Part Number:UCC28070
Hi John,
My PFC prototype using UCC28070 is alive in the sense that it regulates the loaded output to the desired +90V voltage. The DC output signal has a ripple of about 1 Vpp and the ripple has the shape of a 100Hz sinewave. However, there are spikes of a few voltage in amplitude appearing on the output signal. These spikes appear periodically every second or so, which means that the output signal varies from perfectly clean to very noisy. See various pictures in attachment showing the noiseless or noisy output.
I measured the GDA and GDB chip outputs and I can see that the duty cycle of the PWM is not stable. With a 12 Ohms load (90V across it), the two GDx signals vary from about 50% duty cycle to about 30% duty cycle on both GDx outputs. See video in attachment. The variation seems cyclic too and I would say it has a similar period as the spikes appearing and disappearing on the output.
Is this the sign of an instability of the compensation around UCC28070?
besides, the measured VINAC signal has not got the perfect shape of a rectified signal. Every other one cycle, the voltage doesn't go back to zero but stays at about 0.6-0.7V before going back up. See icture in attachment. Could the chip prevent the signal from going back to zero? See picture in attachment.
FYI, the circuit diagram is almost identical to the one used in your eval board, expect that it was designed to deliver 1000W using the Excel design tool that you provide to work out the values of the components around the chip. There is a mains transformer in front of the PFC module dropping the voltage to 48V AC. The output filtering capacitor is much larger than what was recommended in the spreadhseet (recommended 6mF, fitted 47mF). See spreadsheet in attachment.
Thanks i advance for your assistance.
Best regards,
Raphael