I am using the LM5006 with the flyback diode to allow the regulator to take advantage of DCM when there is no load.
On one of my boards, every so often the SW ON pulse lasts longer than normal, allowing Vout to spike up to 7V (from the normally regulated 5V). The 7V then decays back down to 5V, and the cycle repeats - eventually an ON pulse is too long and the voltage goes back up.
Most ON cycles are the correct length and there is no noticeable spike on Vout.
The board seems to behave correctly when a load is connected.
The amplitude of the spike on Vout seems to be correlated to Vin - the higher Vin, the higher the Vout amplitude. I suppose this makes sense, but I wanted to make sure I mentioned it.
This behavior may happen more at certain level of Vin. For example, it might not happen at 20V, then I change Vin to 15V and it begins to happen. I am not 100% sure of this.
I have two other boards. One does not do this at all. The other does it, but very occasionally - possibly only while changing Vin.
I did see the application note about one possible side effect from a no-load situation, but that was Vout drooping from 5V down to ~2V, then back up to 5V. That situation would be ok, and doesn't worry me because it is understood. I don't know why the regulator is behaving like this, and that concerns me.
Any feedback is appreciated. I can provide more data or perform any experiments.
Thanks,
Dave