How I repaired the computer power supply

Yesterday my computer stopped turning on. He turned on for 20 seconds and immediately shut down. First of all I checked the supply voltage and cables, everything looked normal. I suspected that the power unit probably broke. I unscrew the bolts and here it is already on my desk.

I bought this PSU 3 - 4 years ago, the manufacturer Zalman. Not the latest in the market, it produces quite high-quality products. Model ZM500-HP. The rated power is 500 watts.

I take off the top cover. Oh ... and here's dust! Well, well, I confess, I have not cleaned it for a long time. But fully cleaned of dust, it was necessary to completely remove it, disassemble and clean it, it was a time-consuming business and I was too lazy.

The cooler is also in the dust.

I'm armed with a vacuum cleaner and a compressed air cylinder! A couple of minutes and everything is clean. Visually, we can say that the PSU is made of good quality, there are no missing parts on the board. There is an input LC filter, a short-circuit protection and an extinguishing thermistor, which protects against starting currents. Transformers have the appropriate power cores and winding wire. Even a heat-transfer tube with a radiator is, this is not often found in power supply.

One of the most frequent causes of breakdowns is the breakdown of capacitors. Most often this can be seen even visually, by deforming the body (swollen look) or by emitting an electrolyte (there may be something withered above or below). At a cursory examination, this condenser seemed suspicious to me.

It's good to see how the body was deformed. This indicates that it needs to be checked first. To get to the board, you have to untwist everything.

I unsolder a suspicious capacitor, its characteristics are 1000 μF 16V.

I check it with a tester. Well, that's what you had to prove! The loss of capacity is 10 times, instead of 1000 we see 85.25 μF. In addition, the internal resistance of 5.4 Ohm! Yes, condensers (electrolytes) have an internal resistance, and the less is the better. The best samples have the so-called Low ESR (Low Equivalent Series Resistance), i.e. low equivalent resistance.

I throw the spoiled condenser. Now I put a new one, but first of all it is necessary to check, despite the fact that it is new.

Next, put in place and sealed.

I installed a capacitance of 1000 μF per 25 V. It is larger in size and got into the grinding. But there was no other at hand.

I will also say a few words about the breakdown. It happens that the capacitors lose their capacity in the output filters, this is bad, the output ripples, and the constant will be "shaggy". But the power supply will continue to work. In our case, this capacity was in control circuits. The clever PWM chip that controls the PSU has its own protection. One of them in case of deviation from the set operating mode (which was influenced by our condenser). That's why the power supply did not turn on.

In the future I did not find other spoiled details and boldly collected everything back. The golden rule: before you close the body cover - check the operability)) Here without surprises, it all worked like a Swiss watch.

Perhaps this information will be useful to you, or maybe just read to kill time, in any case, thanks for attention :)

With the best wishes, @gleb-jeglov aka Evgeny