T42 shuts down when i move it

Message

rahlquist · #181 Post by **rahlquist** » Tue Sep 25, 2007 11:14 am

So reading my "Make" news feed I see reference to this page;
http://geektechnique.org/projectlab/726 ... ard-repair

And LO and behold the infamous Ibook video problem is based on the same cruddy BGA package used in the T41. What a shock.

rahlquist · #182 Post by **rahlquist** » Wed Sep 26, 2007 4:58 pm

And lastly they ignored all the evidence I provided and just up and closed the ticket. How nice.

#183 Post by **rkawakami** » Wed Sep 26, 2007 5:45 pm

Sorry to hear of your denied request at repair. The BGA package itself is not the sole cause of the problem. And trying to stay away from gear which uses BGA components is probably going to eliminate 99.99% of the modern laptops, as well as most consumer electronics. What seems to be the issue is the use of lead-free solder. This change was mandated several years ago for health and environmental reasons and apparently the first compounds are/were brittle enough that flexation becomes a real issue. Also, there is a problem known as "black pad", which is an oxidation layer that forms between the landing pad on the motherboard and the component's legs. I've noticed this on several T23 components I have re-soldered. The end result is that after some period of time, the oxidation (and probably some flexing, from physical and/or thermal sources) causes intermittent or complete disconnection of parts from the printed circuit board (PCB). The BGA packaging adds to the problem because those weakened solder bonds are hidden beneath the part AND those packages are more rigid than the PCB material it sits on. Thus, the board bends but the BGA package does not. There has been some attempts to minimize the flexing by putting epoxy balls at the corners of the BGA package, between it and the PCB.

Google: "lead-free solder brittle" and "lead-free solder black pad" for more information.

rahlquist · #184 Post by **rahlquist** » Tue Oct 02, 2007 12:01 pm

Eh what can you do, whats sad though is while most computers have a definite EOL (End of life) this little thing kept chugging before this. I have no doubt the heavy use I have put on this thing contributed in some way and even possibly the effects of living in the south where high humidity can really help oxidation along. I wish we had extended the warranty. 3 years for a laptop to be disposable is a bit much for me, heck my wife still uses the Gateway Solo 1450 I bought her before our daughter was born 6 years ago. Trust me its been through the wringer too and I have gone though no less than 3 power supplies and 6 power jacks. But for a celeron 900 it does a good enough job for her.

While BGA itself may not be the fault, the fact that this particular BGA part is failing cross brand would seem to indicate a larger problem that the manufacturers have just managed to be lucky enough to skirt. Since the whole purpose of the lead free solder was to reduce lead in the landfills and cut disposal costs, letting this problem go uncorrected may have a minor backlash effect.

Kidane · #185 Post by **Kidane** » Wed Oct 03, 2007 10:35 am

rahlquist wrote:So reading my "Make" news feed I see reference to this page;
http://geektechnique.org/projectlab/726 ... ard-repair

And LO and behold the infamous Ibook video problem is based on the same cruddy BGA package used in the T41. What a shock.

I think I may have mentioned this on an earlier page, not sure now as it's been several months.

I bought a T42 with the intention of fixing it, attempted to reflow the solder but it was never reliable enough for me and I sold it to another thinkpads.com member (represented properly of course).

What is interesting is that my girlfriend's old iBook G3 had a similar problem and she gave it to me to see if I could fix it as she has a Macbook now. I reflowed the solder in the same way as I did with the T42 and the laptop has been working flawlessly for about four months now. Granted it's too slow to be of much use, but there it is.

Of note is the fact that on the iBook, the Radeon chip sits much closer to the edge of the board. On the T4x it lies nearly directly in the middle of the system board. Additionally, the iBook is a 12" model, whereas the T42 is a 15" model.

My opinion, contrary to what I posted earlier, is that this issue is *not* heat related, but caused by bad solder, exacerbated by board flex. The "long fan" assembly on the T4x puts pressure on the GPU, possibly holding it in place longer. Perhaps the 15" LCD models experience the problem more, since obviously the board will flex more with a larger chassis.

That said, I picked up a pristine T41p which I am much happier with. It is still under warranty, which I will naturally be extending.

Final thoughts, for those of you attempting to reflow the solder with a butane heat gun, don't be shy! I tested the gun on an old circuit board and it takes a lot of heat before the solder loosens up enough to shift the component any significant amount. I believe it is much harder to burn up the GPU than people think, so if you want any chance of success, be sure to get the chip nice and hot. Keep the gun moving of course, but you are going to need to keep it on the GPU for some time to have any effect on the solder.

Good luck!

chan_man · #186 Post by **chan_man** » Fri Oct 12, 2007 7:32 pm

Just a question on this topic.. first, do we know when the T4x series machines got coverted to no-lead? I would be surprised that this change happened more than a year ago. I am sure the T6x series is no lead...

For leaded (Pb) solders, it is relatively easy to put the entire board into a standard reflow oven , just take off any rubber bumpers or covers, kapton tape (clear yellow tape) will survive the reflow as well as any plastic that is part of the electric connectors. It is probably the same if you use a rework tool as long as you have the right tip. The best would be to take the GPU off, then reball the part and re-attach.

If these were lead free, then there are many possible causes of crack solder joints and most of the causes are not mechanical. Lead-free solders are more stiffer so for the same geometry of a leaded (PB) joint, it will take less mechanical stressing to crack it. Also, if the parameters of the soldering process is not done correctly, the life of the joint will be short. Unfortunately, most Contract Manufacturers use their own "special" settings so if there are more than on CM making motherboards for Lenovo, they will no behave the same.

Black pad is just one of the problems croping up, stress corrosion of the solder joint is driving lots of research. Most of these problems will also happen in Leaded solder systems, but were figured out many years ago, we are just going through the same problems now with Lead Free.

This is the major problem in the assembly industry.. learning the solder all over again. The industry is spending millions on getting this right.

kangazoom · #187 Post by **kangazoom** » Sat Oct 20, 2007 1:24 pm

how annoying.
i just started having the same problem a few days ago (t42 restarts when i move it).

unfortunately, i have no soldering capabilities, so i guess it looks like i will have to purchase a new laptop (t61?) with money i don't have.

Jeff85 · #188 Post by **Jeff85** » Sat Oct 20, 2007 4:17 pm

I'm looking to try the heat gun method to fix my T41. However, I'm not sure how much of the motherboard I should cover with aluminum foil. Varying sources show different amounts. This guy on you tube only covers certain parts of his motherboard it looks like:
http://www.youtube.com/watch?v=vR8L3B3eDr0

But another person in this thread basically covered everything except the GPU and a couple of things near it.

willpower101 wrote: Insulate everything with foil and peel off any surrounding plastic. I found that just curling some foil over the little plastic connector kept it plenty cool. This pic is from my first unsuccessful run, later, I cover everything but the video chip.

Will it make a difference either way?

strawf · #189 Post by **strawf** » Tue Oct 23, 2007 3:25 am

I managed to fix my broken T40 by putting two 2-euro coins and one 5-fin-mark coin between the gpu and keyboard. It has now worked for at least 2 weeks without any problems. Still Im not very comfortable to use it, since I think it can jam anytime.

EDIT:

After several weeks of use (I think its something like 4-6weeks) my T40 has now jammed twice this week. Its not frequent yet, but I believe the laptop is coming unstable again.

toad · #190 Post by **toad** » Wed Oct 31, 2007 4:55 pm

OK, my "wedging things in between CPU and keyboard" approach is not bad, but certainly not 100%. Got myself access to a heat gun and will fry the motherboard sooner rather than later.

@ Jeff85
Have you tried it yet? I still have to remove my MB, but reckon I will package it the same way the bloke in the youtube video did...

toad · #191 Post by **toad** » Fri Nov 02, 2007 3:19 am

Right, I fried my board completely

by following the exact timing cycles as in the youtube video. Having said that I wrapped my board with aluminium foil as Jeff85 did in the picture above.

I used a 2K Watt heat gun for all those interested.

Perhaps I'll get another dud off ebay and try again...

#192 Post by **rkawakami** » Fri Nov 02, 2007 5:28 am

I would say that the most important item in the re-flow process is a thermometer. Without it, you are guessing at the temperatures you are subjecting the part(s) and board to. My recommendation is a digital meter using a K-type thermocouple or better yet, an infrared "gun" thermometer. They are not that cheap but it beats frying motherboards. In addition to protecting the parts surrounding the GPU, you also need to be sure that movement of the GPU is kept to a minimum otherwise the balls on the BGA grid may smear or mis-align.

toad · #193 Post by **toad** » Sat Nov 10, 2007 6:04 pm

Up and running again with a spanking new mobo which supposedly does not suffer from flexing

Thanks to the strong Euro it was just as expensive/cheap bought from the States as the repaired ones here on the old continent.

@ rkawakami
so what temperature should one go for?

visionviper · #194 Post by **visionviper** » Mon Nov 12, 2007 1:40 pm

Reading the thread, I am a little bit confused. What is the temperature I should shoot for? 200c or 230c?

EDIT: I am wondering because I am pretty sure I need to do this with the second hand T40 I just bought.

#195 Post by **rkawakami** » Mon Nov 12, 2007 3:53 pm

From the information I posted earlier in this thread:

http://forum.thinkpads.com/viewtopic.ph ... 507#237507

I'd say that you're looking at ~230c being the optimal temperature for lead-free solder. There is a possibility of disturbing the surrounding components and thus causing them to become detached from the motherboard. You will need to protect them from the heat source; one of the solutions seen so far is to cover the other parts of the motherboard with aluminum foil.

unitedunited · #196 Post by **unitedunited** » Mon Nov 12, 2007 5:45 pm

Amazing thread.. both for the info contained and that it is almost a year old and still going strong!

I first started having these sorts of problems with my T42p over a year ago, at that time with help from these boards I diagnosed bad memory and replaced my expansion RAM. That seemed to work until this past August, when problems started cropping up again. Thanks to this thread and others I now know what I'm dealing with.

Question 1:

People seem to have used wildly different materials other than actual post-its (aluminum, tin foil, cork, eraser rubber, springs) under the keyboard to apply pressure to the GPU. So I'm confused whether it's ideal to use the metallic (conductive) materials to dissipate heat versus the insulating ones to block it. I'd guess post-its themselves are roughly in the middle.

Question 2:

I'd like to do something more permanent but not quite motherboard replacement (like so many others here, my machine is just out of warranty), and I'm wondering if there's a repair lab out there with access to a reflow oven that will do the reflow at a reasonable cost--I'd rather pay to have someone else do it properly than take the chance that I'll fry some other component with a heat gun I don't have.

Any ideas on either of these questions will be much appreciated.

#197 Post by **carbon_unit** » Mon Nov 12, 2007 8:17 pm

unitedunited wrote: Question 1:

People seem to have used wildly different materials other than actual post-its (aluminum, tin foil, cork, eraser rubber, springs) under the keyboard to apply pressure to the GPU. So I'm confused whether it's ideal to use the metallic (conductive) materials to dissipate heat versus the insulating ones to block it. I'd guess post-its themselves are roughly in the middle.

Any ideas on either of these questions will be much appreciated.

I think the idea behing Question 1 is to apply pressure to the chip, holding it down so it makes contact where the BGA connections have come loose. Heat transfer is a side effect.

Regarding Question 2, My thinking is that the board is already unusable. I can't make it much worse by trying to reflow and maybe it will be better. Good luck trying to find someone to professionally reflow it, but if you do please let us know.

visionviper · #198 Post by **visionviper** » Mon Nov 12, 2007 8:53 pm

unitedunited wrote:I'd like to do something more permanent but not quite motherboard replacement (like so many others here, my machine is just out of warranty), and I'm wondering if there's a repair lab out there with access to a reflow oven that will do the reflow at a reasonable cost--I'd rather pay to have someone else do it properly than take the chance that I'll fry some other component with a heat gun I don't have.

I know there are places that will redo the solder on the GPUs for Xbox 360s. Maybe they will be willing to do it for a laptop.

aaa · #199 Post by **aaa** » Tue Nov 13, 2007 11:19 am

visionviper wrote:
unitedunited wrote:I'd like to do something more permanent but not quite motherboard replacement (like so many others here, my machine is just out of warranty), and I'm wondering if there's a repair lab out there with access to a reflow oven that will do the reflow at a reasonable cost--I'd rather pay to have someone else do it properly than take the chance that I'll fry some other component with a heat gun I don't have.
I know there are places that will redo the solder on the GPUs for Xbox 360s. Maybe they will be willing to do it for a laptop.

There are. I asked some seller on ebay who's offering the service for $50, they said yes. Haven't gotten around to actually trying it though, and I probably need to know what needs reballing exactly (is it the gpu itself or the things on the gpu?).

unitedunited · #200 Post by **unitedunited** » Sat Nov 24, 2007 3:58 pm

update:

I popped the top, hoping to do the post-it fix, and ran into two problems.

One, I have a T42p, hence a long fan, so I can't actually see what's going on with my GPU. I do know that the slightest of touches on the top left corner of the fan-unit copper covering the GPU creates the screen disintegration/blackout. Two, I can't remove the fan assembly. The black thermal goop layered between the GPU and the copper fan hood might as well be super glue. I know from other threads that removing the fan unit shouldn't require very much force, yet it's as if it was installed in such a way as to specifically prevent removal. Is my Fire GL chip just cursed?

strawf · #201 Post by **strawf** » Sun Nov 25, 2007 5:31 pm

strawf wrote:I managed to fix my broken T40 by putting two 2-euro coins and one 5-fin-mark coin between the gpu and keyboard. It has now worked for at least 2 weeks without any problems. Still Im not very comfortable to use it, since I think it can jam anytime.

EDIT:

After several weeks of use (I think its something like 4-6weeks) my T40 has now jammed twice this week. Its not frequent yet, but I believe the laptop is coming unstable again.

Ok. The laptop is now completely unstable(does not handle any movements).

I would like to offer an untested solution for the overheating of the mobo when using heat gun. Put a tiny piece of soldering tin over the GPU so that you can see when the temperature is high enoug for resoldering. This way you can indicate the temperature of the top of the GPU and estimate the required time to heat the pins up to the soldering level.

I would like to hear any results of this method.

Ive also thought about the stresses in the laptop when moving it. Would you think that ultra-tightened screws would be better than just tightened? What about any rubber buffers inside the case to absorb the slight movements of the outer case? This way we could prevent any rebrokeups as the mobo stays stable no matter how you handle your laptop.

aaa · #202 Post by **aaa** » Sun Nov 25, 2007 5:41 pm

strawf wrote:
strawf wrote:I managed to fix my broken T40 by putting two 2-euro coins and one 5-fin-mark coin between the gpu and keyboard. It has now worked for at least 2 weeks without any problems. Still Im not very comfortable to use it, since I think it can jam anytime.

EDIT:

After several weeks of use (I think its something like 4-6weeks) my T40 has now jammed twice this week. Its not frequent yet, but I believe the laptop is coming unstable again.
Ok. The laptop is now completely unstable(does not handle any movements).

I would like to offer an untested solution for the overheating of the mobo when using heat gun. Put a tiny piece of soldering tin over the GPU so that you can see when the temperature is high enoug for resoldering. This way you can indicate the temperature of the top of the GPU and estimate the required time to heat the pins up to the soldering level.

I would like to hear any results of this method.

The soldering tin would melt far before the balls do, wouldn't it?

strawf · #203 Post by **strawf** » Sun Nov 25, 2007 5:42 pm

aaa wrote: The soldering tin would melt far before the balls do, wouldn't it?

Well it could be so, but without any sensors its really hard to know how hot the GPU is.

visionviper · #204 Post by **visionviper** » Sun Nov 25, 2007 11:26 pm

but without any sensors its really hard to know how hot the GPU is.

That is why it is best to just drop $40 or so on a infrared thermometer so that you can be sure. You can always sell it on eBay when you are finished if you don't want to keep it.

#205 Post by **rkawakami** » Sun Nov 25, 2007 11:48 pm

Earlier in this thread.....

http://forum.thinkpads.com/viewtopic.ph ... 232#254232
http://forum.thinkpads.com/viewtopic.ph ... 029#238029
http://forum.thinkpads.com/viewtopic.ph ... 191#238191

dozer · #206 Post by **dozer** » Thu Nov 29, 2007 2:01 am

rkawakami wrote:Sorry to hear of your denied request at repair. The BGA package itself is not the sole cause of the problem.....What seems to be the issue is the use of lead-free solder. .

Yes, the switch to lead-free solder was your typical gov-induced balls-up...lol...a "problem" that was already in the process of being solved anyway, and in a more sensible ways (i.e. the rapidly growing commercial recycling of all electronics).

In any case, I wanted to offer a few additional observations from the perspective of a mechatronics engineer....

Heat and age both cause grain-growth in solder (i.e. growing metal crystallite size); as does oxidation and other chemical attack. Larger grains tend to crack apart more easily.

If you look at solder joints on wire-antennas for example, which are both old and have been out in the weather for several years, you can even see the larger grains with the naked eye.

Eventually these joints all go intermittent. Most professional antenna installations specify mechanical-bonding in addition to the final soldering (i.e. crimping, wire-interlocking, etc.)

So, keeping the part/board cooler most definitely prolongs time between open-circuits by reducing the growth-rate in grain size.

Second, I have seen little mention of flux in the discussions of heat-gun etc. procedures. Perhaps there is "no clean" flux left over on the board from mfg....but when it comes my turn to do this, I think I will be applying a real rosin-based flux prior to reflow (thinned, so that the spray gets under the BGA part fully); and cleaning it off afterwards with isopropyl alcohol.

Flux is a BIG help in removing oxidiation; not just from the solder, but more importantly, from the ball/pad -interface-. And that's really the key here....not just getting the ball to touch the pad again, but getting that solder to BOND to the pad solidly again (or in the first place? LOL).

So I think that flux has a very important role to play in any repair that's wanted to last LONG term.

One final note....the original processes were either hot-air reflow or infrared-reflow; or a combined process. Thus, comments about keeping heat away from the top of the chip are kind of silly.

Imagine the board in a toaster-oven....with those red-hot coils radiating onto the TOPS of ALL the parts....yup, that's exactly what the original manufacturing IR-reflow process looked like.

Albeit, with a bit more precise controls...lol..

Hot-air reflow, same thing....hot nitrogen blowing ALL over the parts...not just the leads/balls on the bottom.

In fact, if by some miracle you figured out how to heat only the balls to reflow-temp, without heating the rest of the chip, you would likely break the chip, literally....via differential thermal-expansion causing it to warp.

The entire package must be heated -evenly-....that's what the time-temp profiles (preheat and postheat) are for...to insure even heating and no warping of components or the board itself.

The key in this type of rework is to limit the PEAK temp. That's where doing it with a propane-torch can be tricky. You can do it, but you need to be very careful of distance, i.e. max peak temp on the part.

A reflow-oven uses lower peak temp in the gas than a torch or heat-gun; but uses a much greater MASS of heated gas, in order to get the parts up to temp.

ps; there is a man in Arizona USA doing Apple laptop boards for $50. I have not personally had work done by him, but his epay feedback is 230 feedbacks with 98.7% positive.

I believe he states a success-rate of 8 out of 10 boards. A 20% 'failure' rate may not sound good; but you have to realize that not every bad board is a bad BGA-ball to board problem in the first place. And not every BGA 'failure' is in the part-to-board interface.

Be aware that many BGA parts have TWO sets of interfaces. There's the obvious one that we're talking about, where the BGA balls go onto the PCB. But many larger packages also have a 2nd set of bonds, between the chip itself and the TOP side of the "package" (think of the 'package' as another miniature PC board, and you'll have the idea).

They call that "flip chip" or "C4" mounting; and those bonds too can fail; especially from extreme thermal-cycling (i.e. repeated cycles covering a wide temp range). So again, keeping down the PEAK temp, i.e. the total -swing- of each thermal cycle (via better heatsink or cooling, and lower usage/speed of GPU) can indeed prolong the time between open-circuit events.

Great thread....thanks much to all who've contributed and provided links.

EDIT: speaking of which...lol...here is the link for the guy in Arizona:

http://www.firstphasetech.com/

unitedunited · #207 Post by **unitedunited** » Thu Nov 29, 2007 11:33 am

Interesting post... a good side effect of this is that I've learned way more than I ever thought I would (or should!) about soldering BGAs.

Do you have a seller name or website for the apple board guy in AZ?

dozer · #208 Post by **dozer** » Thu Nov 29, 2007 12:52 pm

whups...sorry about that....here is his website:

http://www.firstphasetech.com/

ps; united, about that black goo...you might try a heat-gun on it. Most glues will soften with heat.

azazello · #209 Post by **azazello** » Fri Nov 30, 2007 2:19 pm

This thread is great at laying out the options for us unlucky ones with the BGA problem.

I think one key factor in this is the amount of flex applied over the lifetime of the laptop; I know mine is very high since I throw it into my backpack and carry it around with no padding multiple times every day.

Anyway, my T40 started displaying the usual symptoms a couple months ago (not as bad as some though - just jittery noise on the screen and random lock-ups), the noise would go away after pressing on the trackpoint area a few times - that or it would lock up.

I called IBM/Lenovo and they said there was nothing they'd do since it's out of warranty. I checked on eBay and the best I could find was a $300 system board+base cover assembly, claimed to be new and of a new revision (the seller knows about the BGA problem). Meanwhile the laptop kept crashing every few days. Frustrated, I called IBM again and they escalated the issue. Eventually they told me to use the "email IBM CEO" form, after which someone from IBM Executive Relations called me and said they'd need more time to research this. He got back to me within a week and sent me an advance replacement system board "as a one-time exception". I haven't checked yet if the board is a new revision or not, but I'm a happy camper since the problem is gone.

I've also had another non-critical problem which I haven't been able to find info on online: the left screen hinge on this T40 is wobbly. The root cause is that the arm into which the two big anchor screws in the back screw in is too thin and accepts a lot of torque when the hinge is rotating, so after enough open/close cycles it breaks around the top screw's thread and loses the anchor point. The hinge assembly has been replaced 2 times now and now it's broken again. Is anyone else here familiar with this problem? I assume they redesigned the hinge on later models... but I'm not sure if a T41/42 14" hinge will fit, or has a fixed design. EDIT: The hinge problem was my own fault... or at least the fault of someone who took off the screen assembly in the past. One of the screws anchoring the left hinge was missing, which was putting a lot of torque on the other two. Fixed that problem for now.

ZELAZNY · #210 Post by **ZELAZNY** » Mon Dec 03, 2007 5:34 pm

All models made in China ,are known as very sensitive for screan movements.One & only prevention to fix it is baking!
Instruction movie is shown in the Youtube,surch T42 fixing!

T42 shuts down when i move it

Re: WARNING: quoted IMAGES

Bad soldering (T-series)

Who is online