솔루션 찾기 기술 지원

problems with TNY278G

Posted by: TAStreck on
Dear All, we have problems with a SMPS-Design by the TNY278G. For the application ciruit see the attached file. This design is in production and on a small quantity of boards we do have a special problem. If we switch on the curcuit with the highest input voltage of 230Vac the output voltage of +12V overswing to 14V ... 15V. The TNY278G switched of until the output voltage goes to +12V and then continues with switching. This power off during power on is critical for an additional winding of the transformer and the additional curcuit connected to it. Wenn we removed the capacitor C30/100nF in the design this problem don´t occur. But we we have done all the nessasary tests like funktion, EMC, temperatur and clima with the design with these capacitor. Our question is, if we remove this capacitor do we have any restrictions. Do we have repeat all the tests for EMC, Temperatur and Clima ? Best regards Thomas Streck Bontronic GmbH

Files

첨부 파일 파일 크기
Bild Schaltregler TNY278.pdf 300.11 KB

댓글

Submitted by PI-Spock on 09/09/2008
Your schematic shows C30 in parallel with R10. Thats not how its supposed to be... You need to sever that connection and let c30 be the loop compensation element. Most likely this is the cause of the problem.
Submitted by PI-Chekov on 09/09/2008
Hello Thomas I agree with PI-Spock, break the connection at the node formed by R22 and C30 to R10 (see attachment). Just removing C30 is not recommended. This will leave the TL431 open loop with high gain and may cause very high frequency oscillation. I also noted the value of R14 is higher than normal. I would recommend 220-470 Ohms. See screen shot from PI Expert attached. Regarding re-testing, I doubt there will be any impact on EMC and temperature (didn't understand clima) however I'd recommend checking. For thermals simply comparing the input power is sufficient. If the input power is the same or lower then thermals (temperature rise of supply) will be ok. For EMC a pre-scan of both solutions will confirm this. Cheers PI-Chekov
Submitted by TAStreck on 12/08/2008

In reply to by prasun

We have changed our curcuit as you suggested ( see also the attached pdf-file) and made some expierience in our test facility for some weeks. Now with these new curcuit design we have no problems. We have tested these changes on at least 10 pcbs. Now we want to put these changes in our series devices. Can you look over the attached pdf-file ? If you have any doubts about the function of the changed curcuit, please let us know ! Best regards Thomas Streck Bontronic GmbH
Submitted by TAStreck on 12/08/2008
We have changed our curcuit as you suggested ( see also the attached pdf-file) and made some expierience in our test facility for some weeks. Now with these new curcuit design we have no problems. We have tested these changes on at least 10 pcbs. Now we want to put these changes in our series devices. Can you look over the attached pdf-file ? If you have any doubts about the function of the changed curcuit, please let us know ! Best regards Thomas Streck Bontronic GmbH