zappedagain
Electrical
I'm seeing something odd in an isolated flyback power supply. Per my understanding, an isolated flyback needs a capacitor from the primary return path (lets call it GND) to the secondary (isolated) return path (lets call it GNDI) to give the return current a path other than chassis/earth ground; basically closing the circuit for the stray capacitance between the primary winding and the secondary winding. With the right capacitor, the ripple from GND to GNDI should only show a very weak (10s of mV) copy of the flyback dutycycle.
My flyback is running at 250 KHz, and on a few PCBAs from a lot I'm seeing about a 5 Vpp 250 KHz sinusoidal ripple between GND and GNDI, and the expected 50 mVpp copy of the dutycycle is riding on top of this. I can reduce this 250 KHz ripple by adding more capacitance from GND to GNDI. The attached document shows the difference.
Does anyone know how the flyback is generating a sinusoidal return current at the flyback frequency? Is this a sign of a stability issue?
Z
My flyback is running at 250 KHz, and on a few PCBAs from a lot I'm seeing about a 5 Vpp 250 KHz sinusoidal ripple between GND and GNDI, and the expected 50 mVpp copy of the dutycycle is riding on top of this. I can reduce this 250 KHz ripple by adding more capacitance from GND to GNDI. The attached document shows the difference.
Does anyone know how the flyback is generating a sinusoidal return current at the flyback frequency? Is this a sign of a stability issue?
Z