Alejandro,
I understand it's operating in DCM. But please note that this operation is a problem because Vout ripple is almost 10V in the 28Vout power rail. It looks Vout rises fast with multiple minimun on-time pulse and the switching stops and Vout falls once Vout reaches the OVP level (DS doesn't specify whether LM5576 supports OVP or not, though). Maybe LM5576 has a same problem with LM3478 as reported in https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/726917/lm3478-load-free-output-ripple-issue. Therefore, could you check if LM5576 has burst mode functionality and OVP functionality?
Thank you so much for sharing a good material. I think it can be a possible reason beside the one I just mentioned above. But it looks the material is not covering the same case with LM5576. It's talking about VIN low dropout of TPS54231 but LM5576 is operated with Vin that is high enough. And there're some functional differences as you can see below. The boot capacitor is driven by VIN in TPS54231 but is driven by VCC (internal regulator) in LM5576. LM5576 also supports PRE pin for light load or no load condition.
1. LM5576
2. TPS54231
To sum up, I'm thinking the possible reasons for the abnormal Vout with the high ripple under no-load condition are :
1. OVP operation due to the burst mode --> Could you check if LM5576 has OVP and burst mode functionality? It's not specified in DS.
2. A bootstrap falling below BOOT UVLO and disabling HS FET (similar mechanism as described in the section 2.3 of SLUA928) --> Connecting a diode from Vout to VCC of LM5576 could help to charge C_BOOT by supplying enough current during switching IDLE period?
3. The best compensation values not being adhere to. --> Vout is well regulated in >300mA load but it's showing a burst-like operation with the large voltage ripple in <300mA load. In this case, can we get a well-regulated Vout over all load conditions if we tune Ccomp and Rcomp properly?
Please advise to solve the problem.. Thanks,
