TPC8109 transistor problem ? - Help please

[Gates]

Hi, new to the forum, but looks like I finally found the right place to help me with this problem. I have searched fairly extensively for an answer to my problem and the best info I found was on this forum. Here’s the issue. I have a T43 with power issues. Problems started a few months ago. Laptop would shut off unexpectedly. When I say shut off I mean it was as if it lost all power (AC and Batt). No errors or events logged in event manager. After a shut down like this it was always a problem to get it re-started. Usually if you removed the battery and plugged in the AC as you held the power button down you would see the battery led on the bottom of the display flash. Then you could momentarily release and re-press the power button and it would re-boot. As time went on, the shutdowns became more frequent and the re-starting harder and harder. At this point I cannot get it to boot anymore.

Troubleshooting at this point:
Ran lenovos PC doc and didn’t come up with any errors. Tried it with different pieces of ram, different HD - you name it.
Checked temps with tpfancontrol and everything is good, CPU occasionally hit 70 C, but seems to fall right back into check. AC adapter and batt have been checked and swapped out with known good parts.
Here is really where I need the help. I checked all the fuses for the main power circuit F2, F11, and F12 – all are good. I noticed that when I plug the AC adapter in that one of the TPC8109 transistors immediately gets so hot that it will burn your finger. (Don’t ask me how I know this) It’s the one directly to the left of fuse F2.

I’m no electrical engineer, so that’s where I need someone’s help here that knows more about how this works. Can this be as simply as a shorted FET that needs replaced ?

Sorry for the long post, but I wanted to include as much detail as possible.
Any help would greatly be appreciated.

I know that replacing the mobo would be the easiest fix, but what’s the fun in that?
Thanks.

[richk]

I don't have a working T43 board right now so I cannot check, but I am supposed to have one Thursday. What is the fru of the mb or what is the type number of the mcchine ? (like 2686-44U)

[Gates]

mb is 39T0476 and machine is 2668-44u. Sorry, I guess I should have included that in the orginal post.

[RealBlackStuff]

If you remove the front palm rest, you'll see a big chip with Intel on it, half under the wifi card. Remove the wifi card also.
That is the Southbridge chip.
Try and put some pressure on that in different places, while starting up.
If that then works, the SB needs to be reballed.

[poshgeordie]

Hi Gates and welcome to the forums.

Does the T43 work OK just on battery? If not what symptoms does it have and does this mosfet you mention still get warm?

Also if you run it just on the AC Adapter are the symptoms still as you describe?

BTW for everyone else, it's an ATI based machine

[poshgeordie]

I'm fairly certain that none of the power mosfets get particularly warm and certainly if it's getting warm enough to burn your finger then it's pulling way too much current through it.

The most likely candidate is leaking electrolytic (voltage smoothing) capacitor somewhere along the supply chain, which is leaking current to ground and so causing high current levels to flow through the chip.

I'll identify the chip tomorrow at work and look into this further and see it it's possible to isolate the supply lines.

Do you have any soldering skills because I'll be suggesting removing surface mount components?

[Gates]

O.K. Here's what i found out. Pushing on the soundbridge chip doesn't seem to have any effect - still no response. The computer doesn't respond to anything on just battery, but the mosfet i spoke of does not get hot or even warm. Running on just AC is where all previous observations were made. I tried on AC and Batt and got the same result as just AC.
My soldering skill certainly aren't excellent, but i have a really nice stereo scope that makes it easy to see exactly what i'm doing wrong. I have managed to successfully re-solder some pins on a t23 board where the ram slots are, but really haven't had to remove anything except a dc jack - but i'll give it a try. Let me know what you're thinking.

By the way, thanks for all the input on this. I really appreciate it.

Gates.

[poshgeordie]

I guess you mean the mosfet device that is to the left of the DC connector with fuse F2 inbetween them.

I cannot be responsible for any damage etc for what I'm going to suggest - you do it at your own risk etc etc.

What I'm suggesting is not best practice, but could you short together the three top pins from the top right hand side to the bottom four pins - don't short out the top left hand pin with anything else.
Then try to start up the laptop on the AC Adapter and tell me what happens.

Also there's a good thread here and on the fifth post a circuit diagram of the DC in to the battery connectors.

As I've said before the above test is done at your risk.

[Gates]

Yes, I believe we are looking at the same component (8109 not 8019 like I said). I understand the risk, but at this point the board is pretty much for parts anyway. If i understand this correctly you have me connecting pins 1,2,&3 to 5,6,7,&8 or source to drain and leaving the gate alone. I will try this when i get home this evening and post results.
Thanks for the link. I looked at this when i was searching for an answer to this. It was helpful.
Gates.

[poshgeordie]

That's correct - connect the source to the drain.

[Gates]

O.K. Here's the results i got with source connected to the drain. No heat on the chip, but no response from the pc. Just to make sure that i'm doing this correctly, what components would i need connected to the system board to properly test this? I tried the above on AC power, no battery.

[poshgeordie]

You need the AC Adapter, keyboard to start the board up, CPU + fan, one memory module, the display.

Short out the source and drain, press the on off on the keyboard.

I don't think that chip's the problem, it's only indicating a problem elsewhere like a leaking capacitor or a short somewhere which is causing a high current to flow through that chip.

[Gates]

O.K. That's all i had connected, just want to make sure that was enough. Results were as described then. No heat on the chip, no response from pc. Upon removing the short on the chip, the heat would return. Re-checked votage on adapter and at jack - 16.39.

Gates.

[poshgeordie]

As I said before, it's going to be difficult to know what to suggest now since it could be any number of things.

The only thing if you can do it (warning again!), is to temporarily unsolder the TPC8014 chip at the bottom right hand corner of TPC8109 which gets hot (there's three upright green components which are resistors to the left of the 8014).

See if the 8009 gets warm if you switch it on. Then make a short between the source and drain of the 8014 and see if it gets warm (which I guess if probably will).

You could then try the same with the RH 8014 to the right of the other one and see what's happens.

By doing this we may be able to find what's causing the high current flow through the chip.

[Gates]

O.K. Just to make sure I'm following. I'm removing the 8014 below and to the right of the 8109 and trying to power up the pc, checking to see if the 8109 gets hot. Then i'm making a short between source and drain of the 8014 directly to the right of the one i just removed and checking to see if the 8109 gets hot - or the 8014 that i'm shorting? After that i'm replaceing the 8014 that i removed and trying the same with the other 8014 - Is that correct? Sorry about all the questions, but like I said, I'm certainly no expert on this kind of work. I'll work on this once i get back home this evening and report back.

Thanks.
Gates.

[Gates]

Well, i guess i greatly over estimated my de-soldering skills. One side of the chip went well, the other 4 pins now look like one big pad. I believe i'm done now. Oh well, it was certainly worth a try, the board wasn't doing me any good as it was.

Thanks to erveryone that helped with this, I appreciate it.

I was offered a free t42 board at work yesterday, so I think i'll start a new post to see if everything or anything will swap over.

Thanks again.
Gates.

[RealBlackStuff]

What you probably need is desoldering braid, to 'soak up' excessive solder.
http://www.radioshack.com/product/index ... Id=2062744

[Gates]

Actually, that's what i used to do the de-soldering. Can't seem to get it to pick up the last off the solder, or the pads are gone. Not sure which at this point. Other side went failry well, but looks like i lost one pad.
Gates

[gbarbanti]

I guess you mean the mosfet device that is to the left of the DC connector with fuse F2 inbetween them.

I cannot be responsible for any damage etc for what I'm going to suggest - you do it at your own risk etc etc.

What I'm suggesting is not best practice, but could you short together the three top pins from the top right hand side to the bottom four pins - don't short out the top left hand pin with anything else.
Then try to start up the laptop on the AC Adapter and tell me what happens.

Also there's a good thread here and on the fifth post a circuit diagram of the DC in to the battery connectors.

As I've said before the above test is done at your risk.

I do not understand how to do to join three pins from upper right with the four pins are three pins below .If you potrannno merge only with three pins .I can draw a picture please? thanks-

[gbarbanti]

Yes, I believe we are looking at the same component (8109 not 8019 like I said). I understand the risk, but at this point the board is pretty much for parts anyway. If i understand this correctly you have me connecting pins 1,2,&3 to 5,6,7,&8 or source to drain and leaving the gate alone. I will try this when i get home this evening and post results.
Thanks for the link. I looked at this when i was searching for an answer to this. It was helpful.
Gates.

I do not understand how to do to join three pins from upper right with the four pins are three pins below .If you can merge only with three pins .I can draw a picture please? thanks

[poshgeordie]

It's been literally years since I was last here, and I was very surprised to received a new post notification today! It's also been years since I last repaired a thinkpad, never mind the infrared reflowing etc, so a quick hello to anyone who still remembers me - memorable times!
I still have a working T4x and I know of a T40 still in daily use with a graphic chip I reballed over 10 years ago, and still with that great version of Windows XP that IBM produced for the thinkpads.

OK, back to the topic!

I do not understand how to do to join three pins from upper right with the four pins are three pins below .If you can merge only with three pins .I can draw a picture please? thanks

I've no motherboards to actually check, but it's very likely the three pins 1,2,&3 source, and four pins 5,6,7,&8 drain are already linked together with circuit tracks on the motherboard.
So, to link pins 1-3 to 5-8 you can simply solder a thin piece wire between one of three pins 1-3 and one of four pins 5-8.
Don't connect anything to pin 4.

Hope that helps. I can't help much more than that.

[ajkula66]

It's been literally years since I was last here, and I was very surprised to received a new post notification today! It's also been years since I last repaired a thinkpad, never mind the infrared reflowing etc, so a quick hello to anyone who still remembers me - memorable times!

Good to see you in this neighbourhood again, Nick...

Hope all is well by you. Rest assured that TPF hasn't forgotten you.

[poshgeordie]

Good to see you in this neighbourhood again, Nick...

Hope all is well by you. Rest assured that TPF hasn't forgotten you.

Hi George, and great to be back here again.

Also to you Dominic and thinking of the trans Atlantic work we did together.

Greetings to you both and everyone else.

[RealBlackStuff]

Hey Nick, how are you doing?
Great to see you're still alive!