Firmware

[electric_vehicle_lover]

I used a board we bought from Microworks to this project, to try control a generic EUC I bought some time ago. That generic have lower speed and lower torque that was dangerous to ride because I could fall on small holes.

Now, the Microworks board, first it seems to be to much quick even dangerous to accelerate but also the motor is rotating backwards! Anyone have an idea how to put it rotating on the right direction? -- please see the video:

It's so important to have our own OpenSource firmware...

Please note that I am being using also this motor for the firmware development and it works as expected and the same as the other motor I also use.

 

[vezu4iy]

3 hours ago, electric_vehicle_lover said:

I used a board we bought from Microworks to this project, to try control a generic EUC I bought some time ago. That generic have lower speed and lower torque that was dangerous to ride because I could fall on small holes.

Now, the Microworks board, first it seems to be to much quick even dangerous to accelerate but also the motor is rotating backwards! Anyone have an idea how to put it rotating on the right direction? -- please see the video:

It's so important to have our own OpenSource firmware...

Please note that I am being using also this motor for the firmware development and it works as expected and the same as the other motor I also use.

 

Calibration?

[Tomek]

7 hours ago, electric_vehicle_lover said:

Now, the Microworks board, first it seems to be to much quick even dangerous to accelerate but also the motor is rotating backwards! Anyone have an idea how to put it rotating on the right direction?

hmm, either put the board on the other side of the wheel (well, cable's probably too short), find a way to put the board upside-down (flip on the vertical axis - components towards the shell), ask microworks if there's a way to recalibrate directions, or swap two corresponding to each other pairs of motor and hall sensor cables?

[electric_vehicle_lover]

I made the calibration process a few times and it works well. Also I found that seems to be possible in this firmware, to calibrate the board horizontal.

Now I exchanged the motor phase wires in all options but didn't worked - made me remember my tests during development to find the right sequence of hall sensors. On this board, if the motor ask to much current the board will detect and beep. Other times, the current is so high that the BMS cut so quickly that the boards don't even beep.

I didn't exchanged yet the hall sensors wires as it take to much time.

On the OpenSource firmware, for inverting the motor direction is needed to exchange the sequence of hall sensors... I don't know if will be possible only with wires, on this board.

[esaj]

Like @Tomek suggested, I suspect the board is the wrong way around and the wheel is trying to balance in the wrong direction. Did the same mistake with the custom-frame & Firewheel -main board on first try 

[electric_vehicle_lover]

So I rotated the board 180 degres and now the motor runs on the right direction, however, with to much speed that is not ridable and is dangerous. I wonder if that firmware were developed for a motor that outputs much higher frequency of hall sensors signal.

On current OpenSource firmware, I tried to ride a bit with EUC with constant speed and is not possible :-) -- I think I will move again to test the balance code. Last time I were feeding the output of balance controller to the duty-cycle PWM of the motor but now I will instead output to the speed controller that will then output command to duty-cycle PWM.

The idea is to have zero speed when is balanced and increase speed as the angle go higher, for on direction or the other.

Edited April 4, 2016 by electric_vehicle_lover

[Call-Tech]

I don't know if this helps any, but I deal with this same problem on the repairs of two wheel self balancing scooters all the time. The Chinese never use the same color wires from one day to the next and I have to open up the motors and make sure the hall wires and the 3 phase power wires are going to board in correct order. Too fast speed and uncontrollable is almost always order of 3 phase power wires and or hall wires. 

[electric_vehicle_lover]

About the OpenSource firmware, I plan to do:

(ANGLE) ---> (balance controller) ---> (speed controller) ---> (torque controller) ---> (MOTOR PWM)

ANGLE is the angle output from the IMU;
balance controller will output a speed value;
speed controller will output a PWM duty-cycle value (PI controller for now);
torque controller control the duty-cycle PWM value and the motor max current.

[lizardmech]

I found something interesting the other day while working on my controller design. Toshiba make an inexpensive MCU that has a dedicated BLDC FOC controller on chip. It has an open source configuration utility to configure the BLDC hardware, people run it on PCs and android. You would only need to put one on a BLDC board and add the gyro to turn it into an EUC controller. If produced in moderate quantities I think it would cost maybe $50 usd to build them with not too expensive mosfets.

[Almaz]

Hello! Can someone place link where can I buy e-bike controller

based exactly on stm32? I have experience of developing firmware

for stm32 microcontrollers.

[electric_vehicle_lover]

Hi @Almaz

Sorry, but is not e-bike is instead a unicycle.

What would you like to do?

[Almaz]

electric_vehicle_lover, of course I mean a unicycle)

Where did you bought your stm32 based unicycle? Send me please link for web site of seller?

Here is your pictures of stm32f103:

 

[electric_vehicle_lover]

@Almaz

We bought the boards on Microworks: http://microworks.en.alibaba.com/

You can find more information on our wiki here: https://github.com/generic-electric-unicycle/documentation/wiki/MicroWorks-18km-h-controller-board

Please tell me if you need more information for the development tools, etc. Please tell me your user ID on Github so I can add you.

[electric_vehicle_lover]

On 4/4/2016 at 7:56 PM, Call-Tech said:

I don't know if this helps any, but I deal with this same problem on the repairs of two wheel self balancing scooters all the time. The Chinese never use the same color wires from one day to the next and I have to open up the motors and make sure the hall wires and the 3 phase power wires are going to board in correct order. Too fast speed and uncontrollable is almost always order of 3 phase power wires and or hall wires. 

I tried to exchange the order of the hall sensors but I got kind of short circuit on the other options:

GND	Hall sensor 1	Hall sensor 2	Hall sensor 3	VCC	Notes
Black	Blue	Yellow	Green	Red	Original
Black	Blue	Green	Yellow	Red	Kind of short circuit
Black	Yellow	Blue	Green	Red	Kind of short circuit
Black	Yellow	Green	Blue	Red	Kind of short circuit
Black	Green	Yellow	Blue	Red	Kind of short circuit
Black	Green	Blue	Yellow	Red	Kind of short circuit

I think I will forget to use this board...

[electric_vehicle_lover]

About the firmware, now I have more knowledge and I think it should be more simple... and I need to go back. I were looking to OpenSource firwmare with FOC and VESC is currently maintained firmware: http://vedder.se/2015/01/vesc-open-source-esc/

I asked on VESC forum for suggestions to control the Unicycle and I got this help: 

Hey I'm building an self balancing electric skateboard like the OneWheel (http://onewheel.com)
I connected a MPU-6050 to an arduino nano. Then the arduino is calculating from the angle of the board the right amount of current with an PID-controll loop for self balancing.
The arduino is connected with UART to my VESC.
I'm using FOC because it gives me the best startup torque.
I'm currently testing the board and fine tuning.

So from the MPU I get the angle and the tilt rate. For calculation, I used an control like this: (https://frobotme.wordpress.com/2014/04/29/algorithm-for-self-balancing-robot/)

The arduino give the calculation over UART to the VESC (current value from -50 to 50). I used this library for this:https://github.com/erwincoumans/ArduinoServoTxRx/tree/master/VescUart
I'm using current control to control the torque.

http://vedder.se/forums/viewtopic.php?f=8&t=177&p=1008#p1002

Video showing the one wheel skate in action:

 

 

[Tomek]

@electric_vehicle_lover are you considering moving to VESC? its hardware specs seem almost suitable to EUCs, but probably the transistors need upgrading, and a gyro could be added directly onto the board too...

[electric_vehicle_lover]

No VESC as that would be another barrier for the users, as we can't buy easily VESC working boards.

By the way, I Just received a package from Microworks with the boards that have Bluetooth :-)

I really miss for development a serial port and now I will have it :-)

 

[jbwheel]

26 minutes ago, electric_vehicle_lover said:

No VESC as that would be another barrier for the users, as we can't buy easily VESC working boards.

 

I don't get it, there are a lot of VESC manufacturers following the vedder design. Lot of people use it for diy e-skate.

[electric_vehicle_lover]

@jbwheel may be may lack of knowledge. Can you please point me to a online shop where I can buy working boards of VESC?

[jbwheel]

At the moment, it appears to be cheaper from Australia (!), didn't make a complete check http://www.enertionboards.com/electric-skateboard-parts/vesc-motor-controller/

You may want to check on the e-skate forum. (http://www.electric-skateboard.builders, http://diyelectricskateboard.com)

But vedder also say (http://vedder.se/2015/01/vesc-open-source-esc/) : "From enertion I got a space cell once and ollin gives me a percentage for each board he sells. The best soldering and assembly quality I have seen is from ollin boards. He does a very good job with the assembly."

http://www.ollinboardcompany.com/product/vedder-s-speed-controller

Made in Germany : http://esk8.de/  ;  http://www.hellray.de/shop/#!/VESC-V4-11-12-made-in-Germany/p/56097898/category=15255004

 

Edited April 18, 2016 by jbwheel

[jbwheel]

@Tomek and @electric_vehicle_lover : VESC current is insane (50 to 240A) but it seems a bit short for voltage max is 60V, I dont know if it would be ok, especially during braking...

[electric_vehicle_lover]

60v is not enough for generic unicycles... :-(

Anyway, VESC seems to be a very relevant project to follow!!!! Thanks!

[Tomek]

22 minutes ago, electric_vehicle_lover said:

60v is not enough for generic unicycles... :-(

 

35 minutes ago, jbwheel said:

@Tomek and @electric_vehicle_lover : VESC current is insane (50 to 240A) but it seems a bit short for voltage max is 60V, I dont know if it would be ok, especially during braking...

I didn't dive into the topic but my assumption is that switching the mosfets to higher-voltage rated ones and/or adding a good heatsink would make it suitable for 67.5V+ that EUCs are using, but maybe there are other voltage-sensitive components I'm forgetting.

[jbwheel]

56 minutes ago, Tomek said:

I didn't dive into the topic but my assumption is that switching the mosfets to higher-voltage rated ones and/or adding a good heatsink would make it suitable for 67.5V+ that EUCs are using, but maybe there are other voltage-sensitive components I'm forgetting.

Changing IRFS7530 MOEFETs would be easy, for this one DRV8302 MOSFET you would need solder habilities unless you ask directly to the makers... What about lowering the max operating voltage of the motor ?

@electric_vehicle_lover 

Also for the same motor, more volt means more speed, more amp means more torque and VESC allow a very fine control of it... But increasing the wheel size allow to increase speed, just look at the MSuper2 where the wheel diameter is bigger but not the motor diameter.

Edited April 18, 2016 by jbwheel

[Berus]

3 hours ago, electric_vehicle_lover said:

By the way, I Just received a package from Microworks with the boards that have Bluetooth :-)

I really miss for development a serial port and now I will have it :-)

 

But this controller, in theory, is for unicycle 500w motor- 30km/h, not classic generic unicycle .