Colleagues,

I’m building a SEPIC charger around MCP1631 controller. 3A charge current. I’ve examined several reference designs, and none of them had snubbers. As far as I can tell, they all were close to the ideal SEPIC topology.

Could a SEPIC converter require a snubber? If so, I would design one in just in case. More generally, are there caveats in SEPIC converters that I should watch out for?

Any suggestion, insight or reference is really appreciated!

- Nick

 

Hi Nick,

Regarding the snubber...

If your SEPIC design allows the inductor current to go to zero, then there can be ringing if there is no load on the inductor until the next charge cycle starts. This high frequency ringing usually causes no harm, but does increase the EM emissions from your circuit. A snubber can be used to suppress the ringing signal without effecting the efficiency too much. Your choice.

Most likely your SEPIC design doesn't allow zero inductor current except maybe under an output short condition where the current limit circuit has kicked in. What does your design do?

Regards,
Ifixit

 

My circuit is a 3amp battery charger. Space is at premium, so I'm planning to use as high of a frequency as my controller will allow, which should allow me to use smaller magnetics and capacitors.

Would you recommend an RCD snubber in parallel with the upstream inductor (like in flyback converters)?

I's rather design the snubber into the first version. If it will not be needed - all the better.

 

I cannot say for sure as I am not the designer, but you most likely don' t want a snubber if you are operating in the 100KHz to 1MHz range and don't care about EMI.

 

To some extent, I care about the EMI. There's a possibility that this charger will become a part of a commercial product, which will be submitted to UL for testing. At the moment, I don't know which tests UL will run.