Micro Motor Community

Automated motor testing with a flight controller?


#1

Gday fellow inhabitants of this excellent corner of the internet :wink:

Im using a fairly expensive programmable lab power supply for my motor testing.
Its basically feeding the motor a pre-programmed test pattern, e.g. 60 seconds at 2V, then 60 seconds at 2.5V, then 60 seconds at 3V, then 30 seconds rest. And repeat.

Now I want to run a lot more of such tests in parallel, and dont want to buy more of these power supplies.
Instead, I would like to power up a bench full of flight controllers, each of them running a voltage pattern for 4 motors each.

The challenge is:
How to program a flight controller (MultiWii would be ideal :wink: ), so that it does the following:
After boot, it will start immediately to power all 4 motors based on a pre-configured schedule (similar to the above), without any manual input from a receiver or USB etc.

Now this is a pretty big deal for me, because I need this setup to prove reliability of our motor production to a large potential customer, and I suck at programming.
I know that you guys will jump in anyway to help me, but to increase motivation, Im offering some hardware perks (FCs and motors) for those who are putting their time into this!


#2

Good luck with the potential customer Chief Scientist!
When I read your post I thought of some hardware that I know nothing about. Would an arduino or raspberry pie be able to handle this type of autonomous testing?
We have some of the smartest people in our niche hobby here, so I’m excited for you and will be following. Please keep us posted


#3

I’ve sent you a PM.


#4

No problem programming this on (multiwii) avr code.
Build a lipo discharger from my old multiwii board.

Also have a project named “worc” (WallieOnlineRobotController) based on arduino pro mini and tb6612fng.
Could modify this to do the test you want.
Also implemented the MSP so could give motor status back to multiwii gui if you want.


#5

@sier, thanks for the message, replied!

@wallieonline, would love to see what you got!
Not sure if feedback to MultiWii is useful. I’d love to see current logging, but AFAIK that’s not possible on the board we are using. So far I plan to diagnose the smoke signs :wink:
A very simple test pattern which is played for a definable amount of time would be sufficient.


#6

There will be a difference between Analog Voltage Control like you are using now and Pulse Width Modulation (PWM) from a multiwii board. Hope this isn’t a problem.


#7

I explicitly prefer PWM control, because it provides an even more realistic simulation of the stress that the motors will put through on an actual flight.
I dont think it will make a difference in term of wear, but the more variables we eliminate - the better!


#8

This has Arduino written all over it, one pwm output pin could easily switch a MOSFET that switches dozens and dozens of MOSFETs that each run a motor. At that point they can even ramp up and ramp down instead of hard 2v-2.5v jumps.

If you go the clone route you can probably be up and running testing 10-20 motors at once for $25 (plus power supply cost).
No current sensing though, and you’ll need to use some sort of chunky 4v power supply.

I’m sort of on the other side of the world unfortunately or I’d whack something together and ship it over to you.


#9

I do have enough of the Classic FCs here.
The bottle neck will indeed be the power supply.
My low-volt PS does 10A max.

All I need at this point to move to the next step is a fine sketch :slight_smile:


#10

The WORController is based on a 2$ Arduino pro mini and a tb6612fng (or other fet/hbridge setup) :slight_smile:
Current sensing is possible with the i2c breakout board.

@Benedikt Only have a diy version of a multiwii, not sure if it has the same layout as the boards you use. Wat mcu 328p or 32u4? Which pins are the fets connected?


#11

Is this the board?
Then I have all the info i need.


#12

A dedicated solution with current sensing and logging would be very interesting mid-term.
Ideally using a supply voltage of 12V.

But short-term, running on the FCs I have works best for me.
Its this FC here:


#13

Send you a PM.
Going to test when I get home.