My isolated flyback converter is not starting. This is a 2nd issue prototype - issue 1 worked, but issue 2 which has no significant changes in design, doesn't start. I have observed that the voltage on the feedback pin droops as the pulse width increases during the soft start - even with the opto removed. So this droop is not a genuine feedback effect. When the droop gets to 4V, the soft start terminates. It looks to me as if the reference is being loaded and this is triggering a fault condition causing a restart. The Vdd voltage is well above 8V when this happens. In fact, the restart only occurs when the Vdd voltage has dropped to 8V sometime later (no switching occuring) and then re-charged to 13V.
Since the reference does not directly exit the chip, I can't see what could be loading it. There is nothing unconventional about my circuit.
During the short time that switching occurs, the duty cycle is still being controlled by the soft start ramp because the output winding and the Nb (aux) winding do not reach their final values. I have tried various things like altering the soft start cap, changing the current sense resistor, removing Rovp1, varying the load, etc to try to distinguish cause and effect and while the voltage attained at outputs, the duration of the hiccup, etc vary with different trials, the droop on the feedback pin reaching 4V and then triggering a fault is the only consistent behaviour. I have every reason to believe that if this effect were not occurring, the outputs would reach their intended values, the feedback would kick in and the circuit would run as intended.
Has anyone got any ideas?
