I’ve been away for a couple months, so someone may have already done this. But I thought it was a cool idea. I’ve had an idea of doing something like this for a while now, and finally I just got the prototype built and tested. This is a fully field oriented controller, meaning it has phase current sensors and a bunch of software and is always in current control mode, so you never need to worry about blowing a motor or controller from getting stuck in seaweed. That, and a few other features:
Current mode control for smooth operation and stall detection
A 25A PCB fuse so it doesn’t catch fire if the worst happens
Excellent FOC low speed operation for smooth control thanks to the much-loved VESC open source motor control software
48V input, to keep the battery current low and wire gauge small
30A peak motor current per phase (not continuous!)
CAN bus operation for telemetry, vastly reduced wire count, and easy integration with Ardusub
Current monitoring so you can see when the motor bearings are starting to fail
Industrial connectors for easy connect and disconnect when its maintenance time
Can fit 8 of them in the 4" electronics tube
Easy motor parameter detection via VESC-Tool
Edit: Here is a power test of the controller to the current rating of the T200
Thank you. I would definitely sell it or spread it if other people want it.
The software is VESC (open source), with a few customizations. VESC includes UAVCAN, as does Ardupilot. There is a hardware can transceiver on the board, and pixhawk has a can port as well.
Supposedly Ardusub is part of ardupilot and works with CAN now. I need to check that I can direct the motor commands to the CAN bus without having to mess with the code much.
It’s not very manly compared with my electric skateboard controller . It can probably sustain between 150W and 250W depending on environment and cooling. I haven’t got 6 or 8 of them into a tube yet for test, so I don’t know exactly how they are going to perform in that environment. It does better at lower voltages than high voltages, as one would expect.
Requirements are everything, forums are awesome to get other perspectives. I could easily trade lower voltage transistors for more power, but then i couldn’t use this design in a distributed system. If I used more transistors to spread the load, then that works for a distributed system, but they wouldn’t fit in the tube.
My experience with my BR2 is that I don’t use all that much power on each thruster, so that was the dimension i traded away. I’m always in the 25% gear. If other people are interested in this feature-set and want more power, then I would make it do that.
I realize it is a long time since there has been any activity here, but I am very interested in the FOC controller that you have made. Great work! Is this project still alive and are you selling these?
Also a question, it states that the controller can handle 48VDC. Is there an integrated step-down converter since the T200 only handles up to 20V? It that case, what is the voltage that is delivered to the thruster?
. It can probably sustain between 150W and 250W depending on environment and cooling. I haven’t got 6 or 8 of them into a tube yet for test, so I don’t know exactly how they are going to perform in that environment. It does better at lower voltages than high voltages, as one would expect.