12V-flyback converter-problem with regulation
Hi
We have designed a 300Vdc to 12V converter (12V@10A).
Specifications:
1. Input: Regulated 300Vdc (from PFC)
2. Output: 12V (+/- 5%) @ 10A
3. Operating freq = 132kHz
4. Operating ambient temperature range: -40 to +85 deg C
5. Device used: TOP257eN
We developed and tested the prototype board. We are facing the following issues with the converter:
1. The output voltage is going out of regulation as load increases.
2. When the voltage on ref pin of TL431 was measured, it read a voltage lesser than 2.5V.
3. The waveform on this pin showed a periodic noise burst (around 65MHz), which increases in amplitude as load increases.
4. This noise is present even on output terminals after the post filter.
5. We do not know the source of this noise.
Following are attached for your reference:
1. Schematic
2. 4-layer PCB layout snapshots of the converter section and Waveforms captured on TL431 ref pin and output terminals (pdf document).
Request you to please provide your feedback and comment on the following:
1. Layout - Is it OK or are there any corrections / improvements?
2. Reason for output not regulating.
3. Source of this noise and how to eliminate / reduce it.
Regards
S.Prashanth
Files
| Attachment | Size |
|---|---|
| Noise-on-Vref.pdf | 311.5 KB |
| 300vdc-12vdc-conv.JPG | 308.28 KB |
Comments
Hi
1. The inductor L1 was put to reduce noise pickup. However, I tested by removing L1 and directly shorting its pads. The waveforms that I shared are without this inductor.
2. The circuit was designed using PI Expert Version 8 software and compensation component values are as selected by the PI Expert software. Components marked as NM are Not Mounted. They are only place-holders in case any tweaking is required on the compensation.
Request your comments and advice on the other queries.
Regards
S.Prashanth
Hi,
1. Our datasheets and application notes have detailed guidelines regarding board layout. Please verify that these guidelines have been followed.
2. The typr of noise burst that you have shown is often probe pickup and is not real noise. A somple way to verify that is to connect both terminals of the probe to the same point on the circuit board. If you still see noise bursts that confirms that it is noise pickup.
3. Please use a probe similar to the one we show in our DERs for measuring any voltage ripple on output. This will minimize effects of noise pickup.
Please see page.13 of the file indicated in the link below:
http://www.powerint.com/sites/default/files/PDFFiles/der59.pdf
Regards
PI-Sarek
Hi,
Can someone guide how to drive transfer function of TNY switch, for loop stabilisation/
Hello Dev0509,,
We are more than willing to help you in your concern. However, I would like to request if you can create a new ticket for this one since this thread is old already. We'll wait for your post.
Thank you and Regards,
PI-Lupin the Third
Hi,
Something appears to be unusual about the feedback loop design in your circuit. Why do you need L1? This inductor is shown in series with the feedback divider network.
Some of the component values shown in the feedback compensator circuit appear to be out of proportion. I request you to verify your design using PI Expert which is a free software that can be downloaded from our web site and will provide you the appropriate compensator component values:
http://www.powerint.com/en/design-support/pi-expert-design-software
Should you need further help, please contact our sales office in your country and one of our FAEs will be able to help you with this issue.
Regards
PI-Sarek