DD-Wheel with Hoverboard Motor

Hello,
do you think an arduino + this controller
https://www.ebay.com/itm/5V-36V-350...rentrq:e1c3eba21640a99b87966774fff06aa8|iid:1

like in

would work too?

This is the datasheet of the controller
https://www.simplypumps.com/uploads/pump accessory data sheets/data sheet bmc15 controller r1.pdf

ops...could not edit the previous post.
Would add this controller too: https://www.mikroe.com/brushless-4-click

All this controllers are useless in this kind of application. You need torque mode and for that you also need position feedback for the commutation.

I have no idea how strong 3D printed parts are. All I can say is that high speed corners and going over curbs puts the rig (and my hands) under a lot of stress while using 8 Nm.
So far the wood/metal clamps I am using show no sign of fatigue.

Yes, I thought about that too. After considering price/value for my needs I opted for the hoverboard motor. Yes, a bike motor might work too. It depends on your needs and the motor spec. The controller is capable of 8-50V and 30A(peak).

Be aware, that in stall mode and with quick load changes (the controller pushes the wheel to the right, you are pushing into a left turn, then you hit the curbs which lets the controller alternate left/right very quickly) 30A are easily reached. I had to scale down from the possible 12 Nm (stall torque) to about 8 Nm in the above dynamic scenario to prevent activating over current protection. I am not sure how a higher W-rated motor will react in this situation.

Hi guys, sorry for my bad english, i'm completely ignorant in elettronic but have an idea when open mmos because I have a direct drive wheel with a MY1020, a DC motor that i control with an STM32F4 Discovery and two motor driver the BTS7960B. When you press on the button of "Setup" in the mmos software you can select the PWM mode and between the option there is SVPWM(BLDC). That think is the mode to control the BLDC motor. Is useful for this project?

Hi RSP,

thanks for sharing your findings!
Unfortunately the functionality behind this option is untested by the author of MMos. Some people tried running a BLDC with these signals but did not succeed. Although I haven't read of someone who used an oscilloscope to actually see what the signal are.

-Benu

@Benu
I got myself a couple of the hoverboard motors and the hoverboard controller.
I was wodering if you ever ran the "motor profiler" on the hoverboard motors and if so, what parameters was the result?
I might get a X-NUCLEO-IHM08M1 but they seem to be unavailable everywhere at the moment.
First I will make a mount for the motor shaft and then put an optical encoder on the motor.

Yes, I ran the motor profiler a couple of times on the hoverboard motor. Some parameters were quite stable (Rs, Ls, Max.Application Speed) between runs, others had some variance (B-Emf constant, Inertia, Friction).
The parameters Pole Pairs, Nominal Current and Nominal DC-Voltage were provided by me.
I changed Nominal Current to such a high value to get some headroom before over current protection kicks in. This is definitely not a value you should run the motor at.

These are the parameters I am using right now:


Oskar at the oDrive community made a tear down of a hoverboard motor. He measured the parameters with some proper tools. His results were not too far off the parameters above. You find his tear down here.

You can order the IHM08M1 directly from ST. They sell their evaluation boards on their web site. According to their site the board is still Active ("Product is in volume production"). But I have no idea what they charge for shipping. I got mine from Mouser about a year ago.

I found on internet this motor driver http://www.wantmotor.com/product/8015a.html for €45 on Ebay. It's interesting because is capable to 45A for 3 second but I don't know that it have a torque mode. What do you think?

Well, I don't think it will work. It seems to be built for higher RPM.
For a DD-Wheel we need exact position feedback from motor to driver. We are operating at a speed of 0-200 RPM. The driver you linked to has position feedback for hall sensors. For our purpose this is usually not working good enough.

Would really enjoy to try to replicate this project but the IHM08M1 seem to be unavailable everywhere and although ST indicate "Active" the buy option is disabled.
Instead what is available is the IHM07M1, a three phase BLDC driver too but based on the L6230:
https://www.st.com/content/st_com/e...m32-ode-move-actuate-hw/x-nucleo-ihm07m1.html

Do you think it would be a suitable replacement or be adapted to the project somehow?

The IHM07M1 unfortunately has not enough power to drive a hoverboard motor. The driver is rated 2.8A peak current. That is way too low. You would need at least 10x that current.
It is a pity, that distributors have no stock of the IHM08M1 right now. I have no idea if it is just a shortage until the next batch arrives. As you said, ST lists it as still active and in stock.

Yep, indeed a pity: would really enjoy to try this project over the brushed MY1020 I was planning to try.
But that board is really hard to find, even few pieces I found on evay are priced insanely high( >120$).
Any valuable alternative you are aware of? I was even evaluating the ODrive solution but that would be maybe overkill,
But if street price for IHM08M1 are so high...

While awaiting a solution, please can you articulate more the software side of the project?
I see in the previous post both you and danove_b wrote some ad hoc code for the ST.
Thus there is not an ultimate solution for this kind of application?
Could such code be shared?

I have quite succefully used a Leonardo to control it in torque mode with just the hall elements. Unfortunately I have very limited time right now, but I'm convinced that it should work OK with the encoder. I t should also be enough resources left in the Leonardo to include the ffb firmware

That's great news!
What kind of inverter (board) are you using to drive the motor? Is it prebuilt or DIY?

So far I did not find an alternative for the IHM08M1. As far as I can remember, odrive robotics had a lower priced one motor board in their shop. That was around 90 USD. But that seems to be gone from the shop.
You could ask the ST guys why the board isn't available from distributors or their own shop. Maybe they are awaiting a new batch.

The code is not the problem and yes it could be shared. It reads the PWM+DIR output from the Arduino/MMos-Board and translates that to a current value for the motor driver.
As mentioned in my earlier posts you either need the development toolchain, which is free of charge and downloadable from the ST web site or the precompiled custom firmware which I can provide.

It's a DIY board that I made for another project I work on. It's build around the btn8982 half bridge which fits the application perfectly. You can see it in the video in the thread cousin of open sim wheel.

I bought an AMT 112s-v encoder which outputs the ABZ signal as a 'CMOS voltage square wave'. Will this single ended quadrature signal work OK with the hardware BENU has specified above?

You are powering the motor with 12 volts? Is it correct that higher voltage will only give added speed. It is only available current that gives torque?