If memory passes memtest is it always good?

Semyaza · Dec 27, 2004

I have been trying to get rid of a problem with my computer for a while now. It keeps getting errors seemingly random but always with one thing in common heavy graphic usage. My memory has passed memtest for loooong periods of time and my vid card shows no artifacts or problems when testing it with ati tool. I have everything at stock right now and still get the errors. My cpu, vid, mem all stock but I still drop out of games pretty often.
The main problem I am having is I can't swap parts to test anything cause I don't have the cash, are there any programs out there to test for errors like these? Oh and I took a dvm to my system all the lines are good. And what heat sensors that are built in show good temps.

cetoole · Dec 27, 2004

Check your PSU rails with a dmm.

Semyaza · Dec 28, 2004

cetoole said:
Check your PSU rails with a dmm.

I did I said that in my above post

Know Nuttin · Dec 28, 2004

have you tested your CPU? use Prime95 or other CPU intensive apps to see if it will reboot.

David · Dec 28, 2004

If its possible, you might want to try underclocking your card and see if you still get crashes, in case it is a graphics issue as you said crashes happen during periods where the graphics card is used heavily.

Also make sure you have all the latest windows updates, patches for your games, and drivers.

Semyaza · Dec 28, 2004

Know Nuttin said:
have you tested your CPU? use Prime95 or other CPU intensive apps to see if it will reboot.

I ran cpuburn last night ok, Ill turn on prime when I get off the comp today and see if its alright.

Semyaza · Dec 28, 2004

DOH, turned core and mem down by 10 and doom3 crashed withing 5-10mins. >_<

Semyaza · Dec 28, 2004

how do I print out my bios settings?

I looked up my stock memory settings just to be sure and I am confused about one number, it shows up as 2.5-3-3-7-1 @ 2.6 volts what is the one for?

aNTiChRisT · Dec 28, 2004

You think it could possibly be virus' etc?

If you have a blank HDD try a format and see if you get the problems.

~t0m

Semyaza · Dec 28, 2004

aNTiChRisT said:
You think it could possibly be virus' etc?

If you have a blank HDD try a format and see if you get the problems.

~t0m

I havent had a virus show up in a scan in like a month, I have been updating my antivirus regular as well.

Semyaza · Dec 28, 2004

GRRRRRRRRRR I just took out my mem and looked at it I was using the wrong timings, I was using KHX3200KA settings when mine is 3200K zzzzzz. But now my computer can't seem to accept the change. Every time I set my cas to 3.0 from 2.5 or 2.0 (Tried both ways) it beeps and I have to reset cmos >_<>_< after doing that 6xs its sitting at 2.5 instead of 3 right now. What gives with that? Also I have numbers I dont know the times for I have award bios and it lets me set all the 3-3-3-8 numbers (not -1 for some reason) and 2 others that are at stock 13 and 15 what is that about?

Semyaza · Dec 29, 2004

bleagh said:
It sounds like you might have Kingston modules with BH-5 chips.
These modules often have problems at 2.5-2.6 volts. But usually work great when the voltage is bumped to 2.8 volts or higher (2.8 volts seems to be the sweet spot).

BH-5 based RAM is usually very good RAM, but BH-5 likes voltage. So just bump the voltage up to 2.8 volts and see how it goes. BH-5 often won't run at 200 FSB or higher with 2.5-2.6 volts. But with 2.8 volts or more it should do 200 FSB (400 DDR) with ease. And it should run at good timings (2-2-2-6) also. BH-5 usually won't run at CAS 3.0, but should be good at CAS 2 or CAS 2.5 with enough volts (I suggest 2.8 volts and 2-2-2-6 timings at 200 FSB).

Don't worry about pushing the voltage as this is normal for BH-5. BH-5 can handle up to 2.9 volts with ease (and up to 3.5 volts with good cooling).

Yeah I made sure to get teh bh5 for those reasons I never thought it have a problem being loose or undervolted, ty I will give it a try.

jcw122 · Dec 29, 2004

that 1 at the end of ur timings problably means ur running 1T (2T is another setting for that)...1T though had known to be better...its the command rate thingy for ur mem. And wow thats really weird how if u LOOSEN ur CAS latency u get a problems...well i guess u should try upping ur vDIMM and report back...im really interested in this.

Semyaza · Dec 30, 2004

K 2.8 didnt do it you say it can go to 3.5 safely? Ill keep pumping it up and testing till then Ill report if I find anything out.

Up to 2.9, wrote down my exact settings as well.
T(RAS) - 6
T(RCD-Read) - 2
T(RCD-Write) - 2
T(RP) - 2
CAS - 2.0
T(RC) - 13
T(RFC) - 15
Super Bypass - Enabled
Sync Mode Memory Bypass - Automatic
Data Scavenged Rate - Fast
Command Per Clock - Enabled
DIMM 1 Driving Strenght - Automatic
DIMM 2 Driving Strenght - Automatic
DIMM 3 Driving Strenght - Automatic
DIMM 1 Slew Rate - Automatic
DIMM 2 Slew Rate - Automatic
DIMM 3 Slew Rate - Automatic

Errors found at 2.8, 2.9, 3.0 and 3.1 volts now.
I will keep walking it up to 3.5 then drop to stock and see how it acts with t(ras) set to 11, I read on the forums that seems to be a magic number for n-force2.

Semyaza · Dec 30, 2004

Zzzzzzz..... I can only go up to 3.3 volts. Next I'll try that tras at 11 thing anyone else have any ideas on what else to do?

I have noticed I get more errors the lower volts I give it, found an error in like 30 seconds at 2.7 but it took near 5 mins at 3.3 and there was less errors.

Semyaza · Dec 30, 2004

Hrm....... Well since its a kit I guess Ill just pop one out and test all of it again maybe its just one of the sticks.

Silmatharien · Dec 30, 2004

Semyaza said:
I have been trying to get rid of a problem with my computer for a while now. It keeps getting errors seemingly random but always with one thing in common heavy graphic usage. My memory has passed memtest for loooong periods of time and my vid card shows no artifacts or problems when testing it with ati tool. I have everything at stock right now and still get the errors. My cpu, vid, mem all stock but I still drop out of games pretty often.
The main problem I am having is I can't swap parts to test anything cause I don't have the cash, are there any programs out there to test for errors like these? Oh and I took a dvm to my system all the lines are good. And what heat sensors that are built in show good temps.

Don't always rely on ATi Tool. Run games/benchmarks and see for yourself.

Semyaza · Dec 30, 2004

Silmatharien said:
Don't always rely on ATi Tool. Run games/benchmarks and see for yourself.

Well I know something is wrong in the first place because I am having trouble playing games, alot of the times it just drops out. Even old games like battlefeild,
its been doing it a while it just has never really been a problem, DOOM3 however taxes my system good so it drops out ALOT more like every 1-2mins.

Just ran through one complete memtest with no errors current settings,
T(RAS) - 11
T(RCD-Read) - 3
T(RCD-Write) - 3
T(RP) - 3
CAS - 2.0
T(RC) - 13
T(RFC) - 15
@3.3 volts.

hope it keeps working la.

Semyaza · Dec 31, 2004

bleagh said:
Um what speeds are you testing at?
If you are testing with the FSB at a 200 clock rate (400DDR), then you really shouldn't need more than 2.8 volts. While it's not uncommon for BH-5 to have problems at 200 with 2.5-2.6 volts, it really should work okay at 2.8 volts. People usually only push the voltage beyond 2.8 volts when they are running at speeds above 200 MHz.

Testing each module is a good idea. When testing just one stick then you should be okay with 'Command Per Clock' enabled (that should be a 1T command rate I think). But when testing both modules together you may need to set 'Command Per Clock' to disabled (2T command rate).

Well I jumped up to 3.3 cause it seemed to run stabler the more power I piled on. It ran 50 full mem passes and prime for 10 hours @ pri 10 at the above settings so now Ill back it down and see if it will stay stable at a lower power.

felinusz · Dec 31, 2004

Up to 2.9, wrote down my exact settings as well.
T(RAS) - 6
T(RCD-Read) - 2
T(RCD-Write) - 2
T(RP) - 2
CAS - 2.0
T(RC) - 13
T(RFC) - 15
Super Bypass - Enabled
Sync Mode Memory Bypass - Automatic
Data Scavenged Rate - Fast
Command Per Clock - Enabled
DIMM 1 Driving Strenght - Automatic
DIMM 2 Driving Strenght - Automatic
DIMM 3 Driving Strenght - Automatic
DIMM 1 Slew Rate - Automatic
DIMM 2 Slew Rate - Automatic
DIMM 3 Slew Rate - Automatic

Those timings all look fine, except for Command Per Clock (CPC) being enabled. I had a 2 X 512 Meg set of Mushkin BH-5 that wouldn't run stably at 200 MHz with CPC enabled (1T), regardless of voltage, but that did 236 MHz for me with 2.95V, and CPC disabled.

Try disabling CPC, and see if that fixes your problem, it is common for 2 X 512 Meg sets of RAM to have trouble running CPC enabled; typically only 2 X 256 Meg sets, or single sticks, will run nicely with CPC enabled.

If memory passes memtest is it always good?

Member

Member

Member

Senior Member

Forums Super Moderator

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Senior Overclocking Magus

Similar threads