X31 - please help :( Very strange symptoms (lots)

[riksarider]

Hi fellow forum members,

I'd just bought a used X31 a few days ago. I was told before the deal that it sometimes does
BSODs, but I didn't care much and took that as an easy-to fix OS problem. System is Windows
XP SP3 (XPForever Edition).

Alas, I may have been very wrong.

Symptoms:

-notebook is very hard to turn on (that is, sometimes I have to remove all power sources,
including battery and AC several times, before I can get a boot logo instead of blank
screen).
-rarely, during the blank screen it gives 4-4-4-4 beeps
-it has 2x256 MB DDR memory modules (Elpida, Infineon, both original IBM FRUs)
-battery is operational, it's good for about an 45-60 minutes (when the notebook works OK,
that is)

Random BSODs everywhere (sometimes right at the Windows boot, sometimes out of the blue
during work...rarely it goes for hours without one). BSOD labels I've got so far (I don't
know their error codes by now):

CRITICAL OBJECT TERMINATION
KERNEL MODE EXCEPTION NOT HANDLED
DRIVER IRQL NOT LESS OR EQUAL
PFN LIST CORRUPT
MEMORY MANAGEMENT
PAGE FAULT IN NONPAGED AREA

Boot-time system errors (before trying to boot Windows):

-invalid date/time (remark: I've disconnected the CMOS battery, but it did the same when it
was in)
-invalid CMOS settings
-Resource Conflict – Allocation error static node # 02

What I've tried so far:

-Booting into a live Linux USB (SlitaZ). That worked quite good, with two exceptions (one
freeze at the Linux boot menu and another freeze with distorted graphics while working).
-disconnecting the CMOS battery
-removing both memory modules at the same time (it gave the 1-3-1 beep, which I guess is OK)
-removing only one memory module at a time
-and because I heard about some security chip that on a daugterboard can be loose: taking the
whole poor thing apart (didn't succeed in getting to that chip, so I put it together again)

What the heck could this be ?
I'm pretty desperate by now, as I saved up my last money to get this, so I can do online
freelance work when away from home. What else could I try to avoid putting this thing back
for sale ?

(Also, another, but less significant question: some of the keys (in the middle of the
keyboard) have the famous nice ThinkPad quality feel when typing. But the Enter, Space, Caps
Lock, Tab, Left Shift and cursor keys are horrible noisy and hard. What can be done to get it better ? It may be that when I took it apart before, the keyboard didn't fit in right back...)

Many thanks in advance for anyone, who gives useable answers !

Gabriel

[rkawakami]

Welcome to thinkpads.com!

The first thing you should do with any "new" system is to verify that the memory is fully functional. The quick (and maybe even the extended) memory test that the BIOS does at boot is nowhere near as good as a dedicated memory diagnostic. My choice would be to use memtest86+ which you can download in several forms from here:

http://www.memtest.org/#downiso

Once you have created the boot floppy / USB drive / CD, use it to start the laptop with all of the other major pieces of hardware removed. That is, take out the hard drive, MiniPCI card, any communications daughter card (modem, Bluetooth), PCMCIA card and leave just the memory modules. Assuming it boots normally, let the diagnostic run for at least three complete passes through all of the tests/patterns. If there's any failures (you'll know it by a bunch of text on a red background, or the system hanging), shut it down, remove one of the modules and try again. If that one passes, try the other module by itself.

The 4-4-4-4 beep code is related to the Security Chip so I'd also re-investigate that issue again. If you don't already have one, download a copy of the Hardware Maintenance Manual for the X31 from here:

http://support.lenovo.com/en_US/product ... MIGR-44018

[riksarider]

(UPDATE: On second pass, it's got 37 errors on test 8, with confidence value risen to 236 from previous 181. Highest error address now is 509.5 MB.
I guess that indicates that either both sticks are faulty or the motherboard has a problem with handling those sticks ?
Will let it run for a while and post the results later, too.

UPDATE 2: Memtest hung up totally at 46 % of second pass - no further errors, apart from the previously mentioned ones. Had to press Power buttong long to power off, total frozen.)

Thank you ! I wanted to test with one RAM module at a time, however, I could only switch the laptop back on with BOTH modules in. :O
Regardless, Memtest86 have just already finished with the first pass:

Error Confidence Value: 181
Lowest Error Address: xxxxxxxxx - 26.1 MB
Highest Error Address: xxxxxxxx - 263.6 MB
Bits In Error - Total: 1 Min :1 Max : 1 Avg : 1

Max. Contigous Errors: 1

(on the right side column) Test 7: Errors: 2

Well, like I said, I have 2 x 256 MB. Lowest error was at 26.1 MB, highest at 263.6 MB.
Does that mean that both of my modules are faulty ?
(I've learned, that an Error Confidence Value above 100 means that all listed errors are about 100 % real.)

Also, I've removed the hard drive and disconnected CMOS battery before running Memtest86.

I need to know, if the RAM should be exchanged (or is it the board itself), as I arranged a good deal for an 512 MB DDR module today.

Best regards and thanks again in advance,
Gabriel

[rkawakami]

Beside the possibility of having two bad memory modules, there's also a chance that the motherboard has a problem with the memory sockets (bad contacts to the edges of the modules or pin(s) that are not fully soldered to the motherboard). About the only way of knowing which it is would be to install a known good module (i.e, verified fully passing in another system) in your system OR transferring your modules into another laptop and testing them there. If you get memtest86+ errors on the other laptop, then that's a good sign that your modules are defective. Of course there can be some other problem with your X31 motherboard but I think you need to figure out if your current modules are working 100%.

[riksarider]

I see. I did a research and those BSODs mentoined in my first post usually point to faulty RAM (the last one had a called address 0x00000004, which I think is at the very beginning of the memory pages).

Since my previous update, I've ran Memtest again third times. First time, it give 436 errors (and thousands of errors in test1, test2, etc. column) and froze after 30 seconds.In the second, it gave no errors, just froze at 20 % of first pass.

Should bad RAM cause it to be frozen ?

Most important question is: can bad RAM cause blank screen at power on, even occasionally ?

Yes, I'm afraid it could be a problem with the board itself (the memory slots or maybe the mem.controller ?)

And I'm still very puzzled about those 4-4-4-4 long beeps I occasionally get when turning it on.
If it's the security chip (TPM), then why would the machine turn on again after a few tries ?
(Well, I've just read some cases with the same number of long beeps on ThinkPads and many of them were gone after reseating/replacing RAM.. :O )

Unfortunately, another laptop or another memory module(s) are not available for testing at the moment.

Thank you again, best regards,
Gabriel

[dr_st]

Yes, bad RAM can pretty much give all the above symptoms, and it can also be a bad board. IMO, that's unfortunately even more likely given the combination of issues (I once had a bad X3x board that behaved in similarly weird ways).

There's not much to do, really. One thing (if you haven't yet) is to thoroughly clean the contacts in the memory slots with good quality alcohol. But the odds, IMO are slim that it actually fixes the problem. Other than that, it's a hardware problem that you cannot fix short of replacing the hardware.

If you're lucky - it's the RAM, if you're unlucky - it's the board. But until you test it somehow - no way to know for sure.

[riksarider]

Well, I've just had an "Unexpected interrupt - Halting CPU0" error & halt on Memtest86, with 11 errors before that occurred (it ran for about a minute, at Pass 0).
This is new for me..

(And now, at another run, another of the error above "Unexpected interrupt", but this type 0 errors reported and Type is 'Inv_Op' instead of the previous 'Debug'. I'm over my head with this...)

Contacts seem very clean at the RAM slots (at least, to the naked eye)..

Yes, bad RAM can pretty much give all the above symptoms, and it can also be a bad board. IMO, that's unfortunately even more likely given the combination of issues (I once had a bad X3x board that behaved in similarly weird ways).

There's not much to do, really. One thing (if you haven't yet) is to thoroughly clean the contacts in the memory slots with good quality alcohol. But the odds, IMO are slim that it actually fixes the problem. Other than that, it's a hardware problem that you cannot fix short of replacing the hardware.

If you're lucky - it's the RAM, if you're unlucky - it's the board. But until you test it somehow - no way to know for sure.

[Neil]

Test the RAM sticks one at the time, in each of the slots, to isolate the faulty part.

[riksarider]

Got another RAM (1x512 MB module) stick for it...at first, it turned on instantly and booted into Windows XP flawlessly, had used it for about an hour from battery with Wi-Fi on for surfing, no errors/BSODs at all (unlike the past 2 modules).

However, when I got back home, it refused to turn on.
I removed the battery and AC and performed the "capacitor emptying procedure" (or what...you know, remove all power, then press power button several times, the hold for about 30 seconds, reconnect power and turn on). It seemed to be working, as after it, the machine turned on and went to the BIOS (following some complains for date/time and CMOS settings). I set the date and time and booted to Memtest86. It's still running (07:30 minutes by now) and so far it found 26 errors (confidence value 127 and max. contigous errors 1). They are in Test 4, 6, and 7.

But...it haven't stuck or hung yet, unlike with the other two modules. (Actually, it was rare not to get a freeze with them during Memtest.)

I should try to boot into Windows XP again after it finishes 1-2 passes, but this already bothers me.
Is it possible that I have a faulty board or CPU ?

Thanks again, Gabriel

[Neil]

Is it possible that I have a faulty board or CPU ?

Possible? Sure, but highly unlikely for the CPU. Could be the system board though. Only way to know for sure is to test all your ram modules in a different machine. If none throw any errors in the other machine, but all show errors in this X31....at least that rules out the RAM.

[riksarider]

Well...though Memtest shows some errors, XP ran without BSODs so far.
Note: with the 512 MB module, the lowest and highest error address is both at 124.1 MB.

Then I went and put another one of the old -supposedly erroneous- 256 MB modules in next to the 512 MB module, to see what happens.
I had Memtest tu run for 3 passes: lowest error address is at 380.1 MB (which exactly 124.1 MB + 256), highest is at 508.2 MB.
I guess that means that both modules have some error. Total number of errors was 321 in 3 passes.
(Max contiguous errors 1)

I think I finally found the right configuration (they are both Infineon PC2100 DDR modules, manufactured at about the same time/week in 2003), as XP runs without bluescreens and become quite snappy with 768 MB instead of 512 MB (obviously).
I guess that's fine by me and finally I can use my machine to work without BSODs.

However, one final question remains: sometimes after a long rest, the laptop refuses to turn on (blank screen, but cooler fan and lights are on, as is HDD).
Then I need to remove all power sources and push the power button a few times, then put the battery and AC back and it starts.

What could be the cause ? I've read that this is quite common in ThinkPads..

Many thanks again.

Is it possible that I have a faulty board or CPU ?

Possible? Sure, but highly unlikely for the CPU. Could be the system board though. Only way to know for sure is to test all your ram modules in a different machine. If none throw any errors in the other machine, but all show errors in this X31....at least that rules out the RAM.

[dr_st]

I don't like to be the bearer of bad news, but I'm pretty sure you have a faulty board. It may work well for a while, but eventually the issues will resurface. As I said, I had experienced most of your symptoms with my faulty X32 board. At some point it died and would no longer POST, at all.

Keep using it as long as it works. If and when it stops working - then you consider your options and decide what to do. .

[riksarider]

Well, the symptoms returned soon after, despite replaced RAM.

Weird...sometimes it just turns on with no screen/beeps at all, sometimes it gives 4-4-4-4 beeps (also, with blank screen).
Sometimes it gets to WinXP, then freezes totally in a few minutes and sometimes it runs for hours without problems (both on battery and AC).
Looks like I have a faulty system board, but since Memtest still shows errors in the new (used, but new to this machine) module, I will at last try a brand new RAM there.

It's even weirder that I usually only can get it to work (e.g. boot with display on), when I'd swap the two RAM modules previously..