I wonder if there is drawing mistake on the schematic. Something around the opto's cathode return seems strange. I tried to verify with the actual PCB layout, but it is not that easy to follow. Normally, pin 2 will be connected to the cathode of the TL431 as shown below.
I wonder if there is drawing mistake on the schematic. Something around the opto's cathode return seems strange. I tried to verify with the actual PCB layout, but it is not that easy to follow. Normally, pin 2 will be connected to the cathode of the TL431 as shown below.
The drawing is ok, its just that their connection of rbias is doubtful, you really need to have a resistor across the opto diode, because like its connected there in post#2, the rbias will likely not get sufficient current through it when the smps is on full power.
Anyway, the connection configuration OP speaks of is called the “fast lane, slow lane” connection, and yes, the opto diode does need to be the upstream side of the output filter inductor
The L21 & C23 forms a post filter to reduce ripple / noise on the output this is typical in flyback designs & very difficult to compensate most of the times, refer attached PDF for detailed analysis
As you know, the output LC filter isnt difficult to handle as long as its corner frequency is well above the crossover frequency of the smps..Preferably three times higher at least.
Yes you are correct treez but getting sufficient ripple attenuation & maintaining filters corner frequency 3 times higher is the tricky part / difficult part of the design particularly if your loop crossover
requirement is higher for fast dynamic response.