a high chance it could stabilize your system.
Many chipsets/memory controllers require looser latencies and/or more voltage when you install 4 dimms instead of 2.

now, the fact that you have a dell is another issue entirely :/
hopefully your bios isn't overly gimped
 
paperfist: How long has this setup been running with all 4 sticks, was it ever stable?

I have never had the option to change timings or voltages for OEM boards such as Dell's, Gateways, HP, etc...

It either works, or it doesn't.

Have you confirmed the installation order for the matched pairs? ie, 1 pair is slots 1+3, 2nd pair is 2+4, this is just an example, it could be different.
 
ch33zw1z said:
paperfist: How long has this setup been running with all 4 sticks, was it ever stable?

I have never had the option to change timings or voltages for OEM boards such as Dell's, Gateways, HP, etc...

It either works, or it doesn't.

Have you confirmed the installation order for the matched pairs? ie, 1 pair is slots 1+3, 2nd pair is 2+4, this is just an example, it could be different.
Click to expand...

I added the 2 other sticks over a year ago and it was stable until a few weeks ago. I haven't checked yet on being able to change the timings, but I don't think I can either.

Yes, the matched pairs are color coded and the MB is labeled, they are definitely in the right place.
 
So I overlooked a couple of things. The Dell uses that shroud over the CPU for cooling and it butts up to the front of the case. That was packed with fur so the CPU probably wasn't breathing well. I cleaned that out. Strange that the rest of the case is pretty dust free. Then I took a look at the CPU itself and the thermal paste pretty much just crumbled. I never seen that before. So I cleaned it up and applied new thermal paste. I used compressed air and blew out the DIMMS, all the other sockets, chipset cooler, PSU and GPU heatsink.

Installed the original 2 sticks in the appropriate slots and entered into BIOS. There's no way to alter the memory settings on this DELL so I exited out and into Memtest 86+

Before the first pass was even over it had some errors.

So I got out of there and went into Windows. I installed HWmonitor and an old 3D Mark test. I ran through that and the highest temps I saw was 32C on the CPU which is an AMD X2 3800+.

-----

My next question is does the GPU have an effect on Memtest86+?

For kicks I took out the GPU and setup the onboard video and rebooted into Memtest86+. It's only been 1 pass so far, but no errors yet unlike the first pass with the GPU installed.
 
It honestly sounds like your motherboard is having some sort of hardware failure. Especially if the 2 sticks of the stock ram cause errors (this eliminates bios configuration issues). I personally think you are SOL.

The GPU shouldn't be causing issues on a memtest.
 
What type of GPU is it? If it takes additional power, then how much? A GPU should never cause errors on memtest. Try that extra PSU yet?
 
OVerLoRDI said:
It honestly sounds like your motherboard is having some sort of hardware failure. Especially if the 2 sticks of the stock ram cause errors (this eliminates bios configuration issues). I personally think you are SOL.

The GPU shouldn't be causing issues on a memtest.
Click to expand...

Well with the GPU out and 2 stock sticks of RAM it made it through 9 passes with no errors.

I now have it running with all 4 sticks.

ch33zw1z said:
What type of GPU is it? If it takes additional power, then how much? A GPU should never cause errors on memtest. Try that extra PSU yet?
Click to expand...

I can't check right now, but it's also a stock Dell card. It doesn't take additional power.

I didn't try out the PSU yet, but I'll give it ago after I run Memtest again.
 
Last edited:
ch33zw1z said:
If the GPU changes the symptom, it's more than likely the mobo. But might as well be thorough 🙂
Click to expand...

Ok, I took the thorough route 😉

With the GPU out and all 4 DIMMs in it ran 8 passes without any errors.

Then I swapped out the PSU with another one and with all 4 DIMMs plus GPU it ran 8 passes without errors.

So does that mean my PSU is shot or the MB?
 
paperfist said:
Ok, I took the thorough route 😉

With the GPU out and all 4 DIMMs in it ran 8 passes without any errors.

Then I swapped out the PSU with another one and with all 4 DIMMs plus GPU it ran 8 passes without errors.

So does that mean my PSU is shot or the MB?
Click to expand...


sounds like you found the problem
yep, psu

good job
 
Soulkeeper said:
sounds like you found the problem
yep, psu

good job
Click to expand...

Fark right at the finish line errors start to rear their ugly head!

After the 8th pass w/GPU,PSU,4 DIMMs I'm starting to get memtest errors. I didn't run the test without the video card past the 8th pass so I can't compare the two.

How many passes do you need to run and after so many is it diminishing returns?
 
One last approach would be to look at all the capacitors on the motherboard. If some a buldged or domed, replace them. Those can cause the symptopms you are having.
 
paperfist said:
Fark right at the finish line errors start to rear their ugly head!

After the 8th pass w/GPU,PSU,4 DIMMs I'm starting to get memtest errors. I didn't run the test without the video card past the 8th pass so I can't compare the two.

How many passes do you need to run and after so many is it diminishing returns?
Click to expand...

As many as you feel comfortable with. I run it for 24 hours to test oc's, 8 passes isn't bad, just put it back to the first config and run it again.
 
Compman55 said:
One last approach would be to look at all the capacitors on the motherboard. If some a buldged or domed, replace them. Those can cause the symptopms you are having.
Click to expand...

That's actually something I took a look at yesterday when I had it all apart. I didn't notice any leaking or bulging.

ch33zw1z said:
As many as you feel comfortable with. I run it for 24 hours to test oc's, 8 passes isn't bad, just put it back to the first config and run it again.
Click to expand...

Thanks, I'll re-run. The screen started filling up with errors on pass 8, 9 and part of 10 when I shut it down.

I'm curious if any one has the time to explain how memtest translates into real world usage? Is it possible that the couple of bits that error out on may never get used by the system? Obviously if your using all 4 GB you'll have issues, but I'm not sure a system like this ever sees that.
 
Last edited:
paperfist said:
The screen started filling up with errors on pass 8, 9 and part of 10 when I shut it down.
Click to expand...
How many errors? My bad stick had 5000+ errors on tests 7 and 8 combined. It managed to boot a few times, but programs crashed constantly.

If it made 7 passes with no errors, the problem could be affected by heat, and you might get away with it in normal use. It seems like a bad idea to use an unstable system deliberately.
 
Dessert Tears said:
How many errors? My bad stick had 5000+ errors on tests 7 and 8 combined. It managed to boot a few times, but programs crashed constantly.

If it made 7 passes with no errors, the problem could be affected by heat, and you might get away with it in normal use. It seems like a bad idea to use an unstable system deliberately.
Click to expand...

It was around 10 errors at the point I stopped testing. I always though any errors where bad. The system is usable, but it does hard lock when you run certain programs, it also locks while trying to backup the drive.

I'm on 11 passes now (not quite 24 hours) with 0 errors with the stock PSU, 4 DIMMs and no video card. I wish I had a video card that would fit to test it with.
 
paperfist said:
It was around 10 errors at the point I stopped testing. I always though any errors where bad. The system is usable, but it does hard lock when you run certain programs, it also locks while trying to backup the drive.

I'm on 11 passes now (not quite 24 hours) with 0 errors with the stock PSU, 4 DIMMs and no video card. I wish I had a video card that would fit to test it with.
Click to expand...

What type of slot is it? a bus powered PCI-E VGA can be had for $20 or so...
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top