48v3KW Motor Controller for self-balancing robot

[stridera]

So, I decided to purchase this motor:  http://kellycontroller.com/hub-motor-48v3kw-high-torquedisc-brake-p-158.html

It's expensive, but I couldn't find anyone who could help me spec a good motor so I just bought one that I hope will work.

My question now is:  Should I buy a controller or can I build my own?  (Or is it even wise to build my own.)  I plan on using 2 12v Li-Ion batteries to run the motor, and a PIK MC to ultimately control the motors.  The controller would basically just need to accept a forward/backwards pwm signal.

If I do decide to go with the controller, will any generic 3phase 48V controller work?

Thanks for your help!

[Soeren]

Hi,

So, I decided to purchase this motor:  http://kellycontroller.com/hub-motor-48v3kw-high-torquedisc-brake-p-158.html

Talk about overkill, or did you decide to make a tractor or a dozer? 
3kW is almost 4hp
It will draw up to 62.5A @ 48V
And the weight of 49lbs isn't gonna play nice.

My question now is:  Should I buy a controller or can I build my own?  (Or is it even wise to build my own.)

A BLDC motor controller is not a beginners project (to put it mildly) and you don't wanna be cruising at any speed when you discover your first hiccup, the second or the third

  I plan on using 2 12v Li-Ion batteries to run the motor,

Why??
It's a 48V boat anchor motor and only feeding it 12V will reduce it to under a fourth of its nominal power max., if it runs at all - the weight and it's inertia won't change though, so it's a bit like putting a Lambo motor in a Lada - a lot of wasted potential

and a PIK MC [...]

I guess you're very new to microcontrollers (and hopefully quite unaware of what PIK means in Danish ) in case you're referring to a PIC controller (from Microchip ).

If I do decide to go with the controller, will any generic 3phase 48V controller work?

Yes, if it's a 3kW controller and you feed it 48V.

If you have the option to return the motor for a full refund, do so.
Then have the patience to read up on what you need - your project is absolutely not a beginners project, which means that if you just dive in because of impatience, you're destined to fail - even a trained engineer will plan a lot before attempting such a project (and some of them will fail ever the same, but will hopefully have the experience and knowledge to rise again).
So read, read and read 'till you puke and make good guestimates of total loaded weight, needed acceleration to balance etc. - then you may have a fighting chance

[stridera]

Talk about overkill, or did you decide to make a tractor or a dozer? 

Yeah, it's for a self-balancing robot that I'm planning on being able to stabalize 100-300lbs (100 normal, +200 if I want to ride it. 

A BLDC motor controller is not a beginners project (to put it mildly) and you don't wanna be cruising at any speed when you discover your first hiccup, the second or the third

I agree.  I don't plan on having this thing cruising, I just wanted high enough torque to handle the reverse pendulum balancing.  But I'm most likely going to get a controller anyway.  I just don't need anything fancy.

  I plan on using 2 12v Li-Ion batteries to run the motor,

Why??
It's a 48V boat anchor motor and only feeding it 12V will reduce it to under a fourth of its nominal power max., if it runs at all - the weight and it's inertia won't change though, so it's a bit like putting a Lambo motor in a Lada - a lot of wasted potential

Total mistype.  I meant 2x24v batteries.  (Thinking of these actually.)  I might actually get 4 of them to keep the balance right and double the time they can keep it running.  So, I'm planning on hitting the 48v. 

and a PIK MC [...]

I guess you're very new to microcontrollers (and hopefully quite unaware of what PIK means in Danish ) in case you're referring to a PIC controller (from Microchip ).

I'm new to PICs, yes.  I've been using Arduino for a while, but I have a friend who works for PIC (which makes it especially embarrassing that I misspelled it) and he can give me some nice chips to play around with.  I'm a professional programmer (I work for Microsoft working on Office atm) so I'm more fluent with the code than I am with the hardware.  This project, ambitious as it is, is to teach me some of joys of mechanical and hardware/circuit design.  And, if I finish this, it'll really help my e-pik.    (Yes, I just looked that up.)

If I do decide to go with the controller, will any generic 3phase 48V controller work?

Yes, if it's a 3kW controller and you feed it 48V.

Thanks.

If you have the option to return the motor for a full refund, do so.
Then have the patience to read up on what you need - your project is absolutely not a beginners project, which means that if you just dive in because of impatience, you're destined to fail - even a trained engineer will plan a lot before attempting such a project (and some of them will fail ever the same, but will hopefully have the experience and knowledge to rise again).
So read, read and read 'till you puke and make good guestimates of total loaded weight, needed acceleration to balance etc. - then you may have a fighting chance

I've been reading everything I could find.  I posted on here and over at reddit.  I've been asking left and right, and it looked like I would need the high torque motor that I choose to get the balance I needed.  (Reddit has tried a lot to help me, but there seems to be a lack of expertise in the field and not enough literature on the network for a project like mine.) 

Money isn't much of an issue for me, so I figure I would try it with this and then get a newer motor if I need to.  (Unless, of course, I am given some real answers w/o being told to go to school for a mecheng degree.

[Soeren]

Hi,

I just wanted high enough torque to handle the reverse pendulum balancing.

Torque is needed of course, but so is "speed of reaction" and I'm afraid this motor will have too much inertia to make the micro-back-and-forth adjustments without lag and overshoot, so I think it will be oscillating a bit, but hey, a free massage thrown in

[...] I'm more fluent with the code than I am with the hardware.  This project, ambitious as it is, is to teach me some of joys of mechanical and hardware/circuit design.

As long as you don't have any dead line at all, I'm sure you can larn a lot from this, but imagine someone  who never programmed anything told you, that he'd like to make, say, a complete database manager from the ground up - I think it's a rough equivlent.

And, if I finish this, it'll really help my e-pik.    (Yes, I just looked that up.)

Taking "E" will turn the latter flaccid I'm told

[...] there seems to be a lack of expertise in the field and not enough literature on the network for a project like mine.) 

Are you kidding? The net is brimfull of research papers and balancing "scooters" that were all the rage some years ago have turned out numeous projects - the balancing part would be the same for your project, only the numbers will differ.
Most balancing is done with PID control, but given your background, you might wanna look into Fuzzy Logic as well - it may be time consuming to calculate the parameters, but it works with linguistic variables, rather than numbers which is harder for us humans to relate to (there's even a piece of software to calculate or recalculate the parameters, but unfortunately it's pricey - pen, paper and time will do just as well). There's a good beginners tutorial here

Money isn't much of an issue for me, so I figure I would try it with this and then get a newer motor if I need to.  (Unless, of course, I am given some real answers w/o being told to go to school for a mecheng degree.

I think you misinterpret the answers you get. Nobody can tell you what speed/torque you need without knowing all the details and that is one of the reasons I tell you to read a lot, as you need some foundation, to be able to ask the right questions.

BTW. You might wanna consider making a small scale model of the scooter, to test some of the ideas you get along the way -  it makes it a lot easier and faster and takes up less space, so it can be tested on your work table

[Gyarados]

Hello

You might try putting the batteries for your robot bellow the axle (eg: bellow the motors)
That way gravity is assisting by working as a counter-balance, reducing oscillation.
You should always work with gravity not against it .


Gyarados

[newInRobotics]

Low center of gravity is good for static balancing, it makes object stable, for dynamic balancing the higher COG is - the better. If weight below axle is bigger than weight above it, robot won't be able to move as it's not inverted pendulum anymore (it's normal pendulum with COG at it's lowest point). Simple proof, try balancing a plain stick placed perpendicular to to Your hand plane, then try doing same thing however with significant weight being added to the higher end of the stick. Which way is it easier to keep stick in balance?

[Soeren]

Hi,

Simple proof, try balancing a plain stick [...]

Try with a hammer - fast way to get the idea

[newInRobotics]

Hi,

Simple proof, try balancing a plain stick [...]

Try with a hammer - fast way to get the idea

Hammer is a stick with weight on one end :p

[stridera]

Doesn't really matter, since it's a hub motor attached to the wheel.  All the weight will be above it no matter what I choose.

[Soeren]

Hi,

Hi,

Simple proof, try balancing a plain stick [...]

Try with a hammer - fast way to get the idea

Hammer is a stick with weight on one end :p

Exactly and the weight comes already attached, ready to do balance tests (rather than having to fiddle with a stick, a weight and duct tape )

[Soeren]

Hi,

Doesn't really matter, since it's a hub motor attached to the wheel.  All the weight will be above it no matter what I choose.

It sure does matter.
Try with the hammer. The motor (your hand) will be under the weight whether the head is up or down, but it will be very hard to balance it with the head down.

Don't discard such hands-on as useless, it gives you a very tangible feel for the dynamics involved

[stridera]

I meant that the weight will always be above.  I agree it's easier to balance that way, but I can't put my motor above the weight unless my bot is upside down. 

[Soeren]

Hi,

I meant that the weight will always be above.  I agree it's easier to balance that way, but I can't put my motor above the weight unless my bot is upside down. 

I thought it was the placement of the batteries we were talking about?!
The higher the better.

[stridera]

Ahh, gotcha.  yeah, I agree with that.  They'll be up as high as possible.