400 Watt PC Power Supply Repaired
About a week ago my E3400 dual core pc suddenly stopped working. And I didn’t knew what happened because it happened while I was busy with backing up important data about electronics. It was if had a Power Loss at home. The Fan of my pc Power Supply stopped turning and with a unexpected silence my pc was dead.
I first checked if anything else at home still worked and everything did, except of course my pc. So because it was no power loss, I opened up my pc and first checked the Power Supply. There was nothing burned or anything visible that could explain the reason why it had stopped.
But since I had no equal 400 Watt supply and I needed my pc badly to finish my work, I decided to use an older 300 Watt ATX supply instead. And later on when I had more time I would check if I could repair the now defect 400 W ATX Supply.
And today the 26 of October I was finally ready to check the inside of this Switching Power Supply. Next photo gives the specs of this model ATX-400W / LP-6100. My first thoughts were that probably another electrolyte capacitor had given up. I used my MESR-100 ESR meter to check for bad ESR values. And I checked all of them but they all were fine! (Also I could have used my Blue ESR/Low Ohms meter from Anatek Corp. for this). I also checked all Capacitor values with the M6013 meter. And all measured within their given value as written on the caps.
Next photo shows the AC Power Input side.
So because the Capacitors all checked fine, I had to look further for the reason why my ATX Power supply gave up on me after about 4 fantastic working years. But according to the sticker on the outside it was manufactured somewhere in 2005. I bought it second hand so the ATX supply already had worked much longer before I had it.
Next I checked on the bigger resistors with often low values. They all checked fine too.
It was obviously no short circuit either, because the supply just had stopped in the middle of my work. And I had to look further so next I removed the secondary semiconductors from the pcb. Not by using a vacuum desolder pump. But just by desoldering the Aluminium cooler plate and all the attached semiconductors together by using enough solder and my solder station (My 25 year old trusted Ersa MS6000 !) at the right temperature. And when you do this carefully one side at the time, back and forth, over and over again until all components are nicely removed, parts and pcb won’t be damaged afterwards. This is the best desolder solution in these special cases when the desolder Pump just doesn’t work because the holes are too small to free the component pins. And after you removed all components with cooler and all, you are much better equipped to open up the now empty holes with that same desolder vacuum handpump. Doing it differently will damage your pcb or break component pins and you won’t be able to remove the semiconductors at all!
There were 3 components on the secondary cooler. All were dual diodes, and one was very big for obvious larger currents. They all checked okay too. And after this cooler was removed I could better view all other parts including the primary cooler. Next photo gives an idea how the supply looks on the inside.
Previous photo shows the inside of the ATX power supply with the secondary cooler plate removed. With on it the 3 dual diode semiconductors. In the back of the previous photo the primary cooler plate is visible.
Next photo shows the secondary cooler with the checked okay components.
As you can see no parts were harmed while taking previous photo (LOL). And I checked this secondary cooler plate first because I was thinking that the big semiconductor maybe was a defect switching Mosfet or Transistor. But it checked out it was only a very big Dual Schottky Diodes.
I checked all parts with my digital semiconductor testers. I have 3 different testers. I had built my first one from a project in Elektor called the SC2005 tester that got a firmware upgrade a few years ago from the author of this tester. But as it turned out to my surprise it gave wrong values while testing one of the semiconductors on the primary cooler plate.
But first next photos show you what the solder side of this board looks like with the secondary plate removed. The open holes show you exactly were the plate with the 3 dual diodes was soldered.
And the 2 photos after that show you the controller IC (a PWM controller 494 probably Texas Instruments) and a LM339 (a quad voltage comparator). Both are placed right next to the secondary Cooler plate.
Finally I found the cause of this suddenly bad ATX supply on the primary cooler plate.
Next photo shows the primary cooler plate before removing from the board.
My Semiconductor tester 2005 (Elektor 2005 + FW update, new IC) on previous photo sees the MJE13007 as a PNP transistor which it clearly is not! But the SC2005 tester checks correctly that the Bad MJE13007 is shorted internally. A Bad MJE13007 which is cause of this failing Power Supply.
My China tester (2014-07 by FISH8840) correctly gives a NPN type. But it looks like the symbol of the transistor has a wrong PNP arrow direction. Because the Emitter arrow direction in a NPN Bipolar transistor normally goes out of the transistor. (And it here clearly goes into the Emitter)
Also there probably is no internal diode going internally from point 3 to point 2. At least the datasheet of the MJE13007 doesn’t show any diode either!
From the two MJE13007 transistors was the middle one short circuited on the primary cooler plate. After replacing it by a new MJE13007 the 400 Watt ATX power supply works like new again!!
These Bipolar transistors continuously can switch 8A at max 400 Volt DC. And are specially intended for Power supplies and capable to withstand reverse high voltages up to 700 V! And a power max of 80 Watt.
And before you all react in saying that the price of a new Power Supply is just about 35 Euros/Dollars or whatever. Let me remind you that the cost of a single MJE13007 is close to nothing!! And that was all that took to repair this fine 400 Watt Power Supply for the next years of continues operation.
Following the datasheet of the MJE13007 Bipolar switching transistor.
On above datasheet you can clearly see the right symbol for the MJE13007.
Last but not least 2 photos that unmistakenly show that the Power supply repair was a great success !
The FAN is also turning. And the Power supply tester proofs that all voltages are fine! Another Switching Power Supply repair successfully completed.
Thank you for reading, until another repair!
Albert van Bemmelen, Weert, The Netherlands.
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!
P.P.S- Note: You can also read his previous repair article below:
P.P.P.S- If you wish to learn how to repair switch mode power supplies like a Professional, please click HERE.