T20 hanging, BSOD, etc. Help appreciated

[LymanSS]

Hi all,
I've been trying to help a friend resolve some issues with his T20. When I first tested the machine it would give me a BSOD within a few minutes of starting up, every time. I thought it might be a memory issue, and so we ran a memory check. It came up with errors, so I swapped out the memory for some extra memory that I had laying around. Now the machine seems willing to run a bit longer, but it's still hanging and giving the the occasional BSOD. Then I realized that is was crashing mostly on ocasions when I was picking up or moving it. Now I wasn't tossing it around...just picking it up off of the table to set it on my lap, or vice versa. Sometimes it'd simply freeze (no response to mouse or keyboard) but other times it'd reset completely and I'd get the boot sequence. So then I start wondering if there's a loose component. I opened up the machine a bit (pulled the keyboard and looked around inside....) There aren't many connectors in there that might get loose. I checked everything I could see, but couldn't find any glaring problems. After that I did a BIOS update (which I meant to do earlier, but didn't due to lack of internet access) and experience no benefit from that.

So I'm stumped. Is it a bad motherboard? I'm hoping that it's something I can fix without having to buy any replacement parts. here's a little more info about the machine.

Thinkpad T20
3647-55U
PIII 750 MHz
256 MB ram (currently)
Win XP SP1

Any info or advice would be very helpful. I've been searching the forum for something similar but haven't found an answer yet. Thanks in advance for any help.

Scott

[tfflivemb2]

What errors are given with the BSOD?

EDIT: If the system is freezing while being moved around consistently, then the problem is most likely damage caused by "flexing" the motherboard. This appears to be quite a common problem...especially lately.

[BillD]

so I swapped out the memory for some extra memory that I had laying around.

Had laying around? Not a good idea.. If you really want to try and save the machine I would buy a stick that you KNOW is good and is the EXACT type for a T20 and try that.. I know it's going to costs some money but more times then not a machine that is suddenly acting up it is a problem with the RAM..

[LymanSS]

Unfortunately I don't have a copy of the error code. I tried to write it down, but didn't ever manage to get it before the machine reset. I know that I got at least two different error messages. One of them was IRQL (or something like that) not_less_than_or_equal_to... Then there was a message about how if this is the first time you've experienced this error you should try removing any new equipment you have installed, and if that doesn't work you should disable memory features such as cacheing and shadowing. Halfway down the screen there was a series of number strings (possibly hex, I didn't get to look closely enough). At the bottom it said "beginning physical dump of memory" That's most of what I remember.

Other times I got a totally different message, which I don't recall with much detail. Which information is critical for diagnosing it? The message at the top, the numbers at the bottom, or all of it?

As for the memory, it wasn't literally lying around. It was properly stored, and basically brand new. I had two PC100 128 MB sticks, which I had bought for another laptop. I can't say for certain that it's exactly the right memory for this T20, but I figured that it'd help us determine whether the problem was something to do with the memory that was in there, or if it was something else.

Any suggestions? I'll try to get the BSOD to happen again and see if I can get the info off of that.

If the flexing of the motherboard is the problem does this mean that there' s already physical damage to the board, or simply that there's something that's getting shorted out or disconnected or something when the board flexes? In other words, am I going to need a new board?

Thanks again for the help. I appreciate the responses so far. It's great to find knowledgeable people who are willing to take the time to help.

Scott

[Kyocera]

Try to get the BSOD again and write down the results and post back or google the error message.

Have you installed anything else beside the ram? Like a new pcmcia card or some other peripheral device?

[BillD]

The BSOD are usually a hardware conflict...On a desktop 90% of the time it's the RAM,video card or sound card..

Obviously on a laptop video and sound card issues are moot... So again I'd try swapping the RAM with a stick that you KNOW is good..Also try one stick in each slot,maybe one of the slots is bad.. I have a 600e that is like that.. If I put a stick in one slot, no problems,if I use 2 stick or move the stick to the 2nd slot I have nothing but problems...

[LymanSS]

Thanks for the suggestions. I'll try to get the BSOD code. I'll also try swapping out the memory, and running only one stick at a time to test both slots.

To my knowledge there are no new devices installed in the machine. It's actually been sitting in a box for a couple of years, so I don't know what was done to it immediately prior to the failure. However it has no PCMCIA cards or other attached hardware. The only thing that might be "new" is one of the memory dims that was in it. It has the stock 128 MB stick, plus another 64MB stick which has a picture of a viking or something like that on it, and says "PC100 certified." But I pulled both of those and replaced them with my (arguably questionable) ram.

Often when it fails there is no BSOD. I'll pick it up or set it down and it just resets....as though it had a hard reset button and I just pushed it.


Again, thanks for the advice.

Scott

[Kyocera]

There may be a good clue, as tfflivemb2 said about "flexing of the motherboard" can cause problems.

[LymanSS]

If it's a flexing problem, how could I determine whether it's a problem with the board itself (say...a loose soldering point or something) as opposed to a mounting issue, or a connector that needs to be re-seated? I'm hoping to avoid purchasing a whole new mother board, but if there's a flaw in the board itself I don't imagine I'll have much choice.

Scott

[Kyocera]

I've never had the issue, seems like way back when before i bought a t series a friend demonstrated to me how when he picks his up with one hand it could cause damage, that kind of stuck in my head, always grab with two hands no matter how uncomfortable it might be .

Not sure how you would troubleshoot that and if it could be fixed. Wow that's a big help, huh. Someone else may shed some light though.

[LymanSS]

I've been reading through all the old posts I can find that might be related to my problem. In this post http://forum.thinkpads.com/viewtopic.ph ... t=t20+test

Beeblebrox mentioned that the processors have a history of coming loose after a while. This sounds like one possible source of my problems. I know that jostling the processor loose would probably reset or at least hang the machine. Any thoughts? Thanks

Scott

[LymanSS]

Well I've put some more time into it, and here's what I've figured out. First off, I checked my notes, and the BSOD has only come up once since installing the replacement memory. Since then it's mostly been hanging, and spontaneously resetting. After checking everything again and again, I finally got back to the memory, and discovered that one of my dimms had unseated itself. I pushed it back in, but it simply won't seat correctly, in either slot. There must be some slight dimensional difference with this one dimm....it just keeps popping up. The other stays seated fine.

So I just pulled the faulty dimm and started it up on 128 MB or ram. Wow, does it run slow now! Anyway, I seem to have eliminated the hanging, but it still spontaneously resets sometimes.

So I opened it up and re-seated the processor. It didn't seem loose though, and in the end I don't think it helped anything.

I've also experienced a no-start condition, which I think is similar to that described in some other posts. I press the power button and all the LEDs light up, and then just the power LED stays lit and nothing happens. Is this a sure sign of a bad board?

So, to sum up, I've had four types of failure:

BSODs - these mostly occured with the old memory (which did test bad at one point, but later tested good...all tests done with memtest86

Hanging - no response to moust or keyboard, often after a bump or jostle. this seems to have vanished after I eliminated the dimm that wouldn't seat right

Reset - just goes back to the IBM logo and begins rebooting. Happens with no warning

No start - this has mostly happened after the incidents where the machine hung and I reset it by holding the power button.


My next planned step (unless someone has better advice) is to locate the floppy drive so I can run PCDoctor and run some diagnostics. I'm hoping that has some sort of useful motherboard diagnostic. After that I plan to reinstall windows. If none of that works, I'm out of ideas. I'll probably just spring for a new board. I'd sort of like to go for a T23 upgrade while I'm at it, if the cost isn't prohibitive.


Again, any feedback would be greatly appreciated. thanks for all the help so far.

Scott

[BillD]

Reset - just goes back to the IBM logo and begins rebooting. Happens with no warning

How long is the machine on before it resets? Overheating?

[LymanSS]

I considered overheating, but sometimes it'll reset after just a couple of minutes, or even thirty seconds. The fan seems to be operable, and when I re-seated the processor, I checked the little grey thermal pad, and it seemed to be intact. So I don't think it's a heat issue. But perhaps it's a heat issue in some other part of the machine?

Scott

[tfflivemb2]

Since you're running XP have you tried going into the Event View (Control Panel\Administrative Tools) and see what, if any, errors are occuring at the time of reboot?

[LymanSS]

I haven't tried that. I wish I'd thought of it. Unfortunately the owner of the laptop just finished wiping the hard drive (he wanted all the data removed because there's some sensitive information on the machine) in preparation for a re-install of windows. So now I can't check that. But if it keeps failing after we've re-installed windows I'll check the event viewer. It may shed some light on the problem.

Scott

[Captain Marvel]

When you press F8 to intercept the startup, one of the choices is to disable rebooting when there is a problem. This leaves the blue screen showing until you tell the computer to reboot. You should be able to see what it says. I had this on one of our computers this morning and we finally got to see that the hard drive was causing the blue screen.

[LymanSS]

Well, I've done some more work on it. We had to order a floppy drive for it (five bucks off ebay) so we could run PC Doctor. We got it, and ran PC Doctor through all it's tests and it came out fine. It did hang the first time we ran it, but I thought it might have been a thermal issue because it was sitting on a bed. So I put a cooling pad under it and it ran fine and passed all the test.

So I re-installed win XP. It got through the install fine, and I decided to just use the machine for a while to see if it failed. I brought it to my living room, set it up, booted up, and it gave me a blue screen thirty seconds after booting into windows. I didn't catch the code on the blue screen. It then rebooted on it's own, and I left it for a while. When I got back I had an error message in windows saying that windows had recovered from a serious error, or something like that. It offered this error code.

BCCode: 1000007f BCP1 : 00000008 BCP2 : 80042000 BCP3P : 00000000 BCP4 : 00000000 OSVer 5_1_2600 SP : 0_0 Product 256_1

I tried searching for that on google, but didn't come up with anything. Any help in deciphering it would be greatly appreciated. Also, I'm starting to accept that I most likely need a new system board. I'd like to upgrade to a T23 if it's possible, but I can't seem to find a cheap source for a T23 system board. On ebay there are boards in the housing for about 100 dollars. There are boards on their own for 200 dollars, and there are fully functional T23s for about 150. I don't really understand this pricing. Why is a bare board more expensive then a functioning machine? I'd appreciate any help in finding the best/cheapest T23 board I can.


Thanks a lot for all the help so far. It's made this a lot easier.

Scott

[tfflivemb2]

I know that this auction is about to end, but here is one on eBay that I just bought for $96 shipped. it works great. I am assuming that he will list again as there is only one bid on an auction with 6 boards available.

Click Here.

[starbork]

I would also recommend pulling out the NIC/Modem card from the minipci deal under the panel on the bottom, especially if you are getting IRQ conflicts. I had a similar issue w/ my T22 and it turned out to just be a bad NIC. In fact, you might field strip it down as much as you can-- ultrabay out, any PCMCIA cards out, 1/2 RAM out, etc. At least take those things out of the equasion.

Best of luck!

Ryan
--------------------
T22 900 / 512 / 80@4200 / XGA / "Scout"
T23 1.2 / 1024 / 40@5400 / XGA / "SkullTron"
T23 1.2 / 768 / 20@7200 / XGA / "SuperScout"

[tp23]

The combo NIC/MODEM issue is infamous with T2x systems, removing it would have been my first suggestion, but usually windows setup halts while trying to install a defective device.
BTW, did you buy another system board mounted in a tray? The a/c power adapter jack connects to the I/O card by mechanical connection, the little screw under the power jack needs to be tight.