2 controllers on BBS02B experiment.

Hi all.

Stupid question, I have these 2 Chinese 48 volt 17 amp controller that I use on my bbs02 and was thinking about running both in parallel at the same time to see what happens. They are both identical and both run exactly the same. I wonder if there would be any benefit to this. More amps maybe? Cooler running controllers? They are 6mosfet controllers.

No is the simple answer.

Maybe if you do a bunch of special wiring that would take longer and be worse and cost more in time than just buying one bigger better controller is the less simple answer.

Voltron said:

No is the simple answer.

Maybe if you do a bunch of special wiring that would take longer and be worse and cost more in time than just buying one bigger better controller is the less simple answer.

Click to expand...

Well the wiring is not gonna be a problem I've got that figured out. I just don't want to blow up my battery or controllers when hit the ON switch. Since my controller seem to strain while under load and get very hot I figured maybe doubling up might work. They are self learning so I have set correctly to drive the motor forward and all wires are the same colors. Since the second controller is my spare I don't want to blow it up by some unforeseen conflict. Thanks for your answer though.

The throttle setup is simple, one controller has all 3 wires connected with the second only receiving the signal. Brake switch will be "Y"'d to cut power to both controllers. But my concern is "Y"ing the phase wires. BTW colors of both controllers phase wires for proper direction are identical.

The special wiring means:

Cutting all the traces from the MCU to the gate drivers on one board.

Adding wires (probably shielded, or at least twisted pair) from the MCU on the other board to the gate drivers on the cut board.

Paralleling the battery and phase wires of both controllers.

All other wires on the cut-trace controller are left disconnected, as none of them are active anymore. (unless the gate drive is powered by the low voltage power supply section, and requires something from the MCU/etc to be active).

The reason this is necessary is that you can't just parallel two brushless controllers on one motor (or two motors on one controller unless they have locked-together rotors in the correct relative position), because the timing of each will be slightly different, which could be enough to cause shoot through between FET stages on each different controller, under the wrong conditions.

That could fry both controllers. There've been a number of threads about this sort of thing, if you poke around.

If you cut the MCU of one controller out of the loop, and just parallel the power stages as noted above, then you're only using one control circuit (one MCU) so the timing should be close enough to identical, as long as your wires between MCU and other controller's gate drivers are very short.

There is still the risk of induced noise on those wires, that could trigger gate/FET activity that could cause shoot-thru.

amberwolf said:

The special wiring means:

Cutting all the traces from the MCU to the gate drivers on one board.

Adding wires (probably shielded, or at least twisted pair) from the MCU on the other board to the gate drivers on the cut board.

Paralleling the battery and phase wires of both controllers.

All other wires on the cut-trace controller are left disconnected, as none of them are active anymore. (unless the gate drive is powered by the low voltage power supply section, and requires something from the MCU/etc to be active).

The reason this is necessary is that you can't just parallel two brushless controllers on one motor (or two motors on one controller unless they have locked-together rotors in the correct relative position), because the timing of each will be slightly different, which could be enough to cause shoot through between FET stages on each different controller, under the wrong conditions.

That could fry both controllers. There've been a number of threads about this sort of thing, if you poke around.

If you cut the MCU of one controller out of the loop, and just parallel the power stages as noted above, then you're only using one control circuit (one MCU) so the timing should be close enough to identical, as long as your wires between MCU and other controller's gate drivers are very short.

There is still the risk of induced noise on those wires, that could trigger gate/FET activity that could cause shoot-thru.

Click to expand...

Excellent post thank you, I understand what you mean. You taught a newbie something today Again thank you.

Yeah.. By the time I built the custom harness with all the paralleled etc etc and then learned you had to go into the board and then opened up the controllers and realized they were fully potted on those particular ones... well... I realized everybody that it hadn't worked for in all the threads I searched (thanks to the wonderful search function on here) after actual trying it and having it not work were right.

On the surface this appears to be a recipe for disaster, and it certainly could be, but as I recall paralleling controllers has been tried and found to work, at least in some cases. I don't recommend it, but if the controllers are trapezoidal types, use the same hall sensor signals, and they are using the usual freewheeling diodes (not synchronous rectification), it did seem to work.

It would be worth finding the old thread on this subject and reviewing it. I expected it to blow up, but when it did not some closer analysis was interesting. (just to be clear, it was not my experiment, I just reviewed the results and looked into why it did not fail).

If you analyze the hall and BLDC 6 state drive sequencing only one hall changes at a time, and only one motor lead changes at a time, and the sequence doesn't produce a rail to rail short on adjacent states that inevitably overlap slightly. So if both controllers are wired correctly and have already learned the correct drive sequence the minor timing errors don't produce destructive currents. Or at least they did not for that test.

If this goes wrong it could produce a battery to ground short through the FETs of both controllers which would likely blow both controllers. When it works, the two controllers share current into the motor. A fast BMS protection event might prevent controller damage in the shoot through fault case.

This is totally different from running two motors on one controller. That doesn't work unless the motors are physically timed and locked together. But in this two controller one motor case the controllers are locked in phase by the motor's own signals so they can cooperate.

Here is one thread where two controllers on one motor was tried and it worked, with both sensored and sensorless modes:

https://endless-sphere.com/forums/viewtopic.php?t=75186