T23 RAM problem

[esoh]

I have been having similar problems to those described.

I have a T23 with 512MB RAM (2 256MB sticks). Recently (over the past month), I have been having problems. Freezes, screen goes black, system will do 1-3-3-1 beep sequence on startup, etc. Sometimes it starts up fine, works for awhile, then craps out later. Sometimes it doesn't start at all.

I tried reseating RAM, running memtest86+, running on one chip at a time, etc. None have worked.

Looking at this thread, I guess I have some options:

1. Clean the SODIMM connectors and sockets
2. Try new RAM
3. Replace MOBO

I'm also curious about Ray's "fixing" of the SPD information on the RAM sticks. Clearly, I am nowhere near his league on this stuff, but is this something that is fixable fairly easily? I'm guessing not, but just thought I'd ask.

Frankly, I don't know how much more $$ I want to put into this machine. If it's not an easy fix, it's probably going on eBay for parts.

Ed

Moderator note: Split this off from T22 RAM problem

[rkawakami]

I tried reseating RAM, running memtest86+, running on one chip at a time, etc. None have worked.

Can you be a little more specific as to what doesn't work? Did the laptop boot memtest86+ and then the diagnostic say you had a failure or did the system just give you the 1-3-3-1 beep sequence? I'm also assuming for now that you've been using these same two modules for quite awhile and that this problem is a recent event.

<snip>... but is this something that is fixable fairly easily? I'm guessing not, but just thought I'd ask.

Easily fixable? Not from what I can see at the moment. In order to "fix" those modules from Dersu (and a couple of others from mudhead in this thread), I have used an old bench-top memory tester here at work. It can read the SPD information off of the modules and write back to it if needed. I have not yet found a utility program that can do this with a module installed in a system.

Your problem does not sound like the SPD is scrambled since you indicate that sometimes it works. I'd be more inclined to suspecting something on the motherboard is causing this problem. By all means, if you have access to some other PC133 modules then try those. If you continue to get those memory error beeps, then your problem is with your system.

I would definitely suggest cleaning the modules and sockets. The modules are easy; some isopropyl alcohol on a nakpin or paper towel and then rub both sides of contacts on the module or use a pencil eraser and lightly rub down the edges. The sockets are a lot harder. If you can get a folded piece of paper towel (lint-free cloth is best but not everyone has that in their house) wet down with isopropyl alcohol and brushed across all of the contact pins. In both cases, make sure you are properly static-discharged before touching the contacts on the module or laptop. I have handled many memory modules in my job and have never zapped one bad enough to make it fail, but there's always the chance.

[esoh]

Ray,

Thanks for getting back so quickly. Love this forum.

In regards to what didn't work, I definitely should've been more informative. Here is each solution I tried and the result:

1. Move 256MB RAM chips around in various configurations (singly, changed positions, etc.) - The computer exhibited similar problems. The 1-3-3-1 beep sequence didn't show up immediately, but the laptop would freeze eventually.

2. The laptop would boot memtest86+ and cycle through. The two times I ran it with both modules, I received no errors.

Yes, you are correct in your assumption that I have used both RAM modules for over a year with no problems until the past month.

As an FYI, the chips are:
- 1 256MB Transcend PC133 SDRAM
- 1 256MB AMC 55366 chip (no details on stick about speed)

To be honest, I don't remember where I picked up the second chip, but like I said, it had been working fine for awhile.

Thanks for your assessment about the MB. Clearly, I am not as technically oriented as others and was hoping that my problems were representative of what others have seen. I will try cleaning the modules and sockets and see what happens.

Again, thanks for your help. I will post results as soon as I have them.

Ed

[esoh]

Well, I did the suggested fixes for my RAM problem. Here are the results:

1. Cleaned RAM sticks and sockets using isopropyl alcohol and lint-free clothes.

Result: The computer worked ok for awhile, and then I had the same old problems. Freezes, 1-3-3-1 beep errors, etc.

2. Tried running memtest86+ again followed by the complete PC Doctor diagnostic.

Result: The computer passed all of the tests.

At this point, I think I have it narrowed down to two possibilities:

A. RAM chips are bad.

Thought: Unlikely since the computer was working fine 2 months ago, plus the memory passes memtest86+ tests. If one chip was bad, then the computer still should have been fine when I tested using one chip at a time. I suppose the chances of both RAM going at the same time and not being detected as faulty by memtest86+ are extremely remote.

B. MB is faulty.

Thought: This seems like the only answer to me unless someone can convince me otherwise.

I know I'm grasping at straws here, but does anyone have any other thoughts?

Thanks.

Ed

[schen]

Is the a relatively new T23 or something you've had that "developed" the problem?

I had a similar problem with an A31. My brother sent my wife an A31 he had put together. The HDD from the A30 that she had should have run fine in it after some drivers' upgrades, but the machine kept freezing. So I reloaded from scratch, but the machine would freeze at various time. Usually after it warmed up for a few minutes. I tried everything that you did; different RAM, moving RAM around, one socket, 2 sockets, etc..... I even tried different HDDs. I finally gave up and had him send another one. It loaded and has run fine ever since!

Does that mean that your machine has the same issue, no, but it could.... a hardware failure somewhere on the MB can be d(&((^ hard to track down.

[rkawakami]

B. MB is faulty.

Thought: This seems like the only answer to me unless someone can convince me otherwise.

That would be my guess as well. About the only way it could be a memory problem is if the failure is due to temperature. However, that might also mean that there's something else on the motherboard that is temperature sensitive.

You've mentioned that the system can run fine before encountering the error and also that the system will not even boot. When the laptop does not boot properly, has it first been running for some time? In other words, has it already been warmed up?

If so, about the only thing that you can easily do which does not take any special equipment (meters or soldering iron), is to open up the laptop and re-seat the CPU. As you are getting a memory error beep code during those times that the system does not boot, my only thoughts are that it's the memory or CPU causing the problem. And as long as you are in there with the heat sink off of the CPU, you might as well apply some thermal grease (aka Arctic Silver 5 or similar) to the top of the CPU, after cleaning off all of the old stuff.

[esoh]

Is the a relatively new T23 or something you've had that "developed" the problem?

I had a similar problem with an A31. My brother sent my wife an A31 he had put together. The HDD from the A30 that she had should have run fine in it after some drivers' upgrades, but the machine kept freezing. So I reloaded from scratch, but the machine would freeze at various time. Usually after it warmed up for a few minutes. I tried everything that you did; different RAM, moving RAM around, one socket, 2 sockets, etc..... I even tried different HDDs. I finally gave up and had him send another one. It loaded and has run fine ever since!

Does that mean that your machine has the same issue, no, but it could.... a hardware failure somewhere on the MB can be d(&((^ hard to track down.

It's not a new T23. I've had it for about 2 years, bought it used previously. I added the RAM and have been using it fine up until a couple of months ago.

In your case, when you say that "you had him send another one," was it a HDD, RAM, or computer? Just curious.

Ed

[esoh]

B. MB is faulty.

Thought: This seems like the only answer to me unless someone can convince me otherwise.

That would be my guess as well. About the only way it could be a memory problem is if the failure is due to temperature. However, that might also mean that there's something else on the motherboard that is temperature sensitive.

You've mentioned that the system can run fine before encountering the error and also that the system will not even boot. When the laptop does not boot properly, has it first been running for some time? In other words, has it already been warmed up?

If so, about the only thing that you can easily do which does not take any special equipment (meters or soldering iron), is to open up the laptop and re-seat the CPU. As you are getting a memory error beep code during those times that the system does not boot, my only thoughts are that it's the memory or CPU causing the problem. And as long as you are in there with the heat sink off of the CPU, you might as well apply some thermal grease (aka Arctic Silver 5 or similar) to the top of the CPU, after cleaning off all of the old stuff.

On the boot issue, I can usually get it to boot after reseating the RAM. After I do that, it will go OK for awhile. At some point, it will usually freeze and when I try to restart, I will typically get the 1-3-3-1 beep sequence. At this point, the only way to get it restarted is by reseating the RAM again. As you can imagine, this is a tedious process.

I have reseated the CPU on an old TP 23 before, but frankly, I don't know if I really feel like going through the process again. It might be time to sell this one for parts and see what else is out there. Maybe the T42? I love the T23, but I think the platform is showing its age now by all of the posts I am seeing about RAM problems, connectors, etc.

Ed

[RealBlackStuff]

My $0.02:
you may have an overheating problem, either from the CPU, or there is a bad solder point on the mobo. Something bends/expands after a few minutes...
I've seen posts about coils falling off or hanging loose, yours might have that problem too.

[schen]

In your case, when you say that "you had him send another one," was it a HDD, RAM, or computer? Just curious.
Ed

It turned out to be on the motherboard.

[hiero]

...this is a known problem with many Mobo's on the T23.

I have the same problem and living with it is a lottery.

I have a 256 and a 128Mb and no matter what combo, single or pairs, you get issues. It seems to want to always default to lowest stick value if being a problem on boot.

Mostly, I get blue screen of death Memory Page dump errors come up during memory intensive tasks, and after 2 years I know I have to hit the power, give the keybord/base a tight squeeze between my finger/thumb at the SHIFT key sorta area and hope all barks up seeing the full quota.

I have found inserting either paper padding or latterly cocktail sticks ACROSS the length of the dimms and then compressing the cover onto them gives best results.

You know, surely IBM should have sold silly niggles such as these by the time they got to the T series!!!??? Jeez....

[esoh]

...this is a known problem with many Mobo's on the T23.

I have the same problem and living with it is a lottery.

I have a 256 and a 128Mb and no matter what combo, single or pairs, you get issues. It seems to want to always default to lowest stick value if being a problem on boot.

Mostly, I get blue screen of death Memory Page dump errors come up during memory intensive tasks, and after 2 years I know I have to hit the power, give the keybord/base a tight squeeze between my finger/thumb at the SHIFT key sorta area and hope all barks up seeing the full quota.

I have found inserting either paper padding or latterly cocktail sticks ACROSS the length of the dimms and then compressing the cover onto them gives best results.

You know, surely IBM should have sold silly niggles such as these by the time they got to the T series!!!??? Jeez....

Thanks for this post. I tried the squeeze in the DIMM slot area this morning and got it to boot, immediately after I had the 1-3-3-1 beep problem.

I'm going to try the paper padding trick tonight and see if I can get it to work more consistently. Are there any potential fire hazards or heat issues I should be worried about with this strategy?

Ed

[hiero]

No, I have used this method on this and other laptops in past without problem...

To explain the method of the cocktail sticks further, my connection seemed to have to be enhanced across the row of pins and I used the stick as a spring clip to this effect both ends tucked under the retainging clips and then the cover slammed onto them. Its not totally solving the problem but a lot better than it was.

I too thought that maybe just the single stick (512Mb) would solve the problem, but as yet I have not tracked one down on eBay UK for anything under the equivalent of your $80-90 !! (about £45+) A bit too much under the circumstances. Other than that, its not a bad Laptop. BUT the T42 is so much nicer in every aspect, and I'm jealous of my Partner for this!

[esoh]

No, I have used this method on this and other laptops in past without problem...

To explain the method of the cocktail sticks further, my connection seemed to have to be enhanced across the row of pins and I used the stick as a spring clip to this effect both ends tucked under the retainging clips and then the cover slammed onto them. Its not totally solving the problem but a lot better than it was.

I too thought that maybe just the single stick (512Mb) would solve the problem, but as yet I have not tracked one down on eBay UK for anything under the equivalent of your $80-90 !! (about £45+) A bit too much under the circumstances. Other than that, its not a bad Laptop. BUT the T42 is so much nicer in every aspect, and I'm jealous of my Partner for this!

Sorry, but I'm a bit dense. Can hiero or anyone else help...are you trying to push the DIMMs down into the MB or trying to sit higher? Based on this post and posts from other threads, I guess you are trying to get them to be flatter to the MB, but just wanted to make sure.

Ed

[frankie66]

My T23 worked fine for 3 years, both DIMM slots were occupied with 512M cards (1GB) and life was good. Then I decided to sell the machine. I decided to pull out 1 memory card and leave the machine with just 512M, people on eBay don't pay much more for the half gig of memory and I could use it elsewhere. Then I formatted, reinstalled etc. After that the machine started to randomly freeze, or make funny patterns on the screen. I also noticed the CPU was running much hotter (well it felt like I could dry my hair with the hot air the fan was pumping out). So I tested it to death, I took out the CPU fan cleaned it re-seated it, low level scan the HD etc etc. I passed all the tests from PC doctor but kept freezing randomly. In a last ditch attempt I put back the 2nd 512M card and wow everything was fine. Then I figured it out, I again took out the 2nd memoy stick but then I moved the 1st stick to the top DIMM slot hey presto it was still fine with 512MB. My conclusion is that the bottom slot has a bad connection, and from I read here it may be a design flaw. Do you guys agree? The only other thing I could think of is that the there is something special about the top slot and that must be used if you only have one stick, but I found no doc to suggest this is the case. Also now the stick is in the top slot the CPU is not getting hot anymore. Diagnosing these problems is so hard and time consuming. Everyone suggested ACPI and the harddisk as culprits and I must have scanned the drive a dozen times. I am so grateful forums like these exists and thank everyone who takes the time to write down their stories it's a big help.

[hiero]

Actually, my advice on putting pressure on the Dimms, is actually into MBoard for me, but it may be away for you, or one one way one another.

You need to sit with the laptop on your lap, side on, screen half cocked, Dimm cover off, on Battery power and see what works best on boot.

Boot into BIOS to check the Memory POST result...

[hiero]

...years on! Followup was that the board failed around the SoDimm sockets and had to resort to a new board. Laziness and the advent of newer technology meant I haven't gotten round to fitting it 3 years on! I do however have much use of my older 760ED...