High performance motor controllers

When I got the batch manufactured, I asked CircuitHub to take a video, and here it is!
Industrial pick and place machines are so sexy! Hopefully ODrive will enable inexpensive machines that go this fast ;D

Video link!





Cheers,
Oskar

Nice vid!

I'm totally inexperienced with buying BLDC. What specs should I focus on to get a motor that is light and relatively strong at low rmp, or are they all very similar?

With the worms I'm thinking that I get holding torque for free from the friction, but I guess I'll have to test what works best.

The Hangprinter extruder might require a lot of power, so I'm also interested in fast extruders. Even if movements won't be as fast as in madcowswe's demo, a large nozzle will allow a lot of plastic per second to flow through.

---

torbjornludvigsen.com

Quote
tobben
Nice vid!

I'm totally inexperienced with buying BLDC. What specs should I focus on to get a motor that is light and relatively strong at low rmp, or are they all very similar?

Thanks!
I made this motor selection guide: Link. Let me know if that helps to pick a motor, or if there are some other metrics you want me to help calculate.

Cheers,
Oskar

The spreadsheet is really helpful, thanks
If I understand correctly, I think a low Kv value would be good to limit the need for gearing?

Looks like lower weight goes hand in hand with higher Kv? It that is the case, there will be an optimum where increases in mass of gears outweighs mass reductions in the motors themselves...

This was the best Kv/weight I found by a quick search: [hobbyking.com]

Base speed: 19200 rmp at 24 V bus voltage

Does torque decrease linearly if we lower bus voltages?
Does torque in fall off quickly if we run a BLDC motor below its base speed?

Lines will be quadrupled and I need to go hunting for good gearboxes and/or motors below 100g with Kv below 800.

How do I pay for the PCB and how do you ship it?

---

torbjornludvigsen.com

Yeah lower kv is good for torque. In fact, torque is proportional to [max current / kv]. You can see how it is calculated in the spreadsheet. I would recommend that you make a copy of the spreadsheet and enter your motor parameters and see what you get. You can make your own torque density column.

The motor you picked out is good in terms of kv, but the current rating is very low, so it might not be very strong.

Torque is proportional to current, and stays the same if you lower the bus voltage. The base speed scales linearly with bus voltage. For your application with low speeds, you can easily run at 12V.
Torque does not fall off, it is flat over all speeds.

I'll email you with the details from the info you entered into the form. It will probably be with PayPal and I'll post it to you with some reputable carrier.

Cheers!

Nice, sounds like motors for multi-rotor use driven by 12 V might do the trick =) I also discovered motors for gimbal use (with more poles but often also thinner wire in windings if I understand correctly). These often look usable, but max current is often not specified.

My copy of spreadsheet: [docs.google.com]
Hobbyking's gimbal motors: [hobbyking.com]

Cheers!

---

torbjornludvigsen.com

Gimbal motors might work, but you need to be careful with the max current. Like you said, it's not always specified for gimbal motors, but they do specify resistance. So you can do some calculations:
The power dissipation of the motor is P = 1.5 * I^2 * R where I is the current and R is the phase resistance that is quoted.
Suppose the quoted resistance is 12 ohm, and it is a small motor which we can guess will only handle about 3W of heating, we can use the formula to get that the max current is 0.4A. Which is very little xP.

Cheers,
Oskar

That would give us 2 Ncm of torque from a 140 kv motor... It might just be enough.

Where does the factor 1.5 come from?

I'm thinking about doing the first tests with this motor (only 72 kv): [hobbyking.com]
In hope that we can avoid crazy gearing.

---

torbjornludvigsen.com

The factor 1.5 comes from it being a 3 phase motor. I can elaborate if you want?

Okay, gimbal motors run with such small currents and large voltages that the control mode will be different (open loop current control instead of closed loop).
But that's okay, it shouldn't be too hard to make that configurable on the ODrive.

I added another lighter motor (90g) to the top of the original spreadsheet if that is helpful.

Cheers

I think I understand, 3 wires can carry 3/2=1.5 times as much power as 2 wires if average current in all wires are equal.

Do open loop current control have disadvantages I should think about?

Do you have recommendations for a good introduction to BLDC control theory? I'm having mixed luck with web search and Wikipedia.

The lighter motor is helpful, thanks =)

---

torbjornludvigsen.com

Hm its not quite like that. The power dissipated in a single resistor is I^2 * R. But in a three phase motor you have current flowing in all three windings with unequal distribution.

In the above image, imagine current flowing from A to equal parts arriving at B and C. Then we get Pa = I^2 * R, and Pb = Pc = (I/2)^2 * R. If we sum everything together you get the factor 1.5.

Open loop current control is fine for high resistance motors like gimbal motors.

For a good introduction you can read this post on Shane Coltons blog: Everything You Ever Wanted to Know About Brushless Motors. He also links to a masters thesis in that post that is a good continuation that goes into very much detail. This quick reference is the one I use the most often: DYNAMIC MODEL OF PM SYNCHRONOUS MOTORS.

Cheers

Ah, I see. Nice figure

I read the blog post. He's a good and enthusiastic writer! I learned that energy conservation analysis is an awesome shortcut to a lot of BLDC motor understanding.

Skimmed through DYNAMIC MODEL OF PM SYNCHRONOUS MOTORS. Nice and condensed equation derivations.

Looked quickly at the Master's. It really has unusually good looking hand drawn figures.

Cheers

---

torbjornludvigsen.com

I am excited to let you know that the time has finally come for the alpha stage of the ODrive project!

Alpha release
This consists of two components, the alpha software release and the manufacturing run for ODrive v3.2.
The alpha release will include all the core features of the board, and allow go-to commands sent over USB, as well as step/direction interfacing. You can check which features will be included in the release on the roadmap.

ODrive v3.2
The new design is ready, and can be reviewed on github. The design will go to production on Monday the 17th of April, and you can get in on the production run here.



More info
If you want to read more about it, you can check out the blog post.

Hi,
Just wanted to let you guys know that there is an ODrive community getting started over at [discourse.odriverobotics.com], with some discussion that you might find interesting.

Cheers!