TOP257PN - can't reach designed voltage / voltage drops significantly when output loaded
Hi,
I designed a new SMPS as described in attached PiExpert report, with minor changes. Changes: used SF38G instead of SF35G, C3 has a voltage rating of 1600V, instead of R9 and R10 I used a potentiometer, that was preset to right voltage before startup of power source. This circuit is not intended to work with variable output voltage.
Rectified output voltage is not reaching 24V, and stays at about 22V. At startup, voltage rises for about two seconds to maximum of 22V. Without any load, output voltage rise to 22V, then driver stops working. Under light load of 1kohm, output voltage stays at about 21V. Under 100 ohm load, output fall to about 10V.
I tried diagnosing everything that I was able to do:
Transformer was made as stated in design report with primary inductance of about 1630uH
All traces are free of shorts. PCB was designed similar to my previous SMPS made with TOP252PN that was working without problems.
TL431 seems to react to exceeding 2,5V at it's reference pin, pulling down it's cathode low.
Capacitors are not shorted, and have right minimum voltage and capacitance.
Rectifier diode is working, was replaced by another diode, nothing changed.
Hot side voltage is ~ 308V when powered, falling to 295V when 100 ohm load is connected. At no load it stays 308V.
Protection like in TNY series is not triggering (pulsed output).
Bias voltage is stable 16V at rectifying diode cathode/optocoupler collector, and 5,8V at C pin of TOP / optocoupler emitter.
There is a white noise - like interference that can be heard when connected as audio amplifier power supply.
Oscillation frequency at all measure points taken is impossible to obtain, there are a lot of random pulses that look like normal pulses.
Under 100 ohm load pulse peaks get lower, there is a slight duty cycle increase, but voltage on both primary and secondary fall.
It seems that triggering of feedback loop stops TOP and there is no restart until bulk capacitor discharge. Turning off mains before bulk capacitor discharge and turning them on again don't start the controller.
When adding 100 ohm load when 1k load is already connected, voltage drops to about 10V, then when 100 ohm load is removed, it rise again to 22V.
Sometimes feedback loop is not working at all, and output voltage rises to 40V.
When there is no load, there is a sinus-like wave on both drain side and secondary side, but controller seems to not working - no signs of flyback action on oscilloscope.
I attached design report from PiExpert, and osclilloscope readings at drain and secondary rectifier diode anode. Please note that my oscilloscope is not fully compatible with 100X probe, and you must multiply all readings by 10 in case of drain voltages. Other voltages are in right range. Due to instability, frequency measurements can be not precise.
Comments
Thanks for Your reply, I will try those modifications. If problem will be still present, I will provide further information.
KK
I tried proposed alterations to R6, but problem stays the same. I tried both raising and lowering resistance. I disconnected feedback loop from output side by removing R6 and C11, but nothing changed. Output voltage still can't reach 24V and stays about 21V, and sags when 100 ohm load is used. Waveforms are the same as with loop connected. I attach layout of components and traces. I examined transformer for inductance nad continuity, everything is in specifications, primary inductance is 1660uH, secondary 60uH. Primary and secondary have right phases. I also inspected optocoupler nad 431 and there are no damages. Waiting for further advices.
KK
| Attachment | Size |
|---|---|
| pcb_2.jpg (291.54 KB) | 291.54 KB |
| layout_3.png (655.53 KB) | 655.53 KB |
After testing all TOP257PN devices that I bought by exchanging them on my board I found out that most of them were damaged or counterfeit. All of them expect one were a bit thinner after I measured their casing height, so I suspect grinding of the top and engraving with new part numbers. After installing last one that had a pin marking dot, my PSU works normal and there are no issues with voltage and current anymore. Waveforms look as expected, and feedback can be adjusted for desired voltage. Everything looks stable and react accordingly to changes in load.
Thank for your suggestions, topic can be closed.
KK
Hi Zaznaczony,
Good day.
It seems that the issue may be related with the feedback section. We recommend adjusting the resistance value of the gain limiting resistor (R6).
Moreover, you may consider adding another resistor in series with C12 to form an RC filter, similar to the Type‑II controller described in the design guide.
It would also be helpful if you could share the current PCB layout, as you're previous SMPS design operated at a lower power level.
I hope this helps. Thank you.
Kind regards,