Enable VIPER26K

[PScheerer]

Hi there,

My question is if there is a way to enable or disable the Viper26K with for example an optocoupler.

In the isolated flyback topology this is done cycle by cycle. But we want to enable/disable it in the high voltage buck configuration. Can we short-circuit the comp network for long periods of time? How is the internal startup circuit designed?

Thanks in advance!

[Peter BENSCH]

Welcome @PScheerer, to the community!

That's a very interesting question that, as far as I know, nobody has ever asked!

Most probably this works with an optocoupler between COMP pin and GND pin (with a galvanically non-isolated buck, not GND of the circuitry). The supply via VDD is then no longer guaranteed because the switching has been stopped, but according to the data sheet, section 4.4, paragraph 2, the Viper26K should then be supplied again via the internal power source and the switching should start again.

I would be pleased to see confirmation of this assumption.

Good luck!
/Peter

[PScheerer]

Thanks for the answer.

So theoretically it could work. If comp is shorted to device ground, my understanding is that the controller will try to start using the internal current source. But since comp is shorted to ground, it can't. In normal operation, the internal logic would very quickly disconnect the current source. In the described "disable" state, the current source would be on all the time and the controller would try to start all the time.

So it would be interesting to know if the current source and the controller itself are able to handle this situation and how big the losses will be.

We have eval-boards on hand so we can test it out, currently im a bit scared to test it, but maybe you have a good tip or an internal information about the current source?

[Peter BENSCH]

As you have already written, the Viper26K will permanently attempt to start with the IDDch1 current, which will then also quickly become noticeable through heating corresponding to the input voltage. The question then arises as to whether such a switch-off makes sense and whether it would not be better to replace it with a switch-off of the input voltage.

Regards
/Peter