Both supplies (same circuit) stopped working after cleaning PCB
Hello. I have a design requiring a dual output, so I put two TOPswitch (TOP255YN) based supplies on one board at +25 V each. One of the supplies worked perfectly while the other was not quite as stable - it had slightly more what appeared to be intermittent ripple, triangular in shape. Under load the ripple was much higher. Same exact circuits, nearly same layout. There was some extra flux around the parts on the supply that wasn't working so I cleaned the board with flux remover. Now both supplies don't work. Outputs fluctuate around 50 V. The bias voltage fluctuates around 5- 6 VDC. I allowed the circuit time to dry. Used heat gun to further dry. I then tried cleaning with a solvent to remove any residue. Same thing still happening. Any thoughts? Is there a high impedance point somewhere that is sensitive to the slightest residue?
Schematic attached. Pretty much right from PI expert but the bias filtering cap is a ceramic.
I have used a similar circuit with success on other designs so I know it can work.
Thanks!
SCS
Comments
uploaded a PDF of the schem and layout. Some comments on design: One AC input section and rectifier drives both TOPswitch circuits. They share a common source and DC supply. A jumper option allows the output voltage to be changed but not sure this is going to work since this sets a 25 V supply to run at 16 VDC. All Al electrolytic caps are doubled to reduce ripple current in each.
The PCB dimensions can not be changed.
Update. I seem to have one of the supplies working, but the other refuses to ever work consistently. It doesn't seem to have any feedback, so the output voltage varies from 25 V to 50V. The TL431 gets burning hot, and the 100 ohm resistor smokes. I've replaced the TL431 and the optocoupler about 5 times now. These things seems barely marginally stable
Update: Got both supplies working now. Seems it is necessary to replace every active part if something goes wrong. Plus now using zener clamps on output to prevent voltage from going too high if feedback is lost. The Iset resistor will be set to allow up to 0.75A (0.5A is spec, revised down from 2A). Seems to be around 12 k. Transformers were designed to provide 0.75A. Striving to get noise, ripple down to below 10 mV. Output filtering common mode and differential helps a lot there. Still interested in your thoughts on the layout, and why cleaning the PCB caused such issues. Thanks!
Hi SCS,
Some of the possible reasons for your board to have stopped working after you cleaned the board with flux remover could be:
1. Loose solder balls - Loose solder balls if present on the board could have dislodged from their original location and could now be presenting a short between components/pins of the IC
2. If the board was not completely cleaned, traces of the flux could provide a conductive path between high impedance nodes such as V pin and source pin of the IC and hence cause the board to misbehave.
Hope this helps
Best Regards,
PI-Kryten
Hi SCS,
Thank you for considering Power Integrations for your solution.
Can you kindly share your PCB layout as well?
Best Regards,
PI-Kryten