Flyback LED driver has poor quality feedback loop?

[Flyback]

thanks, yes you are absolutely right, yes
And of course I am sure you would agree that because that kind of additional testing is needed, there's just no point in doing it the way that the first post describes.....instead , just do it ChrisP58's (post #36) way or the other way involving the high side current monitor.-Then avoid the necessity for the additional testing that you describe.

 

[MrAl]

Hi,

Well, what i was also saying was that the designer might have already known about this issue even for some time and it was known that it would work ok, even with the proposed test. I cant say for sure until we see the results of the test, and even then we dont know who designed it personally, or at least i dont, so i cant query them directly. Remember the old saying, "We mock what we dont understand" <chuckle>
Some designers have a wealth of understanding under their belt which doesnt translate easily to paper.


But if you prefer to just bias it then i dont see any problem really. Just a little extra work. Just be sure to test the result thoroughly.

 

[Flyback]

Thanks, the same designer has also used a primary clamp snubber on the 100W flyback which comprises just a SMA UF diode going into a 220V , SMB size TVS.

(ie not an RCD clamp, or an RCDZ clamp, so no power resistor to dissipate the energy from the leakage inductance)
If i had access to the PSU then i'd thermal test this, because such a small primary clamp/snubber is unheard of for a flyback of 100W. The TVS is also on minimal land pads

the spec for the flyback is vin=418v, vout=40v, iout=2.5A.
Switching frequency is 75KHz.
Transformer primary is 560uH, and NP/NS = 3.19
Ipk primary = 2.2A

The opamp spoken of is regulating the output current of this flyback