My IBM Thinkpad uses a power adapter that has an output voltage of 16V at 4.5A current. One day the adapter stopped working.
I decided to try to repair the adapter. In the past I repaired several switching power supplies of PCs and also one AC-power adapter of a Asus notebook. I found out that most of the supplies had similar defects. Often they are ease to locate and to repair.
This Instructable shows how to repair a IBM AC-Adapter but using the same principles it can work with any switching power supply.

Step 1: Things Needed AND SAFETY

First of all you need the defective power supply ... :-)
The you need a screw driver. It could be phillips type or a flat blade type, depending on the power supply.
In case of the IBM adapter you need a Dremel tool and a cutting disc too.
To find out the dead parts you need a multimeter that contains a continutiy and a diode test.
Having a soldering iron and and some pliers is also helpful when replacing parts.

AND NOW! BE VERY CAREFUL!
YOU ARE WORKING WITH LINE POWER HERE!
MAKING A MISTAKE CAN KILL YOU!

- Allways double check connections!
- Before putting the power chord into the socket, have a look at the scenery and try to see the things that are wrong.
- Keep a clean working desk (hard to do ... ;-)
- After pulling the power chord out of the socket wait some minutes for the capacitors to discharge. They keep the voltage for a long time and they keep a deadly high voltage!

Read this article if you like to know more about it http://en.wikipedia.org/wiki/Mains_electricity .

Step 2: Opening the Case

The IBM AC-Adapter is not meant to be opened. The case is made of two plastic frames pressed together and melted at the contact to one piece. To take it apart you need to cut the two halfes using a Dremel tool and a cutting disc.

WAIT SOME MINUTES AFTER PULLING THE POWER CHORD FOR THE CAPACITORS INSIDE THE ADAPTER TO DISCHARGE!

Cut with the disc along the sides of the case.
Be careful not to cut too deep. There is a shielding case underneath the plastic case that covers the electronics. If you see the metal in the cut, you are a little too deep ...
Cutting through the metal frame can damage electronic parts.

Only cut the two long sides. The sides containing the power plugs need not be cut .. we will break them open.

Take the blade screwdriver and put it in the cut you made. Place it on the edges of the case, because these are the robustest points of the case. Twist the screw driver to spread the case apart. The uncut parts of the case will break now. Do the same to the other corners of the case.

Take the plastic parts from the inner electronics. Now you can see the metal shielding. On the picture you can see that the shielding got some marks ... but it is not cut and it still works well.

Now you can remove the shielding and the underlying insulation to get access to the electronics

Step 3: Examining and Understanding ...

Start locating the parts of the power supply. We concentrate only on a few parts. I often found out that these are the most critical parts. Most swicth mode power supplies die when being switched on. At that moment a high current flows on the primary power side. You can see that if you plug in the power chord and have a look at the socket. Sometimes you can see sparks that are caused by the high current.

- Every power supply needs to have a fuse right at the input. This fuse will melt and break power connection if too much current is drawn. In our case the fuse is rated 4A. The power supply itself is rated only 1A. The rest is needed to cover the high current that flows when switching on.

- Switching supplies rectify the ac voltage to get dc voltage. This dc voltage is higher than the ac input voltage. A rectifier in a switched mode power supply has a hard job to do and sometimes they break. If you like to find out more about it, read this http://en.wikipedia.org/wiki/Rectifier .

- Another critical part is the capacitor that stores the input voltage. This capacitor has to withstand high voltages. Most of the high current that flows when switching on is caused by this capacitor.

Many other parts can break inside, but I will concentrate on the three mentioned above, because everything beyond that needs more skill, is more difficult to measure and you require a schematic of the power supply. Often you will not be able to get this.

If you like to know how a switching power supply works, read this http://en.wikipedia.org/wiki/Switched-mode_power_supply .

Step 4: The Fuse

Start with the fuse. Turn you multimeter to diode test (continuity test) and put the test cables on both ends of the fuse. The multimeter should "beep" and show a very low voltage (3mV on the picture). If that is the case the fuse is ok and need not be replaced. Otherwise you have to desolder the fuse and put in a new one.

NEVER USE A WIRE INSTEAD OF THE FUSE!

There is a reason why the fuse melted. If you replaced it and everything work, you are lucky, but most of the time other things gone wrong too and the fuse is only the indicator of a problem.

BEFORE replacing the fuse do the rest of the testing. It could be that the rectifier or the capacitor is broken and that this caused the fuse to melt. Good fuse, if that happened it did the job it was made for.

Step 5: The Rectifier

Next part in the chain is the rectifier. In nearly all cases I saw til today there is a full bridge rectifier used. Here it is a flat one located near the power connector. Again use the diode test for measurement.

From underneath the printed circuit board you can reach the rectifiers contacts easily. If you follow the stripes on the pcb you will see that the mains power goes to the two middle pins of the rectifier. Then the outer pins must be the ones where the dc voltage arrives. There are 4 diodes included in a full bridge rectifier. You should be able to measure all four of them. In one direction the multimeter should show you about 0.5V to 0.7V. Not every diode in the rectifier needs to show the same voltage. They are only nearly the same.

If you find one pin combination where the display shows nearly 0V, the rectifier has got a shortage and need to be replaced. If you find two pins where you get an infinite display the diode in the rectifier is broken and the rectifier needs to be replaced.

During measurement it could be that the display shows 0V for a short time and after some seconds it shows the expected 0.5-0.7V. This is normal. The effect comes from the capacitor.

If you found out that the rectifier is broken ... don't stop do the next step also, because this need not be the source of the problem.

Step 6: The Capacitor

Now use our Multimeter in diode mode to find out if the capacitor works.

Place the measuring pins on the pins of the capacitor and look at the display while doing that. As soon as you place the pins the display shows 0V. Then the voltage in the display starts growing and the display shows infinite. Exchange the measuring pins. The same thing happens again. If you are using a multimeter that has got a beeper, you can hear a short beep when connecting the pins.

If you do not hear a beep, or if the beeping does not stop after some seconds, the capacitor may be broken. To be sure if it is, you need to desolder it and repeat the measurement.

If the capacitor is ok but you measure a shortage on the pcb pads where the capacitor was soldered, the switching transistor may have a shortage. If that is the case you should desolder the transistor and repeat the measuering. If the multimeter shows a shortage, you could be lucky by replacing the transistor. Everything beyond this is more difficult and it would be to complicated to describe here.

Step 7: Repairing

After we found out what went wrong we can repair the power supply.

If the capacitor is broken, desolder and replace it. I tried to find out if this was the only defect part and decided to do further testing before trying to buy a replacement. I did not have the capacitor that was used in the power supply and had to use a near replacement.

If using other capacitors than the original ones you have to respect some rules in order not to burn some things down ...

- Look at the voltage the capacitor is made for. Only use capacitors that have values equal or above the one printed on the original.
If you look at the pictures carefully you will see that I used a replacement with only 400V. I simply took the risk because in the cheaper power supplies only 400V capacitors are used. They should work, but 420V gives you an extra gap of security. In high quality power supplies capacitors with more than 400V are used ... even this ones fail from time to time ... as you can see here.

- Take a capacitive value as near as possible to the original one. The original shows 68uF. I luckily found one that was 100uF. I would have tried a 47uF also, but that would lead to less current on the notebook side. For testing it would be ok.

BEfore desoldering the original capacitor write down a description about how it was soldered. It is importent to keep the polarity on those capacitors.
When soldering the replacement to the pcb be careful to solder the "-" and "+" to the correct pads. Keep the original to remember how it was connected.

To find out if the power supply can deliver the current needed, put a power resistor on the notebook plug.

DO NOT CONNECT THE AC ADAPTER TO THE NOTEBOOK NOW! THE NOTEBOOK COULD BE DAMAGED IF YOU DO!

ATTENTION! DON'T TOUCH ANY COMPONENT WHILE THE AC-ADAPTER IS SWITCHED ON! WAIT SOME MINUTES AFTER PULLING THE POWER CHORD BEFORE TOUCHING ANYTHING!

On the picture you can see that the power supply delivers the 16V as written on the sign. The resitor gets hot very quickly. I selected a 6.8Ohm resistor. That should draw a current of about 2.4A. That is about half the current the ac adaptor is able to give. This is ok for a short test. The resistor needs to be capable of handling 40W in this configuration. It should be a big one.

As you can see on the picture, the test capacitor does not fit into the ac-adapter. Now I need to buy a new capacitor, with the same rating as the old one ...