UCC28951: UCC28951 max load turns off X'FMR current issue

Red,

Thanks for reaching out here. The team supporting this device is in the US and on holiday presently. Please expect a response by Tuesday next week.

Regards,

John

Red,

During the last ~1ms, when the input voltage is minimum, you can clearly see the transformer is in saturation (the sharp "hockey stick" curve on the current waveform). B=V*ton/N*Ae, so you can try:

1. increase the number of transformer turns (N) to support more magnetic flux
2. increase to a larger core size with more cross section (Ae) but the same or larger area product (Ae*Aw) to support the same winding structure
3. Keep everything the same but increase the frequency which will decrease the on-time (ton)
4. There also appears to be a small amount of DC offset present in the current waveform (Von*ton>Voff*toff). It could be that your current probe needs to be degaussed. Try the measurement again and make sure the probe is degaussed first. If you're convinced the probe is at zero, try adding add a small capacitor in series with the transformer primary and see if the current waveform becomes more symmetrically balanced.

I expect the problem you are observing at room temp will be worse at high temp. The last ~1 ms showing saturation may happen over the last several ms at high temp. You need more design margin built into the transformer but at least you've shown the effectiveness of the current limit working correctly. 

Regards,

Steve M