T20 Blink Of Death Circuit Questions

[Rex]

It's been some time since I looked at this thread, but an article in a UK magazine (Micro mart) has stimulated me to write.

The lead article in the current edition is about 'cooking your motherboard.' Apparently, lots of laptops show BOD symptoms and sine the boards are flow soldered, the premise is to remove virtually anything and everything and put it in the oven at about 180C for between 5 -9 minutes. This will cause the solder to soften and flow, but not actually melt anything. The author has 100% success rate at the time of writing.

I guess that since I have two T20s that are nothing more than expensive paperweights, it may be worth a try. I have nothing to loose.

Rex

[Tony.Yarwood]

Very interesting..please let us know the results

Best regards

Tony

Google ain't your friend.
More privacy, no IP logging.
http://ixquick.com/

[Rex]

Haven't tried it yet and will be some time before I do. Not enough time available at the moment!

The mag is the current issue of Micro Mart. It is based upon the premise that when the boards are made, all the components are pre-soldered and inserted into their correct positions. The mb is then heated to the pre-determined temp. (apparently about 200C), the solder flows and voila, a completed circuit board.

So doing this again, 'may' renew all the solder points and hopefully fix any dry joints. Seems reasonable.

Rex

[meander]

I've got a T20 that shows BoD symtoms, too. Through google search I've found a solution posted in a German Thinkpad forum. They also refer to the ADP3421 as the root cause of the problem. But it seems the chip is not defective, as a replacement does not permanently solve the problem.

What they propose is to pull the chip's power good output permanently to Vcc level with a 10k resistor. Other posts say they needed a lower restistance of down to 1k. Someone used 1,8k. So, simply bridge pins 16 and 24 with the resistor - that's all. The voltage you'll need to simulate the power good signal should be around 3V. The downside is that the resistor permanently drains the battery - but that shouldn't be a too big problem. Mine is more or less dead anyway. I'll try the next days and report.

[meander]

Wow! I did it. It was really fiddly to solder the resistor to the tiny pins and I shorted some of them several times with too much solder. But don't worry - you won't blow the chip if you power it on with shorted pins. Just the board won't boot at all anymore. After several attempts soldering, desoldering, measuring the pins for a shortcut and so on, I eventually succeeded. No signs of BoD anymore! Windows installation ran through completely and my T20 is now running since several hours without troubles.

I used a small 12V soldering iron with a needle like tip, a good magnifier, a thin flex wire, which I stripped down to just one wire, and some electrician's tape to fix the resistor after installation.

So, good luck if you want to try it.

[nikki605]

Any chance that you could post some good closeup photos of your final repair?

[Rex]

Some photos would be great; if not some more details as to which chip needs bridging. I'm not concerned about soldering, but have no knowledge of my way around a circuit board without some images.

Well done, and thanks.

Rex

[meander]

Sorry, I don't know where I could upload photos, but there are some in the German forum I referred to:
http://www.thinkpad-forum.de/thinkpad-h ... reparatur/

The datasheet of the ADP3421 is available here http://www.hit-karlsruhe.de/aol2mime/da ... DP3421.pdf

The ADP3421 is just under the keyboard near the keyboard connector, so no disassembling other than removing the keyboard needed. On the link below it's in D6, bottom left-hand corner (D5/D6/E5/E6). Hope that helps.

http://www.kawakami-ca.com/ibm_t2x/t20_ ... legend.jpg

[Rex]

Thanks for the links. Unfortunately the German site take me to a ThinkPad empty page. May be I need to register and log in?

Rex

[rkawakami]

Just a word of warning here...

Bridging a resistor between pins 16 and 24 of the ADP3421 converter/controller might seem to fix the issue but I wouldn't think of it as a permanent, nor safe, solution. Pin 16 is the PWRGD signal (Power Good Output). When the system is turned on, the power supplies start up and are checked for proper operation by this controller. PWRGD is supposed to be near 0v at this time. When it determines that everything is at the correct voltage levels, the PWRGD pin goes "high" (a voltage near 3.3V; essentially what is being supplied on pin 24). Attaching a resistor between these two pins simply means that the PWRGD signal will always be high when the system is turned on. By forcing the ADP3421 to report that "all is well" to the rest of the circuitry, it will ignore any condition where the voltages are out of spec. This may cause some other problems beside the BoD. Think of it as substituting a blown fuse with a piece of wire: you'll get power back but unless you find out why the fuse blew out in the first place, you could be asking for more trouble.

Funny bit, talking about power... PG&E, the local power company, just came by to install the "smart meters" on our house. Power went out for about 15 seconds during the changeover. My wife had to shut down her desktop but I continued to type this up on the T23 .

[Harryc]

What does the 'smart meter' do, transmit usage back to PG&E?

Edit: Never mind...found it...as if we didn't have enough RF in our lives.
http://www.pge.com/myhome/customerservi ... owitworks/