Power brick with voltage problem, what's wrong here?

Yesterday I was looking at one of those store-bought multi voltage AC/DC converter bricks. One of the big heavy ones that let you select anywhere from 1.5 to 12V output.

When I checked this one with a multimeter, the voltage was messed up. It was set to 1.5V, output was 5V. Setting it to 4.5V gave an output of almost 9V. And so on.

Inside, a typical chinese-made rectifier. A multi-tap transformer, 4 diodes, a capacitor, a selector switch and a power LED. Resistor from one of the diodes to ground, and resistor in series with the LED. That's it.

Checked the 4 diodes for shorts or opens, they're fine. Checked the 2 resistors, they both measure up OK. LED lights up ok. And, no sign of problems in the cap.

So that leaves only one culprit: the transformer. Might as well get a new brick I suppose.

Getting into a more detailed discussion, out of curiosity, what's wrong with the transformer inside? Most bricks I've seen that fail just go dead because the windings open. But this one seems to be putting out more voltage than necessary. What causes that?

If there's no voltage regulator you might need a load to pull the voltage down. An unregulated power supply will only be approximate, but if it's off by a factor of two or more that's pretty bad! ...I wouldn't be surprised to see 50% above the rated voltage with no load.

For example, if you buy a 12V 1A transformer it will run above 12V at less than 1A. If you add rectifiers and a capacitor to make a simple unregulated supply it will behave similarly. An unregulated supply also depends on your line voltage... If you power line voltage is 10% high, the output will be 10% high.

If you need a regulated voltage, be sure to buy a regulated power supply.

What is the max current the transformer is rated for? Pick a value something like half that and calculate a resistor that will pull that current at a set voltage then use that resistor to load it.

Bad idea to use an unregulated supply for delicate electronics.

DVDdoug: will only be approximate, but if it's off by a factor of two or more that's pretty bad! ...I wouldn't be

what is the nominal primary voltage and what is your regional supply voltage ?


Resistor from one of the diodes to ground

Interesting, can you post a couple of pictures of the offending item, both sides of the PCB if possible.

Tom.... :)

what is the nominal primary voltage and what is your regional supply voltage ?

Ahh! Good thinking!!! If the power supply was designed for 120VAC and you're plugging it into 220VAC you'd get twice the voltage out. The transformer would probably survive but the capacitor might get over-voltaged and blow. And, you'd have to derate the current capacity to half (twice the voltage at half the current is the same amount of power).

...Not Arduino related, but about a week ago I fried a board at work by plugging it into 240V before properly switching it for 240V operation. :( I saw a flash and a puff of smoke from the regulator. A voltage-spike got through the regulator and 3 other components on the board died too. The power fuse blew, but as usual it was too late to protect the circuitry. Unfortunately, the regulator didn't short to ground or open-up... Apparently, it shorted-through and something like 60VDC was probably applied to all of the 5V chips. It's amazing that most of the chips survived.

So you are going to fit crowbars on your SMPSs now, aren't you?

The weakest part protected the rest. I've even seen this on lightning struck modems ( the burnt missing traces are a dead giveaway. ). Years ago, when we were using these devices, I'd bought a box of 50 of these burnt ones for $5. I was able to make about 20 good ones by swapping parts. Dwight