0
\$\begingroup\$

I'm building a single switched AC/DC Forward Converter and have reached to the point where I have steady o/p voltages and currents.

However, I am not able to remove noise that goes back into the AC i/p line. A multi-meter(DMM) placed across the Line and Neutral at the power line input jumps around wildly once the current starts flowing in the MOSFET. The o/p voltages and currents remain steady, however. enter image description here

I tried reading the AC with an analog meter, but not much difference. As long as the load current is small, it shows correct AC voltage. But as the load current increases, the meter starts showing much higher voltage (roughly 400VAC)across the input. Analog meter reading high I have tried various Y capacitors between i/p phase to Earth, Neutral to Earth, from the MOSFET source to Earth, but to no avail.

I have tried one commercial EMI filter also, but nope. I have tried with a common mode filter that I rigged up, but either my design is wrong, or I haven't got the concept right, because it did not make any difference. Some pointers/URLs for a design procedure to correctly design EMI/RFI filters will be appreciated.

My configuration is a single switch Forward converter with 230VAC Power Line Input, a reset winding (Treset=Tprimary), and a P6KE350 diode as protection from Drain to +ve bus after rectifier and capacitor. The controller is a Micro. Switching frequency is 50KHz. The MOSFET is a 6NB80, 5N90 or IRFBE30, depending on availability.

Single switch Fwd Converter The output voltage and currents readings are very stable, but not the i/p readings.

Your suggestions,recommendations and advice sought. Thanks in advance Edit: Added Schematic. My Measurement equipment are small DMMs, an old 25MHz analog scope, and not much else. I also have a Rigol-1052 50MHz DSO, but not used that for this project yet.

2nd Edit: The Micro is also being supplied from AC, with a smaller 1N4007 rectifer, and a TNY278 circuit on the same PCB. I had thought the combination of the TNY278 ckt and the Fwd converter running together could be the cause of this noise(beat frequency?) so I disconnected the track supplying the ac to the 278 part of the ckt, and supplied the micro from a 12v7AH battery, just to eliminate this as a possible cause. Again, no difference.

\$\endgroup\$
2
  • \$\begingroup\$ There is no standard forward converter - post your circuit and post details of your measurement equipment. \$\endgroup\$
    – Andy aka
    Commented Dec 13, 2018 at 9:33
  • \$\begingroup\$ That looks like the typical crappy $5 DMM. Despite labels suggesting otherwise, those are often not properly rated for safe handling of high voltages; no fuses or non-HRC fuses, not enough clearance, case that cannot withstand explosion of fuses or components. Be very careful with using such DMMs on line voltages and higher. Or better: obtain a safe DMM. \$\endgroup\$
    – marcelm
    Commented Dec 13, 2018 at 14:51

1 Answer 1

0
\$\begingroup\$

EMI filter is for fairly high frequencies (that you wouldn't pickup with your multi meter anyway) but might be a good idea to have one as well.

  • You can increase the 470uF cap at the input, but you also need to add some good ceramic cap, as electrolytic are not good at high transients. Add a (or several) 10uF ceramic cap in parallel to the 470uF cap close to the mos and coil, keep the electrolytic cap far away near the rectifier.
  • Use big traces between the ceramic cap and the mos / coil.

  • Then you can filter the input to avoid injecting noise back into the network with an inductor, place another electrolytic cap and place the coil between them as a filter.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
2
  • \$\begingroup\$ Thank you Damien, for your quick response. Ceramic capacitor of 10uF at 400VDC? I am not able to locate at Mouser or Digikey \$\endgroup\$
    – dynamag
    Commented Dec 13, 2018 at 11:25
  • \$\begingroup\$ you can use lower ceramic values, like this one: C5750X6S2W225K250KA \$\endgroup\$
    – Damien
    Commented Dec 13, 2018 at 11:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.