Don't Miss

Burst Capacitor And Dry Solder Found In ZEBRONICS ZEB 450W SMPS

By on August 24, 2019
Burst Capacitor And Dry Solder Found In ZEBRONICS ZEB 450W SMPS










how to repair atx power supply

This SMPS was brought dead and after opening and cleaning it, observed that there was a burst capacitor on the secondary side, and a lot of dry solders.


toroidal coil in atx power supply


dry joints in power supply

As the ESR of the two tank capacitors, 330uF/200V were out of range, removed and on checking, look at the readings that I found:

330uf 200 volt cap bad

Replaced the two caps and also completed retouching all the solder points on the board. I normally start from the AC input section and follow the patch until I reach the last point on the board.

Usually I complete this work within about 30 minutes or even lesser sometimes. Look at the shine on the PCB after thorough cleaning, dry solder patch up and another cleaning for removing the flux.

bad components in atx power supply

how to fix atx power supply

The SMPS was tested and found to be working perfectly fine, and fit back!

atx power supply tester

atx power supply fixed

I usually use disposable paper cups for storing the screws and cut leads of components.

Another job added to my satisfaction BAG with a ‘bang!’

This article was prepared for you by Parasuraman Subramanian from India. He is 69 years old and has more than 30 years’ experience in handling antique equipment like Valve Radio, Amps, Reel Tape Recorders and currently studying latest tech-classes conducted by Kerala State Electronics Technicians’ Association. He has done graduation in BBA degree, private diploma in Radio Engineering and retired as MD of a USA company. Presently working as Consultant to Hospital and other institutions.

Please give a support by clicking  on the social buttons below. Your feedback on the post is welcome. Please leave it in the comments.

P.S-If you enjoyed reading this, click here to subscribe to my blog (free subscription). That way, you’ll never miss a post. You can also forward this website link to your friends and colleagues-thanks!

You may check on his previous repair article below:





  1. Waleed Rishmawi

    August 24, 2019 at 1:16 pm

    congrats on the repair my friend. I have exactly the same power supply tester and works wonderfully every time. have a blessed day my friend.

  2. Robert Calk Jr.

    August 24, 2019 at 1:58 pm

    Good job, Parasuraman!

  3. Albert van Bemmelen

    August 24, 2019 at 2:57 pm

    Another great job done! Although these power supply testers are great they not always help! Because I have the same tester and 2 others that all show that a 450 Watt power supply that someone brought me works fine. Which isn't working fine when placed in a desktop computer. Because the Fan isn't working and the circuit or sensor that checks the temperature must be defect the power supply is not working correctly! I have tried to find the culprit but was sadly still unable to. And all mosfets and transistors on both cooler plates tested fine when I had removed both from the board! So be aware that these testers are useless in such cases.

    • Robert Calk Jr.

      August 26, 2019 at 12:51 pm

      Hi Albert. Did you test the FET's, transistors, and diodes with a VOM with X10K?
      Sometimes diode junctions test ok with other testers, but fail with a VOM.

      • Albert van Bemmelen

        August 26, 2019 at 2:58 pm

        Yesterday I repeated my search in this 450 Watt Power Supply to find if I could fix the not working Fan. Apparently I was able to startup the Fan by heating up the pcb board with my hot air station. So the circuit to start the Fan works but apparently very poorly! Because it never seems to start on its own! So I still have to repeat my search in finding the temperature sensor that is causing this by simply carefully heating the appropriate area and check that circuit from the moment the Fan starts. By-the-way: All capacitors and semiconductors tested fine Robert!

        • Albert van Bemmelen

          August 26, 2019 at 8:53 pm

          Robert, I finally found the culprit! The Fan is working now! A previously 'working' transistor unexpected turned into a diode. I will make a new repair article about it.

          • Robert Calk Jr.

            August 28, 2019 at 5:08 am

            Hi Albert. Yep, sometimes transistors test good with most testers but fail when they are put under a load. The good thing about the VOM(with X10K) is that it can find the culprits and it is also not very expensive. Nearly anyone in the world should be able to afford one.

        • Parasuraman S

          August 26, 2019 at 9:55 pm

          The coils in the secondary are very important. If we read Jestine yong's world famous book on SMPS Repair, he has covered how important it is to replace the coil, if necessary. I did it once in an SMPS, as it was found blackened. Even salvaged, good looking one will do. Did you check the switching pattern in an oscilloscope? You might already know that connecting the ground of oscilloscope to Hot ground and just keeping the probe on top of the SMPS will give us a good swing. Watch whether it varies or stays stable. If it varies, the problem is in the primary section. Similarly, you can connect the scope to secondary too and check. Not on Transformer, but on the circuit.

          • Albert van Bemmelen

            August 27, 2019 at 6:49 pm

            Thanks for the valuable tips Parasuraman! It wasn't a secundary coil in the transformer or anything like that but a transistor that previously tested as a good NPN transistor on my DCA75, but still was out of spec. I managed to find the culprit after rechecking and yesterday made a new repair about my finally solved not working Fan problem that was almost untraceble until I compared it with the component datasheet and desoldered the components again for a confirming second test!

            • Robert Calk Jr.

              August 28, 2019 at 11:47 am

              That's good to know, Albert. I know that the DCA75 Pro has it's limits but mine hasn't been wrong yet on transistors. It was wrong on some diodes. I will keep an eye on it.
              Just goes to show that we can never trust anything 100% of the time. It's always good to have many meters and not to depend on just one meter.

              • Parasuraman Subramanian

                August 28, 2019 at 2:41 pm

                The DAC is not a sure shot and has proved wrong in my experience. Analogue millimeter in X1 range and recheck on X10 is sure to expose the defect. If the reading between B to C and B to E, is not identical, just replace the transistor. Our expert, Jestine Yong also strongly recommends use of Analogue Multimeter.

                • Robert Calk Jr.

                  August 29, 2019 at 8:21 am

                  That is true, Parasuraman! The more experienced that I get the more that I appreciate having Mr. Yong's e-books, and the more that his teaching is validated.

              • Albert van Bemmelen

                August 28, 2019 at 4:19 pm

                Robert, I noticed something very odd about testing the CBO and EBO max breakdown voltages with our DY294 Tester. I was unable to determine the breakdown voltage of my in a Corsair VX450W repair used S8050 NPN transistor.

                Maybe it is something that can be explained? Because strangely both the CBO and the EBO voltages exceeded any of the in the datasheet given maximum voltages. Although I in my new soon to be published repair wrote that the higher CBO was about 60V and the lower EBO only about 25V, my DY294 showed in both tests about the max 1630V DC but my DY294 display began flickering while I was pressing the test button longer down.

                The CEO voltage my DY294 gave was splendidly 26V as given in the datasheets.

                I also tested the Iceo which was 1574uA on the 2000uA scale test. (IMPORTANT!: many DY294 users are unaware that these I ampere current tests can destroy the weaker diodes instantly because they get hot without even having to press the test button during these selected current tests.)
                And to make sure my transistor wasn't destroyed after the tests, I tested the NPN in my other digital testers which showed a perfect working NPN with a Hfe of 293 and a Uf of 629mV.

                I have no idea what is the correct explanation behind these very high breakdown CBO and EBO voltages of the S8050 NPN under test? (other transistors are giving a decent breakdown voltage!).

                • Robert Calk Jr.

                  August 29, 2019 at 8:16 am

                  Albert. I noticed that also with the DY294 on one of those other settings and burned my fingers. I checked the transistor just like the manual says. Now I only use the DY294 for checking insulation and zeners only. I use my other insulation tester for capacitors because it will discharge them automatically.
                  I am going to check all diodes, transistors, FET's, etc. with my VOM first from now on. And when it fails one I will check to see if my DCA75 Pro passes it.

                  • Albert van Bemmelen

                    August 29, 2019 at 10:05 pm

                    I repeated above breakdown voltage tests on my new S8050 Bipolar NPN transistor with both my perfect working DY294 Testers to make sure they both were working fine and are giving the same results which they did!
                    So it must have been a bad transistor pincontact that previously lead to not reaching the correct breakdown voltage on the 200V and the 1000 (over 1600V!) testscales.
                    Now I got even better results than those given in the datasheets (depending on the manufacturer the voltages are often not the same) and I got normal acceptable values!.
                    So as in the manual is written just use only the E and C connections on your DY294 Meter/Tester to test the BREAKDOWN voltage of any Diode/Triode, and connect in case of a NPN triode like my 8050:
                    1. For the Vcbo (collector-base) breakdown voltage test, simply connect the Collector of the transistor to the C connection on the DY294, and the Base of the transistor to the E connection of the DY294.
                    2. For the Vebo (emitter-base) breakdown voltage test, connect the Base to the E connection of the DY294, and the Emitter to the C connection of the DY294.
                    3. For the Vceo (emitter-collector) breakdown voltage test, put the Emitter on the E connection and the Collector to the C connection of your DY294.
                    And press the test button on the 200V or 1000V scale selections to read the breakdown voltage on the display. The breakdown voltage normally kills any semiconductor but in our DY294 its is a protected voltage test which doesn't destroy your transistor/diode (or capacitor)! But never touch any wires of contacts while pressing the test button!
                    In case of a PNP transistor you need to reverse the Base and Emittor transistor connections because the current (the arrow in the emitter symbol) flows in the other direction.
                    There is also a 4th Breakdown Vces test but that is only a Vceo test with the base connected the the emitter.
                    My DY294 testers gave a Vcbo of 81V (30,40 or 45V depending on what manufacturer), a Vebo of 11V (5, or 6V depending on the manufacturer), and a Veco voltage of 26V (25V in most datasheets).
                    So everything works as is to be expected! Problem solved.

                    And we only have to be careful when we are using the current tests. If we only use these tests for a very brief moment, our semiconductors won't be harmed! Something that probably was not mentioned in the user manual.

  4. Abby karonga

    August 25, 2019 at 3:24 am

    For almost 2 years I had stopped reading due to work loads and my husband discouragment was missing these articles for sure thanks to Jestine Yong who keep posting and I did not delete even one of the so I have a lot to catch up and still need to read new ones.Thanks to all participants who never give up not leaving Mr Albert, Robert Calk,Waleed Rishmawi,Sir Parasuraman,big up to Suranga and others keep the ball rolling.

    • Albert van Bemmelen

      August 26, 2019 at 3:18 am

      Thank you Abby for that much appreciated comment! We all agree that electronics plays an important role in our lives. And we all enjoy reading about it in each other's repairs. And maybe even make new friends too while reading all the comments. I made some interesting new friends thanks to Jestine's free Blog. Even if they live on the other side of the World it is possible thanks to modern electronic technology! Jestine's Blog is a perfect place to learn new repair engineering tricks and helps upgrading our skills! Cheers!

    • Robert Calk Jr.

      August 26, 2019 at 12:58 pm

      Thanks Abby! The majority of people never comment or anything.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.