I had stopped repairing ATX power supply many years back due to the new one cost very cheap. It’s not worth to repair it because the spare parts sometimes were much more expensive than getting a new power supply. Searching for ATX power supply spare parts was not easy as many of them you can’t even find them on the internet. Not only that, many complicated and different designed by power supply manufacturers had eaten up our precious troubleshooting time too because of we need time to understand how all these different designed power supply work.
Some of the power supply designs were using the PWM IC (UC3842) and power FET, some use the double transistors while some use only a single power IC in the primary side. Because of the manufacturers wants the design to be made into compact size, many secondary or even primary power supply circuit were build into a modular board (smaller board). This made troubleshooting even more difficult because many times the meter’s probe can’t reach to the testing point.
The real reason why I had stopped repairing ATX power supply was the profit margin. If you charge to high the customers rather buy a new unit with one year warranty given.
Should you have almost any issues about exactly where and also the way to work with Innosilicon A11 Pro ETHMiner (2000Mh), you can e-mail us in the web-site.
If you charge too low, you may end up in the losing side because of the components replaced, electricity and etc. If you charge reasonable, the profit margin gained can’t even cover your time spent on troubleshooting it. I’m here not to discourage you to stop repairing ATX power supply, however if you have the time, have contacts getting cheap power supply components, easy to access many power supply schematic diagrams and etc then you may go ahead to repair it.
Okay back to the article, one of my customers had asked me to repair his ATX power supply. I told him to get a new one (since it was very cheap) but he said he couldn’t find one that suits his customer’s CPU. He wanted a power supply that is either same size or smaller then the original one with same or higher specification but all he could find was a standard size power supply!
As a favors to my customer, I would do my best to help him to repair the ATX power supply. When the power supply was switch on, measurements were taken. The results were over voltage. The 12 volts line shot up to 13 + volt and the 5 volts line became 5.6 volts. After the casing was removed, I found the inside was very dirty and I used a vacuum cleaner and a brush to clean off the dirt. Then I saw four filter electrolytic capacitors had bulged at the top casing.
As you know, we as electronic repairers can’t just see things at only one side; we have to see the other sides too. What I mean was, try to see if there are any suspicious components that contributed to the failure of the power supply such as broken components, dry joints, loose connection, decay glue and etc before start checking the suspected area.
What I saw was at the primary side there were some components covered with decayed glue as seen in the picture. I have to carefully remove it by scrapping off the layers of the decayed glue while preserving the outer layers of the components. Once it was done, I clean it with the Thinner solution. Decayed glue could cause serious or intermittent problem in electronic equipment because it can be conductive.
If you repair any ATX power supply, make sure you check the fan too because some power supply failure was due to heat caused by a faulty fan. The purpose of the fan is to suck out all the heat generated by the components inside the power supply. In order for the fan to run smooth, you can service it by using a Philips oil base spray as shown in the photo.
Once the four electrolytic capacitors were replaced and the decayed glue removed, I then have to plug it into a junk motherboard together with a hard disk to test the performance of the ATX power supply and measure all of its output voltages. It seems like the output voltages were back to normal. Once everything is okay I then test it in a working CPU to check for the display.