I’m planning on using two odrives for a hoverboard based longboard, and I would like it to have regenerative breaking. I saw a comment that if a resistor is not used on the aux pins that the current will get dumped back into the input.
Is that true?
I plan on way way overbuilding the battery. probably a 13s5p or greater so I’m not worried about over charging the battery. I also will probably only charge it up to 90% from the wall for longevity.
If not, that means I would need a buck/boost converter on aux, into a battery bms/charger right?
My understanding is that the braking system on the ODrive works as follows: When braking the ODrive allows current to flow into the VBus rail. If the VBus voltage rises above a certain threshold (and a braking resistor is installed) a MOSFET opens to pull the VBus voltage back below the braking threshold. If the VBus level passes a second thresholds, the ODrive enters an error state and allows both motors to spin freely. I believe that the two parameters that you need to tweak are odrv.config.dc_bus_overvoltage_ramp_start and odrv.config.dc_bus_overvoltage_ramp_end. I’m personally not aware of anyone who has implemented a useful regenerative braking system with the ODrive so you may need to do quite a bit of experimentation on your own to make something work.
That’s true, but as robopilot says, in the newer firmware (e.g. 0.5.1) the AUX Brake resistor control has been improved. Even with a resistor connected, it will only use the resistor to prevent the battery from overcharging.
When you generate torque in the opposite direction of motion, this will cause charge to go back into the battery. The brake resistor is only used if this causes an excessive rise in voltage.
That said, your battery sounds dangerously large. Why do you need such an oversized battery for a skateboard? How many hundreds of miles are you planning to ride on each charge?
And please be aware, such a battery will become an incendiary bomb when you crash it.
It works for my e-bike.
However, the braking torque is quite severe for even a modest amount (1A) of regen current. It’s usually accepted wisdom on e-bike forums that regen is not really worth it, because the hysteresis drag of having a direct-drive motor (with no freewheel ratchet) outweighs the benefits of regen for a small vehicle such as a bike.
In an electric car/bus/train where the inertia of the vehicle is a major energy store, regen is very important. But for a skateboard, I doubt it’ll make a lot of difference to range.
Well, not exactly. It allows max_regen_current before it starts trying to put any additional regen current through the resistor (feedforward term), and then you can optionally enable the feedback voltage control term.
