Advice for using an ODrive as a dyno controller?

Hi all,

I want to use an ODrive as a load controller for a dyno for testing other controllers and motors. This means I would be running it entirely within the regen braking regime. I have a few questions about this:

  • Is there a way I can use the ODrive with only the logic power supply, instead only using braking resistors on the DC bus supply? How would I go about sizing those resistors?
  • How accurately/precisely can I control very low torque loads if I’m using a low KV brake motor (such as the M8325s 100KV)? Ideally, I’d like one system to be able to work for testing both other low and high KV motors; it would need to have a range of something like 0.05Nm ~ 5Nm
  • Are there any safety considerations I need to be concerned about when doing this (braking resistors coming disconnected and DC voltage spiking?)

Thanks for any advice you can offer me!

Hi! This is a good question.

Is there a way I can use the ODrive with only the logic power supply, instead only using braking resistors on the DC bus supply? How would I go about sizing those resistors?

You’ll still need to power the DC bus (DC+/DC-), but yes it can be with a much smaller power supply. Keep in mind though that there are operating regimes where you’re braking but the motor is still drawing current – essentially where the power needed to put current through the coils is greater than the regenerated energy, usually at slower speeds+higher torques. So it’s good to make sure you have at least a hundred watts or so (e.g. the M8325s can dissipate up to 230W peak at standstill, but a 150W supply should be adequate).

In terms of brake resistors, whether you’re using the S1’s onboard brake resistor driver or an external regen clamp, the resistance sizing should be the same math, example here: ODrive Regen Clamp Datasheet — ODrive Documentation 0.6.12 documentation . Assuming a 48V supply, a 1ohm resistor should be fine – that’ll be able to dump up to 2.3kW (just per 48V / 1ohm = 48A, 48A @ 48V = 2.3kW). Power sizing will be a bit harder – depends on how long your dyno runs are, how much power you’ll be dissipating, etc. Wirewound resistors have a pretty respectable amount of thermal mass and can handle peaking pretty hard, so if you’re doing maybe a ~15s run at 1-2kW brake power, I’d imagine you’d be fine grabbing something in the ~500W category and making sure it has time to cool down between runs. Perhaps five of these in parallel, for 750W rating total?

How accurately/precisely can I control very low torque loads if I’m using a low KV brake motor (such as the M8325s 100KV)? Ideally, I’d like one system to be able to work for testing both other low and high KV motors; it would need to have a range of something like 0.05Nm ~ 5Nm

Pro/S1 current sense accuracy (and thus torque accuracy) is around +/- 5%, and current sense noise is around 100mA (== about 8.2mNm with a 100KV motor). At low speeds, cogging torque starts to impact things as well, though anticogging helps a lot. 50mNm feels a bit low but doable on a 100KV motor. You can also sometimes find a used torque transducer on eBay for pretty cheap, which is helpful if you want a ground-truth.

Are there any safety considerations I need to be concerned about when doing this (braking resistors coming disconnected and DC voltage spiking?)

The ODrive will generally self protect. Connect things well and I don’t think you’ll have an issue. Do keep an eye on the brake resistor temperature though – those big wirewound resistors are usually happy up to around 300-350C, but you definitely want to make sure it doesn’t burn through your desk/carpet (I speak from unfortunate experience). May be worth epoxying on a thermocouple or PT1000 element as well, just to keep an eye on things.

Lots of good info, than answers my questions, thank you very much!

1 Like