Dell Motherboard/Memory Failure... Need help plz!

[Anzy]

Hey guys, quick question, trying to figure out whats wrong with one of my Dell computers... Basically what happened is I started getting blue screens (Can't remember error code off the top of my head) but I ran the Dell Diagnostics and I was coming up with problems on DIMM_!, then I would do another test, no errors, then error on DIMM_3, etc etc... this basically happened with all the slots that have memory. Two of the sticks are identical DDR2 1 gig running dual channel. the other is a stock 256mb stick. Now I figured it may have been dirty contacts so I cleaned the board with compressed air and wiped down the contacts on the memory. Did a test, no errors, couple days later the blue screens start again. The only thing I changed on this particular computer was swapping out the stock PSU with a Cooler Master 400W (Here) and added a eVGA 6600GT 128 card... So I'm quite curious as I do not believe the memory is bad as this problem is happening to all the sticks (stock and what I added which was this...) Any advice would GREATLY be appreciated! Thanks!

 

[Lord Evermore]

Well first off, if you have 3 sticks of memory installed, it's unlikely you're running in dual-channel mode. Take out that dinky 256MB, as it's hindering your performance more than having that little bit more memory can help. (The upgrade of adding 1GB probably helped performance enough so that you wouldn't notice the drop to single-channel.)

Second, if the problem appears to have started when you changed something, even if it seems like they couldn't be causing a problem, change them back and see if it continues. You might have a flaky power supply, or a problem with the video card.

In the end, something may have been damaged while you were working on the system. A static discharge too small for you to feel is enough to damage components. The memory also may just have a compatibility problem with the mainboard; you didn't mention when you added those.

Run the diagnostic tests with only one stick of memory installed at a time.

You did make sure the stock ATX connector was the same as a standard PSU connector, right? It probably was, otherwise you'd likely have just seen a lot of rank-smelling smoke, but Dell I know makes some systems with a connector and mainboard where the pinouts are difference. Having like a 5V wire feeding a circuit that expects 12V probably wouldn't kill the system but probably also wouldn't boot.

 

[Anzy]

The computer in question is a Dell Dimension 5150 E510. I was under the impression that you could run 3 sticks while running 2 in dual channel. If this isn't the case, sure np 256mb won't be anything to fuss over. These problems started when my friend was over testing my gpu on his computer... he might have screwed with something, if it is a GPU issue though, is there anyway to narrow it down to that? I will test one stick at a time, but still it seems that the diagnostics pass sometimes, and sometimes that's not the case. The problem seems to jump from one DIMM to the next without any tinkering. Sometimes 1 will fail, 3.. 2... etc, all without no tinkering with the components. Is there a good third party diagnotics application I can run on a seperate computer to test memory. I can try just running the sticks in question through some sort of benchmarking to test for any problems they may have... If they don't pass there then I would RMA (Though I'm not sure I still have the packaging for Newegg =( ) Anything further on this would be greatly apprecated! And thank you Evermore for your help thus far! =)

 

[Anzy]

Bump, still unresolved issue

 

[LiLithTecH]

Have you checked the 'Event log' for the specific error?
(most likely the infamous "Driver IRQL not Less than or Equal")

 

[Anzy]

Just checked the event viewer, didn't find anything other than the blue screen dumps... 0x0000001a, 0x1000007f, and 0x100000d1

 

[LiLithTecH]

Check the Diagnostic leds's on the back of the case.
If there is a problem with one or more of the led's
being non-lit, blinking, or amber, run the DELL Diagnostics.
(on one of the CD's included when PC purchased)

If all the leds are on, try resetting the CMOS.
(jumper next to battery - remember to disconnect the power cord)


Why did you think you needed to replace the PSU?
The original PSU from Dell is better suited for your new
video card than the new one you replaced it with.

The +12v on your original PSU is 34A (combined +12va & +12vb)
The +12v on the new PSU is 19A

 

[Anzy]

I will check those led's asap, but I'm kind of curious about what you said about the PSU, I'm not completely seasoned into the power requirements of the components but when I did that, I was most (falsely maybe?) about the watt power rating, thinking more would be better since I had purchased the video card, which is why I was concerned about it... so if I switched pack to the original it would be better? Thanks a ton Lil!

 

[LiLithTecH]

Compare the 2 PSU's side by side and I bet you see what I mean.

My guess you were following the recommendation from eVGA about
requiring 350W PSU.

Not all PSU's are created equal.

 

[Anzy]

I finally ran memtest for over 14 hours and got 3 errors...

Tst Pass Failing Address Good Bad Err-Bits Count

6 1 00080091f34 - 2048.9MB 00004000 0004001 0000001 1
5 16 00050c64900 - 1292.2MB 00000000 0002000 0002000 1
5 16 00051c648e0 - 1308.2MB 00000000 0002000 0002000 1


I really have no idea how to read these results or what it means, this is after I switched back to the stock PSU that came with the computer... Help?

 

[LiLithTecH]

What memory modules (brand-model) are you using?

Have you checked the BIOS to see what the system is seeing
the modules as (settings)?

From the errors you are receiving in Memtest, and it being on the 5th Pass,
it sure seems the timings of the memory may be to agressive for the system.

 

[Anzy]

The exact memory I purchased was this... I didn't see anything suspicious looking in the bios that would point to some kind of problem, though what your saying sounds pretty interesting...

 

[Anzy]

Is it really possible this particular memory's timings are too much for this system?