A7V133 unstable at 133

[kremington]

Hi,
Here's a challenge... I'm having an ongoing battle solving a lock-up problem w/ my A7V133 based system. I'm positive I've narrowed the trouble down to either the RAM or the m/b (arrrgh!). I've wasted SO much time trying to solve this on my own. Any suggestions would be gratefully welcome:

Problem:
System locks up with BSD under WinXP Pro or Win2K Pro (SP2) randomly - but most predictably when playing Return To Castle Wolfenstein. I have tried re-formatting my HD repeatedly and performing clean installs of either OS, but no joy. On a couple occasions I even got BSD while installing Win2K just as it got to the graphical stage of the install process - no drivers even loaded at this point!


Attempted Solutions:
At first I thought is was a possibe driver issue related to conflicting versions of VIA 4-in-1 and nVidia drivers. I tried multiple combinations, but needless to say, not a single one solved the problem. I then swapped out the memory from another system, but it is the same brand/part# of RAM - no change. I finally managed to get the damn thing to run with solid stablity by loading the BIOS defaults and de-tuning my FSB/Memory frequency to 100MHz (RATS!). I also manually set the vCore frequeny to 1.75 (ASUS Probe 2.15 still reads it as 1.80 tho).

I spent the extra $$$ on the m/b and RAM to get the extra performance of 133MHz, so naturally, I'd like to find a way to make this b*tch work. Sorry to be so wordy. Here's my current config:

System:
ASUS A7V133 (BIOS ver 1.007a)
AMD 1.2MHz Athlon (266) w/ a serious 100% Cu HS/Fan
2x256MB Crucial PC133 SDRAM
Maxtor ATA-100 7200RPM 40GB.
Peripherals:
AGP: Hercules GeForce3 Ti 200
PCI Slot 1: Open
PCI Slot 2: Open
PCI SLot 3: SB Live! Value
PCI Slot 4: Open
PCI Slot 5: 3COM NIC

Thanx for ne help!
KR

 

[lhgpoobaa]

To test your ram i strongly reccomend u find and download MEMTEST86.
let it run a couple of times to be certain. its extreemly reliable.

Have you considered that your Powersupply (PSU)
is not producing enough juice... or unreliable voltages? that may account for it locking up often under load.

And when u underclocked... was with or without modifying the multiplier?
i.e. were u running 100x9 = 900 or 100x12=1200?
whats the Brand and rating of your PSU?

*waiting for your reply*
Master poo the helpful



The lack of thermal protection on Athlon's is cunning way to stop morons from using AMD.

 

[AndrewT]

Hey we got the same board, almost the same problems. Except mine does stuff like files go bad on drive, it decide to re-install something at boot that was already installed, freeze (no sound loop). Got the same problems with my A7PRO board, except the screwing with files. Getting back the pro this week, going to send the a7v133 in this week. Got them replaced a few times already, once it was my fault, bad flash of bios.

I also did everything possible, removed pci cards, changed vid card, ram... nothing helped. My current a7v133 was problem free for a minute or so after first boot. LOL Will see what happen when they fix it.

..............................................................
I forgot, it also does stuff like everything I try to start gives an error and have to close it, and errors in unknown. :lol: It's not too bad after 2-3 reboots, when I play quake I get random problems like the map doesn't match with the one on server, but it's all good after reboot. It's a friggin pain on the a$$, when I try to set it to anything other than 133x10.5 it show 1050 MHz (even on 100x12.5 it show 1050MHz). Sometimes I feel like opening the case and giving it a golden shower.

 

[kremington]

LHGPooBaa,
Thanks for the great tip on the MEMTEST-86 utility! I straightaway downloaded v. 2.8 and started running some tests. First let me fill in the blanks in response to your post.

P/S is a generic 300W, but it seems to be fairly reliable if ASUS Probe 2.15 is any indicator. Probe of course allows me to monitor the main system voltages (among other parameters), and they always hold flat & steady very close to the ideal values. And I did indeed adjust the cpu multiplier to maintain 1200MHz (12x100).

MEMTEST-86 did reveal a problem tho! Test 5 [Block move, 64 moves, cached] got ~95% through and WHAM... nothing but errors scrolling up te screen. No other tests revealed errors. I did read at the MEMTEST-86 website that this is not uncommon for Athlon systems. They say more conservative memory settings or better quality RAM might be required - DOH! Currently my "high quality" Crucial CAS-L2 PC133 RAM is as about as de-tuned as it can be - using SPD, which drops it to CAS-L3! I guess I can now at least make the error appear consistently, but I'm still left scratching my head - is it the m/b or the RAM? Are there other memory settings I can tweak?

Again, thanks for your kind insight.
KR

 

[lhgpoobaa]

good so see u can now see some evidence of trouble.

so when u set the mem timings to conservative, did the problem go away?
have u tried running just one ram stick?
or possibly changing the slot its in?
Also, if i remember correctly, memtest86 also tests the northbridge and cpu cache at the same time, but the nature of the tests, so even if u do get memory tests its not allways 100% certain to be the memory.
let me know how u go testing at diff speeds.

also, something to bear in mind is my experience with the KT133A chipset.
had two mobo's die on me now. first was the a7v133 and the second, iwill kk266-plusR.
after a long and exhaustive process ive found that the IDE controller in the Southbridge DOESNT LIKE my combination of IDE drives and FSB overclocking.
something there fries the controller when subjected to elevated FSB (thus increased PCI speeds).
i believe its my old (now retired) IBM 16GP ata33 drive thats was the heart of the problem.

and it wasnt temporary either. the damage has be done to the hardware... even with the offending device removed. so beware of old IDE devices and overclocking.

The lack of thermal protection on Athlon's is cunning way to stop morons from using AMD.

 

[kremington]

LHGPooBaa,

Well we have a conlusion, tho not quite a happy ending. through the aid of MEMTEST-86 and a spirited round of musical DIMMS, i can confidently say that my Crucial RAM is A-OK. however, my A7V133 appears to be a lemon.

Memory timings set w/ SPD at 3/2/2 (latency/precharge/delay):

TEST 5 of MEMTEST-86 failed on every attempt, in every single DIMM combination/position when attempting 133/133 FSB/Memory frequency; rock stable at 100/100; but check this out - also rock stable at 100/133!

lastly, i ramped-up all system/memory/graphics settings back to optimal. unsinkable (as the Titanic?) at 100/133!
it appears to me that the RAM is handling 133MHz just fine, but perhaps the northridge isnt?

thanks again for the ideas. i think i'm gonna have to retire (or maybe even return) this board. maybe i'll get me that new Abit KR7A-Raid...

KR

 

[lhgpoobaa]

well thats good news.
at least you have a solid conclusion.
i wasnt too pleased when my a7v133 went bananas...
least i know now it wasnt entirely its fault.

The lack of thermal protection on Athlon's is cunning way to stop morons from using AMD.

 

[Guest]

Your diagnosis re the IDE controller is consistent with my problems. I have trouble booting at 133. When it has problems it often hangs at the IDE/Pnp drive detection.
I will also try memtest86

 

[lhgpoobaa]

hmmm doesnt sound fun.

IDE errors are sometimes difficult to diagnose.

one of the best ways ive found is to create a large ARJ archive.
if you are unfamiliar with ARJ its a command line archive program, similar to zip.
so what i did was create a large arj archive containing a random assortment of files.
for ease i just archived one of my mp3 directories. (freely available on the web)

then u 'test' the archive integrity using ther command:

arj t myarchive.arj

so what it does is do a simulated extraction one which ever drive the archive is... and data corruption quicky shows up as CRC ERRORS for files.
when my mobo went bad it was happening all the time at 150fsb, to the extent that the test didnt get thru all the files. i also moved the archive around too, and it was happening on both my hard drives AND from the copy i burnt to CD, so i knew it had to be the controller and not the hard disks.
for me anyway the IDE problem seems to manifest itself progressivly worse at higher FSB, and also when lots of small rapid read write commands are done, such as bootup, archiving lots of files, or mp3/divx encoding.

Morally destitute, Emotionally bankrupt but a proud and respected member of Toms Forums!

 

[Guest]

I'll do that, I have used arj - sounds like a good way to track errors. I did think that my system was stable once running but it has locked up twice when I have run full virus scans...
Any ideas as how to fix IDE controller probs?

thanks
jr

 

[lhgpoobaa]

that arj trick is also good if your using extra long IDE cables over 18 inches long. and other insidious data corruption problems.

as far as i can tell IDE controller corruption is pretty terminal... couple of things u could try, or theoretically try. only choice D. is guaranteed to work.

A. use shorter cables if they are over 18" long
B. stop fsb overclocking immediately. (could try 100 instead of 133)
C. solder on a new southbridge (yeah right)
D. get a whole new mobo.

soonish im gonna go for option D.
by the way... what chipset do u have? is it a KT133 or KT133A? also, what IDE drives do u have on your system.
ive had this problem twice now with two kt133a chipsets.


Morally destitute, Emotionally bankrupt but a proud and respected member of Toms Forums!

 

[Guest]

I have a kt133a - I spend half the time running at 100Mhz the other at 133. Now I know about the arj method I do that to confirm if there is a problem with the IDE at 133 (or at 100 for that matter).

I have a new (6months) WD 40gb 7200 rpm drive. Currently there is only one drive in the box, however I can put in an old 4gb quantum...

 

[lhgpoobaa]

*nods & smiles*
sounding more familiar all the time.

as far as ive been able to work out, the problem comes up with the following hardware/software combo.
kt133a chipset + fsb overclocking + new? old? IDE drive.

my first mobo was a asus a7v133 with 2 HDD's, a 60GXP (40gb) and an elderly 16GP (16.8Gb)
first time it happened it took ages to diagnose, and i thought it was due to a dud mobo. so i replaced it with another kt133a, the Iwill kk266-plusR.
after about the same amount of time running at 150fsb, 3-5 months, i got similar problems comming up. not exactly the same, but pretty close.

example, on the old asus the problem went away when i ran at 100fsb. now its actually worse when i use 100fsb on the iwill.

with mine i thought it was the elderly IDE drive... but i was never able to fully confirm that. when it was removed it was a bit more stable, but by then perminant damage had be done.

Morally destitute, Emotionally bankrupt but a proud and respected member of Toms Forums! <P ID="edit"><FONT SIZE=-1><EM>Edited by LHGPooBaa on 03/27/02 00:25 AM.</EM></FONT></P>