possible mem conflicts

Joined
Jan 25, 2006
Messages
57
Hi there
I am having some problems with my mem, or at least i think i am...!

Setup is as follows xp athlon 2600 and x3 256 ddr 400 (I think) sdram modules
I have heard some A socket systems can fail the memtest 86 if the settings aren’t right however.

When the memory is taxed on this system the system crashes and will automatically reboot, occasionally it will settle into an infinite cycle of rebooting once a certain point in the booting process is reached. I have remedied this problem in the past by chopping down the latency speed. In fact a solid was to stop this from happening is to under clock my FSB at a paltry speed of 100 MHz there is no problem outside of appalling performance.
I recently got some extra ddr 400 modules but I can’t run them at 200 MHz speeds, the retailer may have skimped me on a ddr 400 and sent me a 333, or lower or something so my first question is

How do I isolate what the clock speed on each individual module is? I plan to make sure each is a 400 to make sure conflicting latencies are not a problem. Are conflicting module clock speeds to blame if that is the problem.

Or have I another problem such as poor voltage setting on the modules?

Sorry I haven’t formed this question so good but if you ask me the right question im sure we can figure out what the hell I have screwed up!

I’d appreciate any and all help and recommendations I want my 2600 baaaccck
 
Joined
Jan 25, 2006
Messages
57
you mean just run it in mem test or try similar operations as i would normally, with one module untill i find the felonious stick? if so have tried, slow but stable ;)
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
so all sticks work on their own, at ddr400 like you want?

also, why not just look at the sticks and find the rated specs for each?
 
Joined
Jan 25, 2006
Messages
57
some bad news has transpired, found our that one is 333 the other is 400 howerver the bad news is that when i tried to test the sticks one a time the system was so unstable it couldnt even boot up much past the mem test!

so i tried all 3 sticks in the first slot no luck and two in the second slot all unstable at the lowest possible bios clock speeds.
i just have the two ddr400 in dimm slots 1 and 3 to allow for dual channel capability.
seems fine at the moment but i cant say how long it will be stable or what kind of preformance i will even be able to get
this looks like a failing mobo to me... thoughts???? :confused:
i was planning on a new 939 system but i not this soon
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
take a close look at the mobo's manual to see how you should populate dimm slots. i've seen boards that like the outermost slots filled first, not the innermost
 
Joined
Jan 25, 2006
Messages
57
hey (cf)Eclipse appreciating your input here!
i have been on to the mobo manual and im convinced it is no longer a mem prob. i have the FSB at 100 and the mem freq at 100 also, ran mem test with a 266 ddr400 in slots 1 and 3 dual channel, and it was solid, stable no probs. so i ran the mem at 200 mhz with the FSB at 100 and again solid as, pass with no spontainous reboots, so i jacked the FSB up gradually starting at 133 and mem test fails i.e full reboot at the begining of test 7.

have i got a bung setting somewhere possibly in the bios, or is the dual channel tripping the system?
i think the FSB setting is rebooting the system but isolating the problem could take some time.
I might run slots one and two and try upping the FSB to see what happens, what da ya think?
hmmmmmm :confused:
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
when you change the fsb, what is it changing the cpu speed to?

that does seem kinda odd though, almost as if it's the chipset having issues at those speeds
 
Joined
Jan 25, 2006
Messages
57
default setting of the FSB at 100mhz has the processor doing about 1.25 gigs up to 133mhz the processor is about 1.6 or so
do you think i may need to investigate the chipset? i also changed the CPU clock select jumper from 166 to auto detect 100/133 thought it may have made the system stable but nope!
 
Joined
Jan 25, 2006
Messages
57
yeah cf i was lucky enough to find someone at my local university to do it, didnt think i would but there you go.
how would you suggest i go about it?
run mem test, or maybe test the modules in different slots??
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
run memtest to check for errors.. though it won't fix any other problems you have. trying different slots should be one of the first things you try ;)
 
Joined
Jan 25, 2006
Messages
57
well cf looks like a bit of bad news i ran all mem through a p4 processor the fsb was at 199 mhz.
i ran the ddr 400 in slot one
then in slot one and three at dual channel interleaved
the i added the ddr 333 in slot two for assymetrical dual channel and it all passed with no problems.
the ram was running @ 159 mhz ddr318 according to mem test if i tried for those specs on my machine i think it would crap itself. so i guess that leaves the fsb or processor.
both of which i cannot be boned replacing, is there an easy way to find out which it is?
 

(cf)Eclipse

Freelance Overclocker
Joined
Feb 18, 2003
Messages
30,027
well... does it work at that lowered speed? i'm not sure if i can really figure out what the issue is with what you wrote :(
 
Top