v2000z stability issue

Hi,
The lappy work great until the last couple of days that weird things happen. I got 3 blue screen of death (two from resuming from hibernation, one when I was typing first version of this thread!) and one occation that the graphic went bad (I was using Firefox) and the system reboot itself.

I am running RMClock and undervolt but I don't think that's a problem since I didn't undervolt the minimum at 800 MHz. I try to test using OCCT, it's fine. Tested with Memtest 86 for about 2 hours, no error....

What's wrong??? It's my beloved 1 month old lappy

My config:
XP Pro, latest patch
Mostly latest driver from HP or other HW manufactuerer
Corsair PC3200 512MBx2
Seagate 5400.2 60GB
Sempron 3000+ under to 1.15v @ 1.8GHz

I'll try to test with Prime 95 soon. Any suggestion on testing would be appreciated. Kinda frustrate now!!!

***EDITED***
Just happen AGAIN!!! The graphic went bad (lots of black blinking horizontal line) and it REBOOT itself again!!!!!
ARHHHHH!!

 

Did it happen before you undervolted?

 

then dont undervolt and tell us how it is lol close rmclock

 

i had the same problem. You mentioned you were using FireFox, I' am assuming you are using the latest Synaptics driver for your touchpad in order for scrolling to work in Firefox.

I found out that It's the drivers from Synaptics that gives me the frequent blue screens. Try using the drivers from HP (no scrolling on firefox) or if you can find version 7.7.5 (something like that)

 

I feel, It has something to do with Undervolting. Generally screen goes crazy and hangs/reboot. You said you have not undervolted at 800Mhz, but even if you undervolted at next stepping, it could happen when proc try to step up.

Rule of thumb is add atleast 0.025 volts to your lowest stable voltage at every stepping.

What P-states you defined and at what voltages?

BTW, I think it is good idea to try RMClock 2.0, esp in your case, because, new version introduced transition stabilization voltage when proc steps up for few milli secs, like if your voltage at 8x is 1.15v and when proc steps up, it will apply 1.2v momentarily and then reduce back to define voltages.

This kind of stability issues I also went thru, and this happens during transition, so I am guessing you are on border at the voltage at 1.15v at 8x( or any intermediate step you may have defined). Anyway I am not seeing those once I upped the voltage by 0.025 volts at every stepping. ( in your case make it 1.175v at 8x from 1.15v).

 

For undervolting, I don't undervolt the minimum (at 800MHz). It's running at 1.075v @ 800MHz as normal. When it hang, it is running at 800MHz so I think it's kinda normal.
I have mem dump if someone wanna take a look at.

Also, HP site mention something about CHKDSK and I did it. It fixes something but later on it hang again (and CHKDSK fix something else again)
http://h10025.www1.hp.com/ewfrf/wc/...name=c00372718&product=1144546&dlc=en&lang=en

I will try to up some volt later.

 

CHKDSK basically just checks the HD for errors and fixes them.

 

That is what I am trying to say. It happens only when it try to step-up, but process does not complete, so we do not see processor at next stepping.

 

1. stop using rmclock for a while and see if you still have the problem-- if you dont...something with undervolting is causing it.

2. You changed your memory- even though many of us use corsair in our v2000z, that doesnt mean all the chips are perfect... put back in the orginal memory and see if you still have the problem--- if you don't--- there's your answer...

my 2 cents- give it a shot!

oh some people might say reinstalling windows may do the trick as well.

 

Use MemTest86+ to test your memory for at least half an hour. Bad memory is notorious for causing BSODs as you described.

If MemTest86+ says you're OK, I concur with the above posts about rmclock.

 

Well, I ran MemTest 86 for 2 hours without any error.
Try reseat the RAMs but error still occur. Last blue screen was IRQL NOT LESS OR EQUAL but I can't read the entire message since it happen so fast.

Now I disable RMClock and performing bios HDD check. Let's see if up the volts to normal make it stable.

 

Hope it works out, but if not:

You're not being thorough. Memtest86 should be ran longer... many say overnight. If you run it for a short amount of time, it'll detect short-term errors. People with these errors probably have frequent crashes on their PCs- you might not have these errors, so more testing is needed.

Also, the memory problem for the v2000z is a bit different. It has to do with a conflict between the memory and the computer. I've heard it's the GPU that causes the conflict. Well at any rate-- good memory that works on other notebooks don't always work on the v2000z so maybe memtest wont even catch it. That's why I said to take out the new sticks and put the old ones in to see if there's a problem.

 

I'll try to leave the memtest for tonite.
BTW, I still got file system error (need CHKDSK to run again). This is after the RMClock turns off.
...hmmm... I'd never found anything like this before.... maybe the seagate went bad? (Should not. Run full SeaTool diagnostic didn't found any physical problem, but only file structure problem).

I can't run CHKDSK/f to get rid off the problem yet. It says the Chkdsk need to be the only thing to (exclusively) run.. How can I do that?

 

Did you try the Touchpad drivers thing that Neero described? Uninstall everything you installed after you got the computer and see what happens. What is your BIOS version?

 

I didn't try the touchpad driver yet but I have been using the same driver since beginning.
The only thing I can't get rid off is file system error. I can't run chkdsk/f . Anyone know how can I boot to command prompt like the old days? Even I schedule the chkdsk to run at windows start , it won't get rid of the file system error...
Have to find a way to run chkdsk/f.

 

UH OH!
A little update. I reran memtest again and I got 27111 errors in the first 7 mins!!!!! (address 294b4 ff f00 0)
I'll try to reboot and rerun again.

EDITED: Reboot, set video mem to 32 MB (so Memtest will test more, I usually set at 128MB). No error yet so far...
I kinda rule out the file system error now since that might be from memory error. I have few theories.
1. One mem chip is bad, occationally.
2. Chipset bad?
3. CPU bad? Hardly...

Anyway, MemTest suppose to catch #1. I don't know how to test for #2 and #3. Any comment?

 

You can boot to command prompt by hitting F8 before windowst starts loading.
I guess you've tried running in SafeMode? Do you have problems while in SafeMode?

 

I still can't run chkdsk/f in safemode (got same message, chkdsk need exclusive right, etc).
Anyway, now I'm focusing on the Memtest 86 error. It is not consistent, but I got 2 occations with error now. Both for Memtest 86 and Memtest 86+.

 

Maybe one or all of the memory sticks are not seated properly. Try only each one of them alone at a time; I mean boot windows with only one of the sticks. I think it's the memory

 

I also think (and hope) it's the memory, not the controller and other things

Oddly, the full 2x512MB memtest last night for 10+ hrs has zero error. I stop for an hour and rerun the test and got 438853 errors!

I remove the lower slot and retest one of the 512MB on the memtest now...

 

Instead of chkdsk, go to Control Panel | Administrative Tools | Computer Management | Storage | Disk Management, right-click on the partition, Properties | Tools | Error checking.

Reseating the memory makes a lot of sense. I had to do that when I upgraded to 2GB, gave me a bit of a scare when my notebook didn't boot up. I reseated the memory and it's been OK ever since.

 

After reseating the memory (only single stick). The BIOS reset itself to default stage (also wipe off all protective password). Is that normal?

BTW, access that error checking from drive property is similar to chkdsk. If you check "Automatically fix error" it will say the same thing and not allow you to do that.

 

You probably could save yourself the trouble by putting the original memory back in only for a while and see if it crashes. Stock memory from HP/Compaq rarely will have issues.

 

I'm confused you mention single stick while your first posts says you have 2x512 (two sticks). Perhaps you are now only using 1 of the 2 sticks.

I have a single stick 1GB corsair. It simply cannot co-exist with my 200M card. But in my case, I get blue screen during the windozes startup consistently if I simply put the stick as the only mem stick in the system or put it in slot2, no matter for Bios F1.0A or F13 that I'm now using.

Since the 200M tries to share RAM from high-end phy address. I put this 1GB stick to slot 1, and OEM stick to slot2 (so let the OEM mem share with the 200M), I never have a single problem (except I have to run at PC2700 ) even though I frequently run mem hungry apps.

I'm not sure if it is also motherboard version/chipset related. I got my v2000Z last early Aug (ordered in Jul), one of the first production batches I believe. At that time there was no report about the lappie being picky about memory and what memory speed the system is compatible with.

If I were you, I'd like to do what breezie said, use the oem memory first (w/o undervolting), to ascertain:
*if it is system problem
*if it is add-on RAM problem, this this case, you can use my above method to further determine see if it is a faulty ram or ram compatibility issue. Typically it is the latter.

Goodluck.

 

I keep reading problem reports from people who bought Corsair memory. Their quality control appears to have deteriorated in the past few years. I had a pair of Corsair 512MB SODIMMs fail in my zv5000z when I bought them early last year. Replacing them with Crucial/Micron sticks fixed everything.

edit: Er, make that early 2004, not early last year.

 

if all things fails just do a complete factory restore and only use the updated drivers from the HP site. then you can add from there. mem,undervolt etc . . .

if you said your synaptic drivers were the stock ones from HP then I'm assuming you cannot use the scroll function in Firefox.

 

Hi all,
A little clarification. I have 2 x 512MB Corsair PC3200 Cas 2.5, but it's actually run at Cas 3.0. I have these two sticks for around a month before this problem occurs. Before that, everything is perfect.

I have not changed any settings. BIOS was always F.13.

I am testing each Corsair stick (I don't have the original PC2700 stick with me now) and both (when exist individually) occationally got hang and got Memtest errors.

Now I got errors in Memtest which might indicate the hangs do not relate to the undervolt. I stop using RMClock anyway, just to eliminate the possability.

BTW, which one is slot 0 (shared with video?) which one is slot 1? The upper or lower slot is slot 0? (I guess the lower slot? But the system seems to boot and get into Windows fine if a stick exist on either slot).

I'll try to see if I can find the OEM stick now. Maybe it doesn't really support PC3200 all the way

 

The slot 0 (slot 1 in my post as I use 1&2) is the underneath one from the prospective when you open the bottom cover.

So in my case, I put corsair mem in first to slot0, then seat the OEM stick (in slot1)

 

Hi tempoct,

It has been reiterated several times before also, the Corsair is not CAS 2.5 at DDR400, it is CAS 3 as per SPD timing table( even as per COrsair it is CAS 3 at DDR400), It has been listed incorrectly.

As per the memory slots, yes you can keep memory in any slot, but I remember if slot 0 is populated( I think lower one), video card takes memory from that one.

If you have used A64 Tweaker before, please load default in bios. Because fiddling with A64 Tweaker caused lot of problems and bios loading incompatible memory timings.

So, load biod defaults, adjust video memory as you wish, and do the necessary tests.

 

Chinna,
I meant the same thing as you said. The Corsair is actually CAS 3.0 but were advertised everywhere (except by Corsair itself) that it is cas 2.5.

Anyway, I think the cause of instability is both sticks are BAD now. From this evening, I run many stress test with the stock ProMos 256MB PC2700 and was not able to cause any error or blue screen hang. The test is run on Windows as well as MemTest.

What should I do now? Exchange for the new Corsair or get another brand? I can't stay like this long since performance on 256MB PC2700 is freaking slow! Too bad that OCZ 512MB PC3200 sales at newegg seems end already... Any good deal on that? (Hopefully OCZ should be more stable. What other brand to recommend? How's Crucial?).

Nevertheless, I will run full night Memtest with the stock Promos tonight and will report tomorrow morning.

 

it's very very unlikely that both sticks are bad. crucial is a well known brand after all. Try them one by one or in combination with the original; rotate their places. I don't know if it's possible but it might be that one of the slots is damaged

 

I totally agree with you but I can't find any other explaination than this.
I can't really fault it with the original Promos stick. The Corsair were working great for a month. When I test the Corsair, one of the stick crash very often while the the other one crash only two time. Anyway, I don't really want to risk.
How's PNY? The 512MB stick is on sale at bestbuy now for about $40.

 

OK. The stock ram doesn't give error.
I can't return the RAM now since it's over 15 days (just notice the place I brought have 15 days return on RAM). So, I call up Corsair and they mention that the lappy support only PC2700 and running it at PC3200 will have problem in the long run.
I have no other choice so I am getting replacement from them.... Anybody with Corsairs success @ PC3200? Chinna? How long/ how extensively you have used the RAM?

NOTE: One thing I find out about is, I believe that PC3200 really give instant 10% improvement to my system. When I run at PC3200 512MB, super PI 2M is 2:04min, PC2700 256MB is 2:11min. PC3200+2700 @ 768MB (running at 2700) is actually lagging at 2:14min. I did some batch processing of raw image processing and have some improvement on PC3200.

 

So are you 100% sure the memory is the cause of your previous problems?

Easiest way for anyone to test their notebooks is to yank out the new ones and put the original ones back in b/c stock memory usually is tested by hp/compaq before they ship out a unit.

At any rate, sorry to hear about your problem- corsair usually won't give you bad memory so I'm guessing it's just a conflict with your mobo or gpu chipset vs the memory chips. I bet you if you put that memory in any other notebook, it'll run fine. Also corsair doesn't know that the v2000z actually does support pc3200 with the newer bios.

I'm suprised that Corsair gave you issues, a lot of users including myself havn't had issues with it. I'm using one in my v2000z as well and have it there since like October. I put the corsair in the memory slot that doesn't take memory for the video, so maybe that's why I'm not having problems.

Btw, don't get the PNY, I hear MANY v2000z users had the same issue with PNY.

 

I have been using this since last 3 months without any problems. I use atleast 10-12 hours a day as this is my main development machine for now. No have not see any problems.
I had some problems when I tried different timings with A64Tweaker, but when I loaded setup defaults in bios that solved the problem.

I even OCed it for testing to DDR416 speeds, but that is just for test.

But, did you try to load setup default in bios and tried running the memory you said has less errors? I would give it a try.

It is possible those two memories are from same batch and on border and causing error when extensively used. I would not hesitate to get a replacement from Corsair.

 

Hi,
I did test the stock ProMos works fine.
I'm reflashing the BIOS (F.13) again and will try to run some more test while I'm waiting for RMA from Corsair.

 

what is your current Bios? isn't it F.13?
BTW I suggested to load setup defaults in bios, not reflashing.

 

hey, did you happen to get my PM?

 

Chinna,
It was F.13, I reflash it again (just in case something in the BIOS got messed). I also load default setting after that.
I retest the less error stick, it given me Memtest error again after reflash the BIOS. So, I think both sticks are bad.

Hegemon,
I don't get your PM. Did you mean to ask Chinna?

tempoct

 

::::UPDATED::::
Actually, it might be my motherboard. I got some Memtest error with stock RAM now!!!!
OMG!!!!

What's the next step to test???

 

Guess you have to contact HP and get it replaced if bought within last 30 days.. Otherwise, you have to send it to HP for repair.

 

just replied!

 

You know what, it just got pass 30 days.... (got it on Dec 18).
I already contact Corsair for memory. Maybe I should try to send it to HP first. Memory has life time warranty anyway.

EDITED: BTW, is there anyway to run full diagnostic on the motherboard? Maybe the chipset??

 

past 30 days for a return but i believe you still have 1 year warranty. Did you do a complete 100% factory restore? both hardware and software.

 

Well, there's rebates involve for complication if I took it back.
Anyway, I contact HP late last night and they schedule FedEx to pick it up tomorrow. Everything is prepaid by HP.

So far, the support experience is quite good. While I have to go through the 'usual' process (repts always think user doesn't know anything), it's not that long. I have a desktop so I'll be good for a while. Missed the portability though.

Let's cross the finger and hope it gets back to me soon