PDA

View Full Version : stability problems


(v)estra
07-27-03, 09:58 PM
ok could my ram cause stability problems i cant even pass the 1st test of prime95 at stock speeds

Yodums
07-27-03, 10:10 PM
What is the error?

Try using memtest86 to verify your memory isn't causing errors. Check your bios settings as well.

(v)estra
07-27-03, 10:17 PM
fatal error rounding was .4879989624 expected .4
hardware failure detected,consult stess.txt
thats at stock 1.44 on my 1700+ :(

Yodums
07-27-03, 10:19 PM
Well, I highly suggest the memtest to make sure... Make sure you research to make sure there's nothing weird about the board. I remember I had memory errors at stock speed with my ol' KR7A and then I found out I had to overclock it a tad because the memory traces were too long, therefore it would create some errors.

So, try searching the Abit Forums for that :)

But leaving the memory overnight to examine wouldn't hurt either.

(v)estra
07-27-03, 10:22 PM
just ran memtest no errors... but still no matter where i set it i get errors on prime 95... could it be a cpu/ram conflict?

Yodums
07-27-03, 10:24 PM
Using the latest version of Prime? I really doubt it could be the CPU. If it boots ok, it should run fine, unless there are chips. Mmmm, this one has got me puzzled...

c627627
07-27-03, 10:35 PM
nForce2 mobos have a PCI lock, your VIA mobo does not.
Don't you have to have 166 FSB for your /5 divider to kick in?
133 to 165 is /4 divider isn't it and if so 165/ 4 = 41 (How are your hard drives still alive that far out of 33 specs?)

There's probably more to it.... of course there is... you listed your dead system confusing the heck out of people trying to help you ;)

Please be kind, remove dead wood from your sig. 165 is probably what killed it.

instead of the dead stuff post one of the most important parts of your system: the power supply. Watts don't mean much, brands do, it may be the PSU.

(v)estra
07-27-03, 10:53 PM
hope my new sig helps

(v)estra
07-27-03, 10:54 PM
ooh forgot 1 thing in my sig the steppings on my cpu are jiuhb dlt3c 0310 i think thats all

(v)estra
07-27-03, 11:08 PM
ok i reset my cmos jumper and set my milti to 7 and fsb to 200 and it passes prime 95... so i know it can be stable at 200 fsb now all i gotta do is figure out why its not stable at anythinh higher than 1500 mhz at 200 fsb...

(v)estra
07-27-03, 11:09 PM
heres a thought could my psu be going out? i know the electricity in this house is outa wack and has fried psu's b4 yes its on a serge protector but i have already been through 5 psu's

(v)estra
07-27-03, 11:17 PM
odd ball question but can too much v-core cause instability?

Constantinos
07-27-03, 11:26 PM
MAybe its a software problem. What OS are you using?

(v)estra
07-27-03, 11:30 PM
xp pro its wierd though its running stable now at 2000 mhz 10X200 ran 10 tests on prime b4 erroring gonna up the v-core a little and then up the multi... maby i was over volting it?

(v)estra
07-27-03, 11:35 PM
well running prime at 2100 mhz @1.5 v-core (my mb under volts its set at 1.575 in bios) and stable as hell go figure

(v)estra
07-27-03, 11:39 PM
testing 2300 mhz at 1.6 volts (1.65 bios) see what happens now

(v)estra
07-27-03, 11:40 PM
dang no luck instant failure gonna up the v-core to 1.7 and see brb

Constantinos
07-27-03, 11:46 PM
IT seems your system has a problem with the voltage. It may be that their is too much heat and its causing you to lock up.

(v)estra
07-27-03, 11:47 PM
i doubt its heat 34 idle 42 load @1.8v core any other possibilities?

(v)estra
07-27-03, 11:50 PM
i take that back it seems that it runs a hella lot hotter at higher speeds at 1.68 v-core 2300 mhz 50*load this hs/f sux... o well theres always the slk800 i am getting from altec i cant wait for that :) o well 2200 mhz @ 1.6v-core works for me

Tyranos
07-28-03, 02:13 AM
There was like 5 minutes between the post before you said you ran memtest, and the one saying you did. :P Did you only run it for 5? It needs quite a bit of time to finish. After testing a crapload of ram, I either got crashes or errors in the first 15mins, or in the last hour or so. In memtest, press C, 2, 5, type 10 press enter, and 0. Test #10 was the one that always caught the errors if the others didn't.

(v)estra
07-28-03, 08:42 AM
im gonna run mem test all day while im at work hopefully i come home to some errors