Thinkpads Forum

hey. i'm having some weird ram issues with my 600 system. i just bought two 128mb 144 pin pc66 sodimm chips. i tested both pieces one at a time in the first open slot (i'm assuming the first slot is the one dirrectly above the built in 32mb ram) and they both show 160mb which is correct if you add 32 and 128. when i put both pieces of ram in at the same time, i'm only showing 224mb. that second ram slot is only recognizing 64mb of the 128mb chip. i've tried putting only one chip in that second slot at a time and still only shows 64mb. even when nothing is in that first slot. it does the same thing for both chips, no matter the order, when they are both in. am i missing something crucial here? i though for a minute that i was maxing out the system, but according to everyone i talk to and IBM (Lenovo now) says 288mb is the max for this sytem. so that is not the problem. is there a setting i have to change in the system or what? help please!

update!!!!!!!!!!!!!!!
i checked the bios and it says

Installed 294912 KB
Usuable 228800 KB

that shows all 288mb being recognized. how do i make it all usuable?

It looks like all of your RAM is working. There will always be a few megabytes that's not usable that the system reserves for the hardware and BIOS and the rest which is the majority of the RAM goes to the software. When the amount of RAM usable and installed differ by more than 8mb, that's when I'd start questioning things.

that's the problem. there IS a difference of more than 8mb. it's a whole 64mb! and with a system that only supports 288mb, 64mb is a pretty big chunk...

I think it's just because of the 2nd memory slot, which shares the capacity
with the internal 32MB RAM. So the BIOS can recognize 128MB,
but only half the capacity - 64MB can be actually used.

32+128+32=192MB
32+128+64=224MB (228800kb)
32+128+128=224MB (228800kb)

It's common with some early models of MMX/P-II laptops which have
the 430TX (not 440BX) chipset installed on their motherboards.
770(430TX) and 770X/Z(440BX) have the same problem.

I'm not sure, but I think if you rip out the internal memory or disable
the function of it, maybe you can make the 128MB RAM fully functional.
Just my own guess.

thats what i dont understand... it has the newer i440BX chipset. i wouldnt think that would be a problem...

Having almost identical issue! I recently bought 2 sticks of 128mb pc-66 ram. I know the ram is good - used for a while ( individually!) in my c240 no probs at all. Installed both sticks in my 600E(2645-AL3). Most of the time it boots fine recognizing all 288mb ( well -bios reports slightly different but close enough!! ) However a couple of times - only about once every 10 days approx - i've switched on and noticed bios screen showing 16xxxx approx ram ( i'm at work - cant remember exact figure!!) Basically meaning one of the slots hasnt been recognized at all. My solution was a switch off - take both sticks out - swap sticks over and reinsert! Switch on and all fine again for a few days. I've tried all the usual cleaning contacts / blowing air etc - reckon the contacts / fit on one (both?) of my memory slots is slightly "iffy"??? As I say - it only seems to happen quite rarely - and I reckon I can live with it. But from browsing forums it seems that 600 series and memory arent always the best of friends!!!! (LOL)

Just remove the internal memory, then you'll see the result!

Yeah, better to be down 32mb instead of 64mb.

shirleywhite wrote:Having almost identical issue! I recently bought 2 sticks of 128mb pc-66 ram. I know the ram is good - used for a while ( individually!) in my c240 no probs at all. Installed both sticks in my 600E(2645-AL3). Most of the time it boots fine recognizing all 288mb ( well -bios reports slightly different but close enough!! ) However a couple of times - only about once every 10 days approx - i've switched on and noticed bios screen showing 16xxxx approx ram ( i'm at work - cant remember exact figure!!) Basically meaning one of the slots hasnt been recognized at all. My solution was a switch off - take both sticks out - swap sticks over and reinsert! Switch on and all fine again for a few days. I've tried all the usual cleaning contacts / blowing air etc - reckon the contacts / fit on one (both?) of my memory slots is slightly "iffy"??? As I say - it only seems to happen quite rarely - and I reckon I can live with it. But from browsing forums it seems that 600 series and memory arent always the best of friends!!!! (LOL)

Have EXACTLY the same problem with my 600E (2645-4BG).

Got this 600E May 2005 with 2 x 64MB memory modules.
With the onboard 32MBs i should have had 160MBs of memory in total.
Within a few weeks of having the 600E i noticed that sometimes it would only show 96MBs total memory.
After a while i noticed that once i'd got all memory recognised - 160MBs total - it wouldn't fail again til i moved the laptop.
I'd then need to go thru the process of reseating the modules to get them both to register.
I upgraded one of the 64MB modules to a 128MB module and the problem persists - mostly it recognises 224MBs but sometime 160Mbs.
When a module is not recognised i find the BIOS has detected it and the total installed memory is always correct with the useable memory showing a module missing.

I've tried swapping my 128MB and 64MB modules around but then found a misisng module happened more frequently so put them back to 128MB in slot 1 and 64MB in slot 2.

I suspect that picking the 600E up inevitably leads to the memory flap door getting a fair amount of pressure on it and a module is becoming 'slightly' unseated and causing the problem.

This would explain why the module only goes missing after the 600E has been physically moved.
But it doesn't explain why when the module is not usable it's still detcted by the BIOS.

In the past couple of weeks i tried another solution...
I disabled the 600E's onboard 32MBs memory and tried various combinations of my 128MB and 64MB modules.
It made no difference - sometimes a module would be unrecognised regardless of whether the onboard memory was enabled or disabled.

I get both modules working now and am extra careful when moving it not to pick it up with any pressure on the memory flap door - but still once in a while a module goes undetected.

warwound

I'm now getting occasional non-recognition of one of my 128mb sticks.(rather than half amount!) Out of habit now I always look out for amount of memory which bios reports. Sometimes it's 160mb - sometimes 288mb ( ie the 6 digits near equivalents!) If I open the memory door -take out the sticks and swap slots - every time so far the next boot has reported full 288mb.So like warwound am suspecting a loose slot/connection (and I suspect you are correct that moving or pressure could be cause - or at least contributing etc )

Wonder if "padding" between memory chips and the "door" would help at all? And if so what material? ( I once had hard drive in a box which only seemed to work with a sheet of cardboard underneath it in the box!?) Of course it could be that packing/padding should be "under" the memory sticks rather than on top.?? Any thoughts please?? ( Is heat an issue with memory sticks? Presume not?!) Or should I just leave well alone and accept the slightly random memory situation on my machine?

Your idea to use some sort of padding under and/or above the SODIMMs is a good one.

I think heat may be an issue though.
For the padding to stand a chance of being effective it'd have to fill out the memory compartment height ways.
And then with no air surrounding the SODIMMs i think heat would be a problem - the padding may even be insulative which would perhaps guarantee a heat problem!

I bet there's some suitable material that could effectively pad the memory compartment though and will watch this thread for any good suggstions.

warwound

Interesting thought. I might try to construct something for the TP600 that I have. I am picturing one maybe two thin strips of something that would add the thickness to keep the ram in place, but not too wide, so that there is still room for cooling.

tfflivemb2 wrote:Interesting thought. I might try to construct something for the TP600 that I have. I am picturing one maybe two thin strips of something that would add the thickness to keep the ram in place, but not too wide, so that there is still room for cooling.

An interesting thought indeed - might be of use to others with this problem (not me, thankfully).

There may be another avenue to explore - could it be that the two little clips that lock the SODIMMs into place are somehow deformed? I would think that if this were the case, the SODIMM in the affected slot might wiggle a little more than in in the other one.

If memory serves, the locking tabs move sideways relative to the modules, but there's a dimple of sorts at the end that seems to exert downward pressure (push them aside, and the module pops up so you can grab it to remove it). If this "dimple" were not deep enough, it might not be exerting sufficient pressure, and causing some sort of intermittent contact.

To verify, you could try putting a small bit of black electrical tape over the spot where the clip rests (only on top though). A layer or two should be enough to slightly increase the pressure exerted by the hold-down. Fire the machine up, and try your "move test" to see if the intermittency is gone. If nothing else, this would at least not impair cooling, since the area covered by the tape would be very small.

Rob

I have the same problem, except it stays at 224Mb. Never changes. I also have the ram cover off, no effect. I'm using two sticks of Toshiba 128MB PC133. Now I've just gone back to 1 stick of 64mb pc66 IBM and 128MB pc133 Toshiba, since I can't get past 224MB. I wish I could use the other 64MB though.

How does IBM expect us to get that full 288MB?

socalismylife wrote:I have the same problem, except it stays at 224Mb. Never changes. I also have the ram cover off, no effect. I'm using two sticks of Toshiba 128MB PC133. Now I've just gone back to 1 stick of 64mb pc66 IBM and 128MB pc133 Toshiba, since I can't get past 224MB. I wish I could use the other 64MB though.

How does IBM expect us to get that full 288MB?

Are you using these PC133s on a TP600? (I assume yes) Part of you limitation could be the result of using the wrong memory. The memory should be PC66 or PC100, even though there are some PC133s that are low density and backwards compatible (such as yours), but they will some times effect the full amount of ram.

Quick update. I padded out the space between ram sticks and door with a couple of business cards. Placed vertically "across" the memory they just fit the compartment. When I tightened the screw down - could feel a slight pressure being exerted from the last few turns. Switched on and it booted the full 288mb. Have rebooted a few times and (so far) havent seen the "non-recognising " problem ?!! But of course it was intermittent before- so no real way of knowing if the "pressure pad" is doing anything or not!! ( I'm hoping/thinking that heat woudnt be too much of an issue?!)

Will keep using like this and report again in a few days!

Make sure that there is still enough room to keep the ram chips cool, or else you will run into a new problem. (ie. system overheating, dead ram)

This is why I recommend trying to create narrow strips, to allow for "breathing room".

Thanks for reply tfflivenb, Rebooted tonight and only the 160mb showing!! So this time have left the sticks in same slots but put a small strip of card under the modules - so when press down into place a reasonably tight fit. First reboot showed 288!! Appreciate your comments re heat etc Have used quite a small strip ot card - and thus seems to be quite a lot of clearance around the bit of card - but does the job of "packing" the width!!! Will monitor and report back in a day or two.

Three days and about a dozen reboots - every time so far bios reports full ram!!! ( 288mb) Fingerscrossed but maybe my particular ( loose slot/connections) has been solved! Small strip of card UNDER the memory sticks - press down into clips - a slightly "tight" but solid fit but clicks into place ok. ( Must mean that the connectors on the ram are pressing slightly tighter on the "upper edges ie to outside of machine.) Thanks for replies and suggestions all. Will report again if anything else changes!!

Awesome, I should actually try this now on the TP600 that I have, so that I can sell it with 100% working ram.

Well i've just done exactly the same mod to my 600E.
Small strip of cardboard under the 64MB module - left the 128MB module as it is as it always detects anyway.

First reboot and back to 224MBs - it's been booting with 160MBs for a week or so now and i've not been bothering to coax the 64MB into being detected as it only detects for a boot or two then disappears.

Tried a reboot and again 224MBs reported.

I'm going out with my 600E today so it'll be a good test.

Will post again with my results.

warwound.

Two days of use and i have good news to report.

The original mod using a single thin piece of card seemed to work but after a few reboots my 600E booted up with 160MBs of memory instead of 224MBs and i was thinking that the mod had failed....
I got some more card and made a narrower strip but this time i folded it double - twice as thick as i'd tried previously.
Made sure it wasn't too thick and that it wasn't obstructing anything and re-installed the SODIMM.
Two days later and my 600E has booted with it's full 224MBs of memory each and every time!!!!!

Ye hah!

Very pleased so far - another week of some real tests, mainly involving moving my 600E from one location to another and if it still works ok then i'll consider it fixed.

Many thanks for the inspiration from this thread.

warwound.

Warwound ( and others) - after a week of reboots showing full ram I switched on tonight and got the old 160 I'm afraid!! I certainly think we were on the right lines with the "loose connection/socket" idea. I've just "repacked" with a very slightly thicker piece of card and first boot is back to 288. Hope this works = or at least lasts longer than last attempt. Good luck.

Well after a week or so i have similar results...

Mainly my 600E is booting with it's full 224MBs of memory but occasionally it only boots with 160MBs.
Again it seems to be moving the 600E from one place to another that causes the second module not to be detected.

Yesterday i got a piece of foam - it was part of the packing for a Radeon 9550 card - the foam is about 8mm thick and lightweight stuff.

So this morning i booted my 600E and it booted with 160MBs of memory.
I cut two small strips of foam and placed one under the non-detected memory module and one above.
The module is therefore padded on both sides.
I used thin strips of foam and don't think that heat will be a problem.

First two or three reboots after using the foam and it's booted with 224MBs each time.
But i have yet to test it after moving the laptop from one place to another.

Shall keep this thread updated with my results.

warwound.

Another quick update ( mainly for warwound!). After repacking with slightly thicker card I have so far had "full reboots" ie 288mb. I have moved the machine around quite a lot - not deliberately slamming it down(!) - but certainly it's had a few gentle taps etc. So far seems to recognise the full amount of ram. I reckon in both our cases it was almost certainly a slightly loose ram socket and that somehow we are now making a better connection. For me it's certainly an improvement - and if I now get just an occasional "loss of memory" I'll bear with it - otherwise I'm very happy with my 600E!!!!!

shirleywhite wrote:Another quick update ( mainly for warwound!). After repacking with slightly thicker card I have so far had "full reboots" ie 288mb. I have moved the machine around quite a lot - not deliberately slamming it down(!) - but certainly it's had a few gentle taps etc. So far seems to recognise the full amount of ram. I reckon in both our cases it was almost certainly a slightly loose ram socket and that somehow we are now making a better connection. For me it's certainly an improvement - and if I now get just an occasional "loss of memory" I'll bear with it - otherwise I'm very happy with my 600E!!!!!

If this continues to work well, perhaps you would be kind enough to take some measurements on the strips you cut, and perhaps an idea of the type of card you used? I see a how-to in the making here, that should probably be added to the "600 Upgrades and general Info" thread if it can be proven to work consistently.

Now, at the risk of stirring the pot some more, the engineer in me popped another thought on this issue while I was reading the thread this morning: what if we're not seeing a loose memory socket, but one that's slightly out of tolerance on the large side? Manufacturing processes aren't always perfect, and if the socket is just a hair too large, a memory module that's slightly thinner than average might need "help" to make proper contact.

Of course, to lend credibility to this idea, we'd have to show a correlation across the units exhibiting this process (similar build window, same parts lot for the affected slot's connector, etc.).

Rob

Well still good news with my foam padding - not a single boot-up so far with anything less than all my 224MBs of installed memory being detected.

So for anyone that's interested i've taken a couple of photos of my mod.
Had to reduce the image sizes to make them suitable for a quick download on the internet but they show the dimensions of the foam strip which i have used.

[edit]image links dead so removed[/edit]

The images are obviously the strip of foam that i placed across the top of the module - a similar strip of foam also pads the module underneath.

Laptop running perfectly - an AV scan and some other stuff have had it working 100% for a while and it's showed no problems (ie heat build up on the padded memory module) so far.

warwound.

Thanks to Warwound for the pictures. I've used card rather than foam. I cut the ends off a couple of my business cards. The dimensions look almost exactly the same as warwound's foam ( in terms of width and length sitting across the memory module.) In terms of thickness - I originally tried just one thickness of card but upped it to 2 pieces. For me I placed it UNDER both modules then gently pushed down until it clicks into place via the side clips. I have no padding etc on top of the modules so no pressure from door when rescrewed down. So far this seems to be working!!! Regards all.

Well all good things (apparently) come to an end so i must report my thinkpad reverting to only detecting one of it's installed memory modules.
It's been booting with 160MBs instead of 224MBs for a few days.

So this morning i made a change.
Same foam mod as before but this time i made the foam under the module twice as long and folded it lengthways to make it twice as thick.
I was planning to do the same with the foam strip above the module too but there was just not enough free space to pad it any more!

I've been out with it all afternoon and it's been running perfectly - 224MBs since i reseated the module earlier.
Only using standby whilst out so i didn't get to see if both modules would remain detected if i rebooted.
Had a good 2 or 3 hours out of my fully charged battery and then it was flat, i returned home and plugged it in to charge.
Rebooted and still i had a full 224MBs of memory.
Fully charged i switched it off a bit later.

Needing my afternoon's work i rebooted a short while ago and....
Just one memory module detected - 160MBs.

I've now made one last change to my foam before i more or less give up the concept of always having both memory modules detected.

This time i have the memory module padded underneath as before - a single thickness of foam.
And i've now padded the memory module above with the longer strip of foam folded lengthways to double it's thickness.

First reboot or two have worked but that's no test - i'll be out and about with my 600E over the course of the next few days and that'll be a good test.

Will report my results back here.

warwound.