Low voltage drop high current unidirectional diode/circuit

[deepak4you]

Experts!

Need help in continuation of a circuit I was working on, for which I had great inputs from the community. This was the original post:

https://www.edaboard.com/showthread.php?380027-MOSFET-switchover-for-E-Bike-controller

Referring to the above, I had managed to get a LFP battery with inbuilt BMS. That helped me create a MOSFET based switch alright (since charge and discharge connections are separate unlike the earlier circuit), but there is still one nagging issue I cannot find a solution/suitable component for.

I intend to use a diode to ensure that current is flowing only in one direction - from the battery to the controller and not the other direction. But 0.7V - 1V drop across the voltage takes away a lot of usage capacity out of my battery. The battery is a 48V (nominal)-52V(maximum) 15S4P LFP battery. If I deduct even 0.7V, it is considerable amount of energy wasted across the diode.

Is there a diode/schottky diode/any other active circuit that would have very low voltage drop/current consumption that would allow flow of current only in one direction? Might be I could use a high current MOS without a body diode which I could turn on in one direction along with the main MOS, and leave it OFF and blocking in the reverse direction?

Thanks a ton!
-Deepak

 

[d123]

Hi,

It's still 8A - 10A? 5.6W to 7W, or 8W to 10W...

Never used MOSFETs without a body diode so couldn't say. Back-to-back MOSFETs is another option, just double the RDSon, if chosen wisely with very low RDSon and checking the curves in the datasheets, should/could be low enough. Schottky diodes are supposed to have "high" leakage, but it might bring diode PD down to ~4W max, depends again on checking device datasheet graphs for reliable assessment.

 

[KlausST]

Hi,

0.7V from 48V is just 1.4%.
Because the current and time is the same: you waste just 1.4% of the energy.. or you have 98.6% efficiency with this. Not that bad.

Is there a diode/schottky diode

--> go to a diode manufacturer ir a diode distributor web site and use the interactive selection tool.

any other active circuit

--> look for "ideal diode" circuits or ICs.

Klaus

 

[deepak4you]

@KlausST - forgot to mention that the operating range is just about 42-52V. So 0.7V is also still 7%. That's a lot of juice wasted in an e-bike situation, hence trying to find the most optimised solution. Any diode seems to be having
@d123 - this time its in the discharge path, so higher current (23A-25A @ 48V). I actually thought of back-to-back (N-channel and P-channel combo) MOSFETs as you suggested, and drew a crude diagram as well. I think that might work. Even if the RdsON adds up, should still be better than . I'll build it and see if that works well. Will keep posted.

 

[KlausST]

Hi,

Your efficiency calculation is wrong. You have to calculate with the average (or the nominal) voltage, not with the difference.
It's not 7% loss.

Don't expect a noticable extension of battery lifetime.

Klaus

 

[dick_freebird]

I would think that an e-bike would -love- current flowing
from controller to battery - at least, during regenerative
braking. Why the concern over reverse current flow?

 

[deepak4you]

@KlausST - I'll try with the diode anyway as a start to simplify things and assess the impact.

@dick_freebird - output of the controller is unregulated and not fix whereas the battery is an LFP one. As I understand, and from some damaging experience, that I cannot just let unregulated voltage feed the LFP battery pack. Hence the need for this circuit. If you read the original post, there's a buck-boost controller as well feeding the battery to ensure that the output to the battery is regulated.

 

[schmitt trigger]

There is another thread discussing ideal diodes:

https://www.edaboard.com/showthread.php?381734-Protection-of-15A-DC-power-line&p=1637071#post1637071