PDA

View Full Version : Urgently Need Some Help


Pages : [1] 2 3

poer
09-13-03, 01:37 PM
Hi,

I am new to this forum and wondered if you guys could lend me a hand. I posted a message in another forum at another site and I am now posting it here, looking desparately for help or guidance.

Please see what follows....

Posted below is my original post, made some days back on another forum with the symptoms and my system specs (easy to find all the info this way). The followup to the original post is that I was sent a new 5900 FX card and once installed I am now getting an error called BAD_Pool_Caller on start up. It reads STOP 0x000000C2(0X00000041, 0X852FE000, 0X000052FE, 0X0003FF2F). This error appears to be associated with the end result which is WIN XP instability at desktop, frequent reboots, and cycles of 5-6 reboots just to get to desktop. If the PC does get to desktop, it usually winds up being fixed at 600X800 resolution in 4 bit color, with no option to adjust and the videocard tab gone - no 5900 FX options. Alienware is clueless. They have had me reinstall this system 2 times and have had me start/stop this computer with components (RAM, Soundcard, etc.) in and out with no joy. They have just sent me 2 sticks of replacement RAM, and if this does not work, they want to have me ship the back so they can test/repair it (who knows how long that will take?). CAN anyone of you make some sense of this situation and tell me what the problem might be and if the RAM could be the problem?

Thanks

Russ

Original Post below

Need some advice about a problem that I have with my video card. Recently I started having trouble with my post (you know the part on boot up with the black screen, white letters)at startup. This happens when the PC is cold and is started up, whe this is done it displays vertical columns across the screen with the letters "a" and "~". If allowed to complete the boot, the WIN XP desktop will have blue vertical lines with dashes in them across the screen and the screen will blink on and off. A restart from here will cause a lockup on black screen and accessing a program will cause a crash of the program. If you do manage to restart WIN XP the desktop will either look and run normal, or, it will come up in 640X480 in basic RGB and the video card will not be recognized.

Before I give my system specs, I should state that this is an Alienware Area 51 built last July. I have used both 45.23s and completely removed them when I got CTDs. With the 44.03s, I have had better success and it was with these that I noticed with games such as BF 1942 and IL-2 (graphics card intensive) that the temperatures were reaching upto 120C (using the Temp monitor)and also the card would interupt the program in progress to notify me that it was not getting enough power and would shut down features to avoid damaging my equipment. Since the temp cutoff was 140C, I suspect it may have hit that. When I looked inside the PC I noted that Alienware has placed a ribbon cable leading to the soundcard inside the cooling fan to the videocard - it was not able to run and probably never had. Removing the cable allows it to run cool (in safe range 40-60 C). Crashes as described above still happen. All power cables from the power supply to video card have been checked and are connected. Card is seated fine.

Is the card to blame for all of this? And what is the possibility that the motherboard is damaged due to high heat? Suggestions? Alienware is replacing the videocard.



System Specs.

Intel P4 3.00 GHz, 512 K L2 Cache 800 Mhz FSB
Intel D875PBZ Motherboard (bios updated regularly)
2X512 MB DDR PC 3200 RAM
Seagate Barracuda 120 GB 8 MB Cache HD
Nvidia GF Ultra 5900 256 MB (44.03 Drivers)
Creative SB Audigy 2, 6.1
Plextor Plexwriter Premium 52X
16X DVD
Win XP Home (fully updated with critical updates)
420 watt power supply
US Robotics V.92 56k PCI Modem (yes in the sticks here)

bkswaney
09-13-03, 04:34 PM
wow... sorry but I have no idea.

Untill now this was almost unheard of in nvidia drivers. :eek:
I guess even nvidia can have unstable drivers.

Well, at least we don't see this anymore in ati's drivers. ;)

GlowStick
09-13-03, 05:10 PM
Hi,

Anyways, this defiantly sounds like a Hardware problem. I strongly belive that something has broke, or worn out somehow, due to manufacuter defect.

Definatly, my list of suspect hardware is:
Video Card
Ram
Hard Drive
MotherBoard

I think the ram could be a good place to start, but i have a feeling that somehow your card could be just bad. Did this suddenly start happing, or did smaller symtons build into this one?

I hope this will help you, but if alienware dose need to take it back, do you have a warrenty on your machine still? I hope you do, the hardware is not that old.

You could also swap out your video card with firends and see if the same thing happens, but for warrenty issues, maybe you should let Alienware look at it.

poer
09-13-03, 06:44 PM
Thanks for the replies. It is still under warranty and so are all of the parts. I will keep the list of culprits in mind. Since Alienware is sending the RAM, I will start there. I hate to have to ship this to them for a month, especially when it is less than 3 months old.

Is there any way to test the ram in the machine now?

Thanks,

Russ

sebazve
09-13-03, 07:20 PM
Originally posted by poer
Thanks for the replies. It is still under warranty and so are all of the parts. I will keep the list of culprits in mind. Since Alienware is sending the RAM, I will start there. I hate to have to ship this to them for a month, especially when it is less than 3 months old.

Is there any way to test the ram in the machine now?

Thanks,

Russ

can you borrow some RAM stick from a friend or something and try with that if you still have problems then i dont think its the memory.

Another thing hight heat could have damage the silicon from your card so that could be a problem. 140 was too damn high. Try another video card.

poer
09-13-03, 10:39 PM
I will be getting new ram from Alienware next week. That one should be easy to eliminate. The card that is in the PC is the new 5900. The old one was sent back. It was the old one that made 140 C.

Russ

GlowStick
09-14-03, 02:16 AM
Hi,

http://www.memtest86.com/

I belive that MemTest86 should stress test your ram. I think it can detect when bytes arent what they should be, and if it continues a crash 'would' occur.

Please share your results if ya test it!

poer
09-14-03, 11:19 AM
Results of MemTest are:

1 pass only:

Test/Errors

0/0
1/3
2/0
3/26
4/17
5/213
6/5
7/2

Safe to conclude bad RAM?

Russ

aapo
09-14-03, 11:25 AM
Originally posted by poer
Safe to conclude bad RAM?


Certainly. Never seen a computer that can actually run the Memtest86 without lockups when a RAM module is that badly broken. :D

EDIT: Actually, after I did some more thinking, I'd say you should test again and see if the failing memory addresses are the same every time. If they are, it's certainly the memory module(s). If the failing addresses are changing, it could be the memory, the motherboard, the processor or even a weak PSU that is the culprit. Most probable is the memory, though.

poer
09-14-03, 11:30 AM
Thanks - never used this one before so was not sure how reliable it was. I think at this point the original problem may have been two fold. First the blocked ribbon cable that caused the card to overheat and then the bad RAM - not sure if this has anything to do with the videocard.

At this point the system runs itermittantly. I wonder if I am running the chance of damaging it further by continuing.

Appreciate the tip on MEMTEST.

Russ

GlowStick
09-14-03, 11:30 AM
Whoa, yeah, i beilve Alienware is right about your ram, they are cross shipping right?

poer
09-14-03, 11:36 AM
Cross shipping?

Some additional info - not sure that address is changing, but the count sure is with each pass on memtest. See this for 5-6 passes.

test/error

0/0
1/22
2/0
3/114
4/44
5/1278
6/25
7/2

Mean anything to you guys?

Russ

The Baron
09-14-03, 11:40 AM
Bad RAM.

GlowStick
09-14-03, 11:44 AM
For cross shipping i ment, did Alineware send you the ram, and when you get it you send the old ram back? And once you get it your problems should be fixed : D

aapo
09-14-03, 11:49 AM
Originally posted by poer
Some additional info - not sure that address is changing, but the count sure is with each pass on memtest. See this for 5-6 passes.

Because there are that many errors, it may be hard to collect all the broken addresses. But if you see a lot of same hexadecimal numbers under the label "Failing Address:" (like e.g. $00a076a0) on each passes, the memory module is the culprit. If the failing addresses are completely random in all passes, the problem is not necessarily in the RAM.

The amount of errors doesn't tell anything; a given broken address may randomly work in some of the passes and fail in others - that's the reason to multipass.

The easiest way to check if it's the RAM is to run Memtest86 for a while (about 1 to 2 hours). If it doesn't lock up, it's the RAM and your computer will probably work ok with the replacement RAM. Random failures in processor or motherboard will lock the Memtest86 sooner or later (usually later), whereas bad memory failures lock the Memtest86 almost instantly and 'normal' memory failures lock it never up.

poer
09-14-03, 11:56 AM
Yes, they are. New chips to me and old ones go in the mail.

How can I verify that the memory locations are fixed for the errors? Also is there any way to set the number of passes the program makes?

You guys have been top notch in helping. Thanks so much.

Russ

aapo
09-14-03, 12:06 PM
Originally posted by poer
Also is there any way to set the number of passes the program makes?

It will make passes forever or until lockup or until reboot, whichever comes first. ;)

But it seems more and more that your problem really are broken RAM modules, so I'd change them. In an unlikely situation it doesn't help, I think you should send the whole system to them, because you have the warranty. :)

I usually troubleshoot older system with no warranty anymore. Once it took over a day to run the Memtest with a slow computer to get enough passes to be sure. Lucky you! :p

poer
09-14-03, 01:34 PM
Let me make sure I understand. If I run, and I have done this, for 2 or so hours and make 7 passes and I am getting large numbers of errors but MEMTEST does not crash....then this means normal RAM errors AND the RAM is BAD. Right?

Russ

aapo
09-14-03, 01:52 PM
Originally posted by poer
If I run, and I have done this, for 2 or so hours and make 7 passes and I am getting large numbers of errors but MEMTEST does not crash....then this means normal RAM errors AND the RAM is BAD. Right?

Exactly.

Of course, there is always an "EXCEPT, IF"-situation (e.g. Memtest86 code fits entirely inside your processor's cache - that's one of the reasons why you can select to run the tests without using any cache at all but that's very slow). But believe me, this time it's the frigging RAM module that's broken. :afro2:

poer
09-14-03, 03:26 PM
Whew! That is good to know. At least there is an end to this madness. I am relieved. Thanks for your help. I guess next step is rip the RAM out and replace it when the new ones come.

GlowStick
09-14-03, 03:36 PM
Great, Besure to post after you get it installed and give us a follow up of if your problems go away!

aapo
09-14-03, 03:48 PM
Originally posted by poer
I guess next step is rip the RAM out and replace it when the new ones come.

Yes, but be careful not to touch any wirings on the RAM modules, and touch a bare metal in your computer case before you start so you are grounded to the same level as your computer's ground level.

poer
09-14-03, 03:52 PM
Will do. Now at 30 passes. No crash at all for over 5 hours of operation in MEMTEST. Literally thousands of errors though....

poer
09-18-03, 10:06 AM
Follow up to the original post. I got the ram and installed it. Checked with MEMTEST and 0 errors. Looks real clean. Reinstalled WIN XP and during install got the bad caller error on the reboots during the install program. XP recovered and finished installing. I am still getting bad caller errors with a very low frequency now. Not sure why this should be happening. All drivers are updated. No conflicts in device manager. Does anyone have any ideas?

Thanks,

Russ

GlowStick
09-18-03, 11:21 AM
Originally posted by poer
Follow up to the original post. I got the ram and installed it. Checked with MEMTEST and 0 errors. Looks real clean. Reinstalled WIN XP and during install got the bad caller error on the reboots during the install program. XP recovered and finished installing. I am still getting bad caller errors with a very low frequency now. Not sure why this should be happening. All drivers are updated. No conflicts in device manager. Does anyone have any ideas?

Thanks,

Russ Ah, well there was defiantly something screwey with the ram.

Now, I have had a silimar experince where the installer crashes, and it was a power/hard drive issue.

Anyways, i think since there could be a ton of diffrent things that could be wrong, you may end up having to send it back to alienware, and since they have all the parts, they can systmaticly switch every part out untill it works.