Has anyone done any mobo level hardware troubleshooting?

[perls]

Howdy All--

I've got a troublesome A21m. Works like a champ on it battery, but shuts down when the AC supply is plugged in.

I wonder if anyone has done any component level troubleshooting on these Thinkpads?

I gather schematic diagrams are super secret double guarded and locked away 10 miles underground at some undisclosed location...

The AC supply works perfectly as I have a pair of these A21m's.
The battery works great as well.

The computer was working fine, then all of a sudden shut down under AC power.

A resistor on the mobo overheats when the AC supply is connected. Everything in the laptop shuts down when under the AC supply.

*ALL* of the fuses on the mobo are OK-- or at least all the fuses that I could find-- there must be at least 15 of them.

No shorted tantalum capacitors (a common menace) that I could find..

There 's a dual op amp (LM358) the function of which is a mysery. Voltages on it's pins are wildly different when switching between DC and AC power. This could be normal, but the voltage levels look suspicious.

I realize thise laptops are obsolete so don't command much money on resale. I'm an electrical engineer so it's sort of a challenge to see if I can actually find and clear the fault in this machine.

I simply don't have the time or the patience to reverse engineer the mobo. (It might not be possible without extraordinary measures)...

So, anyone ever see a schematic, block diagram, or know anything specific about the mother board in the A21m?

LMK--

Regards,

Perls.

[rkawakami]

Welcome to thinkpads.com!

If you poke around the various forums you might bump into these threads:

T42 Shuts Down When I Move It
T20 Blink Of Death Circuit Questions
Pulsing Fan on a Dead T23?

and many others... So yes, there has been some attempts at isolating faults down to the component level. This is being done without circuit schematics as nobody seems to have seen any. For general power problems, my money is on the Analog Devices or Maxim power controller/conditioning chips, or their related circuits.

If a resistor is heating up then I would guess that there's a short somewhere downstream from the source of power. It may be a situation with "tin whiskers" as that is a common problem with lead-free solder. From looking at various T23 motherboards, I'd have to say that type of solder was used as there are instances of cracking/cracked joints and evidence of "black pad" formations.

[perls]

Howdy rkawakami--

Thanks for the links which I've investigated and your ideas.

My suspicion is that very likely a tin whisker is causing my A21 trouble.

The tin migration problem is evidently a common problem with this RohS compliant stuff (I'm being nice by calling it "stuff"). Laptop mobos aren't the only surface mount technology that's plagued with metal migration.

It looks like all of the power supply controllers are common to both AC and battery power.


The search continues for a while at least..

[SMA]

Is it possible that power with reverse polarity has been attached to the machine?
Most likely, one of them universal AC adapters.

Have a look at this.

http://forum.thinkpads.com/viewtopic.ph ... ht=#311309

[perls]

Howdy SMA-

Thanks for your advice.

AC adapter polarity reversal could have been the source of the trouble.

The computer functions correctly on battery power.
It shuts down immediately as soon as AC adapter is connected.
The AC adapter is the genuine IBM unit for the A21m.

I've been wondering if the trouble isn't caused by a protection circuit triggering- like a voltage crowbar.

The other interesting thing I neglected to mention is that when the computer is operating with the battery and then the A/C adapter is connected, the computer "latches" in the fault condition until the battery is disconnected.

The battery power is obviously holding somthing in it's fault mode.

If you're familiar with the A21 motherboard, there's a curious mystery circuit built on a ceramic substrate. It's labeled "VR2". This circuit looks like something to do with system current demand.

I hate to diassemble my working A21m and use it as a "standard".
I could easily wind up wiht two non-working Thinkpads.

Perls

[SMA]

I wonder where that overheated resistor is located? Maybe I could locate it on the board I have laying around here.
Can you read the small white numbers on it? Not, that it really matters what the number is, but if it is no longer readable,
then it would indicate that that resistor has been very hot.

VR2 must be part of the power circuitry. It is indeed a funny thing.

The protection circuit seems to be functioning correctly. The behavior of latching the fault condition is as designed.
It is just a side effect that the protection circuit is reacting.

Sound to me like the problem is caused by a short. It must be in the part of the circuitry that is only in use when operating under ac power
or in the electronics that takes care of switching between the 2 power sources.
(Actually there are 3 power sources, if counting the 2 battery connecters individually)
Also under suspicion, should be the components involved in battery charging.

Have you had the board out of its housing. Otherwise I would say it is time to do so now, so that you can do a visual inspection on the bottom side.
And while at it - do another visual inspection on the top side.
Concentrate on the area 8x8 cm that is limited by the hdd, keyboard and ultrabay connectors - both sides of the pcb.
Also take a look at the tpc8106 mosfet on the bottom side, near the dc input connector.
The lm324, on the bottom side (below vr2) also needs some extra attention.
Look for something that looks burned.

The board I have been looking at is fru 10L1412, hopefully not too different from yours.

Just some thoughts - take it for what it is!

[perls]

Howdy SMA-

I have the entire Thinkpad disassembled (except the display).
In fact I have the disassembled machine up and functioning so that I can probe the circuit while the unit is operating.

The resistor that gets hot is 100 ohms-- marked 101
It's probably a 1/4W or 1/8W resistor.
It is located on the bottom (memory card side) of the mobo.

Of course I have forgotten to write down the voltages during battery operation and when the AC adapter is connected.

My mobo is Thinkpad A21m, fru 12P3488 or is it 12P3408-- hard to read even with a magnifying glass.

Here's a similar mobo on the 'Bay--

http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... :IT&ih=018

This one isn't exactly the same--
uses a different clock generator chip, no S-video output facilities.

With the main battery connector facing you (processor away from you)
Resistor is directly above pin 2 of the battery connector. Looking at the "bottom" (memory card side) of the mobo battery connector "pin one" is to your right. Then there is a thick polarizing "pin", then four more pins proceeding to the left.

Pin 1 on the main battery connector is marked "1".
It is either plus or minus battery power.
Pin five (longer than the other pins) is either the plus or minus.

I suspect that the other pins communicate battery run and charge status to the computer.

I've been reading the data sheets on the power supply circuit devices--

ADP3410, ADP3421; MAX1632; MAX1714

I want to find out more about the MAX1632 and how it integrates into the power supply system.

I suspect that this part is the master controller of the power supplies -- 1.8V, 3.3V, 5.0V & 12V

Too bad we couldn't co-ordinate an "open source" reverse engineering project to come up with our own schematics of these boards!