switching power supply with TOP234Y

Hi,
Can you please send some waveforms describing your problem. I am bit confused with your question.

For the audible noise, see if you can trace the noise to a specific component or area of the PCB. Common culprits are the transformer, clamp circuitry, inductors and ceramic capacitors.


If the noise is coming from your transformer, try varnishing your transformer. We use Hi-Therm BC-359MS from a company called Dolph's (www.dolphs.com) for our transformer varnish. Another option for a noisy transformer is to reconfigure the windings so that you have a lower peak flux density in the core. Decreasing the flux density is easily accomplished in PI Expert by increasing the number of secondary turns manually.


For clamp related noise, try replacing any ceramic capacitors with film capacitors. Ceramic capacitors have a very low ESR and can experience very high peak currents that will generate audible noise. Another solution besides film capacitors is to use a resister in series to increase the "ESR" of the ceramic capacitor.


If the noise appears to be coming from the inductors in your EMI section, try replacing them with inductors rated for higher peak currents. This should bring down the peak flux density experiences by the inductors and can help minimize noise.


For your second question, I don't really understand what you're asking. Can you please elaborate? If possible, please attach your PI Expert design files, transformer build diagram, schematic, PCB layout, etc.

-The Traveler

noise coming from the transformer that has no glue, attached files, thanks for your answer

ssyn -


Thank you for sending in your design files and PCB layout.


I'm noticing a few things that need your attention:
- You didn't run the optimization feature in PI Expert. This is highly recommended.
- The transformer you picked is too small for the power in your design. If you insist on using the EE16 core, I would decrease your output current to 1.0 amp.
- The heatsinking you have for your output rectifier and TOP243 device is insufficient.
- You have no fuse in your PCB layout.
- For the device you've chosen, you probably need to use an external current limiting resister. This could be the source of the audible noise you're experiencing.


For the PCB layout...You need to completely redo it and start over:
- The loop area created by your drain trace is HUGE.
- The loop area created by your clamp circuitry is HUGE.
- You're running your HVDC- trace through the bias supply ground. This is a big no-no...you're going to have major surge/ringwave problems.
- On the secondary side, move your feedback components closer to your opto.
- Redo your grounding scheme on the secondary side of the PCB.


Don't be afraid to copy one of the PCB layouts we have in our RDK boards. We have many different RDKs for the product family you've chosen as well as the power level. We provide these reference designs to make your job easier when putting together your own designs. Feel free to copy/use PCB layouts from our RDKs (provided you buy our parts of course).


-The Traveler

I will redesign the PCB and attached
the core is EE19, I hope there some tutorial for PCB design right from the sources

ssyn -


An EE19 core should get your pretty close to your power requirement. If you give me some specs, I'd be happy to setup a PI Expert design file for you to get started.


For the PCB layouts, I would *highly* recommend going through our reference designs and find something with similar specs using the same IC product family. Like I said, there is nothing wrong with modifying an existing PCB layout or completely using one all together. That's why we provide them!

-The Traveler