Buck converter, reverse flow diode, forward voltage drop

[rks96]

In my application case I am charging the battery with a buck converter using CC and CV mode. For example, when charging 3cell lithium battery I would use 12.6v CV and 1a cc. However, my use case involves leaving the buck converter connected to the battery while input is disconnected, which causes power being fed back in buck converter (wasting power, ect).

My solution is to connect a diode in series with battery to allow current flow only in one direction. I am using IN5822.https://datasheetspdf.com/pdf/675676/CHENG-YIELECTRONIC/IN5822/1

My question is what CV voltage should I set on a buck converter to charge battery fully? Does diode affect maximum voltage battery will be charged or as charging current drops so does forward voltage drop resulting in battery still being charged fully without rising CV above 12.6v in the previous example?

Thanks!

[Paul Rose]

See attached chart: forward voltage drop vs forward current for 1n5822

[Manul]

Lithium batteries need really accurate charge voltages to be safe and in good condition. Diode forward voltage is not stable enough, it changes a lot with current, temperature and also manufacturing variations.

If you use series diode, then you should at least take feedback voltage after the diode, then buck converter will auto compensate for diode drop. Probably you need: output capacitor - series diode - second output capacitor. Take feedback from second output capacitor. There could be some loop stability issues though.

I would strongly suggest to use IC dedicated for charging. They let you charge more accurate, safe and are designed to have a battery connected at the output. Also you probably want some cell balancing for 3 cell battery.

[Zero999]

A diode could be placed between the feedback potential divider and buck converter output, but I echo the idea of using a dedicated charging IC.

Another option is to use an ideal diode controller IC, as I mentioned in the post linked below:
https://www.eevblog.com/forum/projects/diode-to-prevent-current-backfeed/msg3353524/#msg3353524

[David Hess]

A diode can work if the requirements for output voltage regulation are not too stringent.  Otherwise the common solutions are:

1. Use a high feedback divider impedance to minimize current draw.
2. Buffer the feedback divider with a follower.
3. Disconnect the feedback divider with a FET when power is removed.

[Doctorandus_P]

I have not much experience with Li-Ion, but just thought:

Three of these cells in series need a charge balancing circuit.

so if the final charge voltage is "just a little bit" too high and current limited. Then maybe you you can use the balancing circuit to eat up the difference.

Better is probably to put the whole charge circuit near the battery, and just apply a raw voltage that's plenty to charge the batteries.