UCC28711: Very slow startup and the Vout voltage drops the load is connected

Hello Joaquin,

I'm sorry you are experiencing this problem. Obviously, something is not working correctly and I'll try to help you to debug this circuit. Without more information, it seems like the system is behaving as if it were heavily loaded even before you apply your external load. It would be very useful to have waveforms from various points in the circuit to see what is actually happening.

But before that, it is a good idea to inspect each part to make sure that it is the correct part for that location and installed in the correct polarity. Although the schematic diagram shown (closely matching that of the UCC28711EVM-160 board) seems correct, some parts on the board you are debugging may not match the schematic. A wrong value may have been mistakenly installed, for example.

Assuming the components are all correct, your description of the behavior sounds like the circuit is trying to start up into a load that is virtually at the current limit, with barely enough spare current to charge up the output capacitors to 5V over a 7-second time span. Adding more load sounds like it exceeds the current limit and it may be cycling the primary-side VDD. But the values of 3.3V, 4.1V and 3.8V sounds like you are reporting digital multi-meter (DMM) readings, not observations from an oscilloscope. DMMs will show average voltages over their sample window and these readings can fluctuate widely if the waveform is oscillating high and low. So we really need to look at waveforms to debug this.

I suggest to collect waveforms of the output voltage, VDD of the controller, and of the high-voltage Vbulk. Also useful will be the drain voltage of the main switching MOSFET when Vout = 5V after start-up and when Vout is lower when externally loaded.

Without any of these waveforms, the only things that come to mind that may internally load down the system are: output caps installed backwards or leaking a lot (but not yet ruptured), or primary snubber clamp TVS (D10) value may be too low and acting as a load which reduces the energy transferred to the secondary side. Another possibility is that the current sense resistor R6S may be installed with a 19.6 R instead of 1.96 R, severely limiting the power being transferred. These are just guesses, and waveforms will help take the guess-work out of debug.

Regards,
Ulrich

Joaquin,
Thank you for clarifying this issue has been resolved. That feedback is greatly appreciated. If you experience any other problems perhaps this guide will be of use to you in debugging. www.ti.com/.../slua783.pdf

Best of luck with the rest of your design.

Regards
John