Isolated Feedback for Flyback

[berni80]

isolated feedback

Hi,

I would like to have you opinion on this simple topology (see the attached image) to
get an isolated feedback from the secondary of a flyback.

Does this circuit work?

I think the CTR of the opto has a big influence on the regulated supply.
Is it true?

Could you suggest me the simple way to get an isolated feedback?

Thanks!!

 

[FvM]

flyback feedback

Variable reference voltage/shunt regulator ICs as TL431 or LM385-ADJ are often used in commercial products and suggested in application circuits. You'll find many references at EDAboard.

 

[berni80]

isolated feedback

ICs as TL431 or LM385-ADJ

I saw a lot of circuit that use these ICs, but which kind of advantage give these ICs?
Compared to the solution I propose.. does it work?

I would like to avoid an ICs on the isolated side, for space needs.

 

[FvM]

flyback with primary side feedback

If you don't need exact regulation and can accept the temperature drift introduced by the optocoupler, your circuit can be used.

I don't think, that the shown pin assignment meets any usual opto coupler, B.T.W. They would require to left pin 6 unconnected.

 

[berni80]

feedback control loop + flyback

Thanks for your reply.

The opto with this pinout is the ACPL-M43T by Avago.

Here the datasheet:
https://media.digikey.com/pdf/Data Sheets/Avago PDFs/ACPL-M43T.pdf

I do not need a good regulation, but I'm worried about the Current Transfer
Ratio (CTR). It has a typycal value of 45, but can vary form 32 to 80.

Does the CTR influence the regulated voltage??!
Or it is compensate in some way by the control loop?

 

[FvM]

flyback opto isolated feedback

Does the CTR influence the regulated voltage??!

Obviously.

Or it is compensate in some way by the control loop?

Not in your circuit.

 

[berni80]

flyback isolated feedback tl431

I try to better explain my doubts..

With some approximation, in steady state:

CTR = Ir1 / Ir2 = Vout/R1 * R2 / Vfb

Vout = Regulated Voltage
Vfb = Voltage feedback usad by the swithcing IC
Ir1 = current flowing into R1
Ir2 = current flowing into R2

So, Vfb = Vout * R2/R1 * CTI

This mean that the Vfb is strongly influenced by the CTI.
I'm right?

But also using the TL431 as a reference the situation is the same...
I gain thermal stability but I cant compensate the CTI.

 

[FvM]

isolated feedback flyback

The TL431 does the voltage compare on the optocoupler primary side and transmits a PI amplified error signal. Because of the integrator, CTR variation doesn't affect the output voltage.

For a rough regulation, a simple Z-diode feedback as suggested in a Powerint application circuit could be sufficient.

 

[hakeen]

isolated flyback feedback

Hi Berni,

I guess your problem is linearity, not to keep forward current constant. This is why you search for an tight input-output spec, right? ACPL-M43T is an digital coupler rather than analog (OK it is open collector output, but CTR is loose).

I would recommend a high linear optocoupler like HCNR201 from Avago. CTR is much tighter. Moreover on the primary side there is an error feedback photodiode for further correction.

Datasheet:

Hope it helps.

 

[FvM]

flybac feedback

I would recommend a high linear optocoupler like HCNR201 from Avago.

The discussion would be meaningful, if voltage adjustment at the SMPS primary side is intended. Otherwise, it doesn't make much sense to improve a basically limited design concept instead of using a known better one, I think.