Neat. I'll order a DBB* battery combiner and test it out. Should be pretty easy to wire a charger to the output and two low batteries of different voltage to the input. Then test and make sure the higher battery doesn't charge the lower with a simple voltage check at the lower battery. Then power on the charger and check and make sure the voltage from the charger is showing up at the battery terminals.
If it does, regenerative braking and controllers that need a direct connection to a battery to feed back surges to should be supported.
As long as FETs are being used and not diodes, voltage drop and wasted watts and heat generated should be low.
Supporting both a charge port and regenerative braking tends to be rare. Even the company that makes the original combiner circuit this thread is about doesn't offer regenerative braking support in their charge port model:
I had been contemplating upgrading my ideal diodes to these parallel arrays for even less voltage drop:
Or just building my own contactor based solution. Contactors have no voltage drop and are bidirectional, but need some low voltage power to work (electromagnetic coil to keep a switch closed, etc.). Feels like it would be pretty easy to use a voltage comparator like:
Then wire it like:
Battery A positive -> resistor A -> voltage comparator Vin pad
Battery B positive -> resistor B -> voltage comparator Vin pad
Voltage comparator out -> Contactor on/off pad
Battery A -> Contactor normally closed pad
Battery B -> Contactor normally open pad
Contactor out pad -> controller
Feels like there'd always be a battery directly available to the controller for regen that way, but one battery would never charge the other.
Previously I bought both a two battery port and a three battery port plus charging port MOSFET based battery combiner box off AliExpress over the years. Neither supported passing regenerative current back to the packs. Kills my baserunner if I go fast downhill with them too since no where for power surges to go.
Practically, this is because these devices typically just use an IC without paying for the copyright for it, and the IC chosen usually prevents reverse current intentionally to protect power supplies. E.g.:
Description
Positive High Voltage
Ideal Diode-OR with Input Supply
and Fuse Monitors
Features
n Replaces Power Schottky Diodes
n Controls N-Channel MOSFETs
n 0.3µs Turn-Off Time Limits Peak Fault Current
n Wide Operating Voltage Range: 9V to 80V
n Smooth Switchover without Oscillation
n No Reverse DC Current
I suppose because MOSFETs need to be powered to work and they don't add the circuitry to power them when the voltage at the output is higher than the voltage from either pack, like happens in a regenerative braking situation.
Conceptually, though, it also takes some extra components to determine if battery A is seeing a higher voltage than its state of charge because A) battery B is higher voltage and driving the load, a condition that requires blocking that incoming voltage to battery A to prevent the batteries from charging each other, or B) the controller is performing regen and trying to charge it, something we want to allow. Both manifest as higher voltage at the load.
You can see in the simplest possible regen dump circuit that needs to detect and redirect incoming current from the controller there's, surprise surprise, a diode just like the original circuit diagram in this thread:
So you sort of need to do that, but the opposite. Allow the regen current, but block the batteries from charging each other.
You could isolate the currents with a bunch of ideal diodes into separate charge and discharge buses, or put an ideal diode OR between the batteries and the regen dump so it is only comparing the highest voltage battery vs. the controller output, but that increases cost and complexity again.
