Can a faulty motherboard or overclocking damage RAM?

[mdg900]

I recently built a new PC, but it kept completely freezing every so often.

I ran memtest86+ and consistently got errors in the same place during test 6. (Stock settings). After I removed one of the sticks of RAM it passed all tests, so I assumed the RAM was faulty, and have contacted crucial for a replacement.

All seemed to be OK until the PC froze again today, although the FSB was overclocked to 333 MHz, so I'm guessing that was the reason for the freeze.

I'm wondering if its possible that the RAM was working originally and has been damaged by the motherboard or by me overclocking it, and if I should stop attempting to overclock my system.

Also, what's the best windows app to test RAM? I'm using prime95 at the moment, but this doesn't seem to pick up any problems.


Thanks in advance for your help. PC spec is below:

Core 2 Duo E6400
Abit AB9
2 x 1GB Crucial 533MHz
Radeon X1950XT
2 x IDE HDs
Antec P180

[Supershanks]

HDI Design Memtestwill check free windows memory. Has an advantage over memtest86 in that it's testing memory under xp enviroment, downside is it don't test as much memory as memtest. Memtest #8 is supposed to be the cloest test 2 an xp enviroment.

SP2004 Orthos Edition is specially designed for dual processors. Blend test both cpu & memory, Large tests memory more while small tests cpu more.

Have you tried adjusting memory voltage slightly & retesting.
If you have issues running 2 modules try switching over ^& rerunning, sometimes one module is weaker than another & swapping sorts it out.

Do both modules pass when run singly ??

if you've always run within recommended voltage then its doubtful you have done any damage.

luck

[Clunk]

I dont know much about the AB9, but if its anything like other c2d boards, you will have to manually enter all the mem timings and sub timings. I couldnt run at all with 2 different types of ram until i put the subtimings in properly, after that, it was fine

p.s. memtest is what you need to use, test 5.

[mdg900]

I've tried 1.8V and 2.0V and get the same error. (1.8V is the default value.)

The module I removed gives the same error even on its own.

Is there any significance in the fact that the error occurs at the same address every time?

I've tried setting the timing manually and that didn't help. (Although I left "Command Rate" and "Refresh Cycle Time" on auto as I'm not sure what they are.)

[Clunk]

so, one of the modules gives errors and one doesnt? does it still give errors if you put it in a different slot?...have you tried all the slots?

[mdg900]

so, one of the modules gives errors and one doesnt? does it still give errors if you put it in a different slot?...have you tried all the slots?

That's right. I haven't tried slot 2 or 4 yet though.

Both sticks in slot 1 and 3 - errors
Stick B in slot 1 - errors
Stick B in slot 3 - errors
Stick A in slot 1 - no errors

I just ran memtest86 for over an hour and all tests pass apart from test 6.

I'll try the rests of the slots now.

[even steven]

Funnily enough I have a similar problem. Got a pair of 1GB Corsair XMS2 Dominator PC2-6400C4 TwinX RAM modules, and when I slotted them into my new mobo (Abit AW9D-Max) the PC failed to post, stopping at code C1 (Detect memory). When I removed stick B from slot 3, the PC completed post and booted sucessfully, but A and B in any combination of slots failed to post, as did stick B in slot 1.

I figured I'd just got one dud RAM stick, but now I'm starting to wonder if it's a problem with the motherboard.

[Supershanks]

If the module fails on the settings Clunk gave you i'd say it was faulty & get an rma for the pair.

luck

[mdg900]

Just tried the rest of the slots.

If stick B is in the system on its own or in any combination with stick A then test 6 fails. Stick A on its own passes.

Sounds like its just a dodgy stick of RAM but are there any other BIOS settings I can try? Does any one know how I can find out what the "Command Rate" and "Refresh Cycle Time" are and if its worth entering them manually?

[Clunk]

Just RMA it mate, if its making errors now, its not going to be reliable a few weeks/months down the road.

[Supershanks]

Command rate isn't relevent on 975x/965 chipsets all memory runs at 2T. only the new nvidia 680i supports 1T.

Not sure where "Refresh Cycle Time" is on your mobo sorry.

Looking elswhere on the crucial sticks there supposed to run stock quiet easily at default voltages below crucials spec.

You might find this useful
Memtest86+ Guide/How To's
also
TCCD DDR600+ DFI nF4 Ultra/SLI guide - See Method although for tccd & AMD the methodology & explanation is sound.


luck

Edit:

Just RMA it mate, if its making errors now, its not going to be reliable a few weeks/months down the road

I'm with Klunk

[storms33]

this may be some years on, but i had simular problem.

kept getting random BSODs no matter what i was doing with different error codes all pointing towards memory. ran memtest 86+ v 2.11. stick 'A' came up with multiple errors in slot 1. took that out and put in stick 'B' (which was in slot 2). this came up with no errors at all after running 3 times.

so i left it in and got ready to return stick 'A'. less then 3 days later i've started getting the same BSOD errors. ran memtest again and now stick 'B' gets shed loads of errors.

am i right to assume that i need to now return both sticks AND the motherboard?

could updating the BIOs have any affect in this matter? i upped from F3 to F4a (a beta that has been out for months so i would assume is safe)

sed computer component retail supplier is going to think i'm taking the 'pi' double '55' after i've just returned 4 - THATS 4! - Seagate ST3500320AS 500GB Hard Drives (outcome of which i will never buy a seagate again! - thats another rant all together)

reguards to anyone who can help.