[no_reluctance]

Hi everyone, I've been lurking here for the better part of 2 or 3 years and want to integrate into the community a little bit more. I'm not as much of an expert as I'd like to be but I've got project ideas that could use a few tips.

I like brush-less motors and am reasonably familiar with 3 phase control of them but I find a lot of the really important design critical motor feed-back type stuff is often left as an exercise to the reader. For instance, how to tell if a motor is trapezoid or sinusoid type in its design scheme. Its frustrating to say the least.

I also understand there is a fairly wide price gap between a high performance AC motor (used here to mean brush-less DC as well) and the second or third hand industrial motors and small hobbyist variants, for reasons that seem rather trivial.

I wanted a to select a motor that had at least stator cooling (shaft cooling would be awesome, but I don't know if there is any way to acquire those outside of parting out a Tesla, (which I read is permanently affixed to its reduction gear/differential) Which is looking to be both too large, and unsuitable for my application.


I wanted to try my hand at a differential steer/differential drive. using 4 independent wheels and...for lack of there being a good/cheap/available motor. I was wondering if anyone had any thoughts/cautionary tales about using hobbyist RC motors stacked into planetary arrangements to drive a lighter weight vehicle directly. something like a hub motor arrangement (though my project will be a tad different mechanically)


I know of 2 issues with this. firstly, its wasteful volumetrically to have 5 or six motors where 1 large motor would have worked. this is irreducible but I figured it may offset some of its shortcoming because the individual motors are pretty high performance for their weight and have good balancing and performance properties. The second issue is that RC motors of the power you'd need to power a vehicle, even working 6 or 8 in tandem per wheel, need to be beefy indeed. 5 or 6kw continuous at a minimum, if you want a vehicle of even modest performance.

there is a third issue about vehicle dynamics at low speed causing excessive wheel hop. I think the only way to reckon the correct wheel geometry is through experimentation.

Here's the stuff I would like some advice on: Never attempted regenerative breaking with a bldc motor and controller of the hobbyist type...doubt many have either. I also know many prefer these motors to be very (overly) simple, lacking hall sensors and encoders etc... and the controllers themselves may prove unsuitable. Anyone have any thoughts into this sort of arrangement?

 

[Sonikaccord]

Hi everyone, I've been lurking here for the better part of 2 or 3 years and want to integrate into the community a little bit more. I'm not as much of an expert as I'd like to be but I've got project ideas that could use a few tips.

I like brush-less motors and am reasonably familiar with 3 phase control of them but I find a lot of the really important design critical motor feed-back type stuff is often left as an exercise to the reader. For instance, how to tell if a motor is trapezoid or sinusoid type in its design scheme. Its frustrating to say the least.

You can't tell from the outside of the motor. You have to know the motor's winding pitch and magnet pitch. Typically, full-pitch windings have a trapezoidal back EMF, therefore have a square drive need. Others are sinusoidal-ish.


I also understand there is a fairly wide price gap between a high performance AC motor (used here to mean brush-less DC as well) and the second or third hand industrial motors and small hobbyist variants, for reasons that seem rather trivial.

I wanted a to select a motor that had at least stator cooling (shaft cooling would be awesome, but I don't know if there is any way to acquire those outside of parting out a Tesla, (which I read is permanently affixed to its reduction gear/differential) Which is looking to be both too large, and unsuitable for my application.


I wanted to try my hand at a differential steer/differential drive. using 4 independent wheels and...for lack of there being a good/cheap/available motor. I was wondering if anyone had any thoughts/cautionary tales about using hobbyist RC motors stacked into planetary arrangements to drive a lighter weight vehicle directly. something like a hub motor arrangement (though my project will be a tad different mechanically)

Stacked motors are fine. Theoretically and ideally it's the same as increasing the axial length of the motor. So for every motor you add, you can multiply your torque(and current needs) by that amount.

I know of 2 issues with this. firstly, its wasteful volumetrically to have 5 or six motors where 1 large motor would have worked. this is irreducible but I figured it may offset some of its shortcoming because the individual motors are pretty high performance for their weight and have good balancing and performance properties. The second issue is that RC motors of the power you'd need to power a vehicle, even working 6 or 8 in tandem per wheel, need to be beefy indeed. 5 or 6kw continuous at a minimum, if you want a vehicle of even modest performance.

Not necessarily wasteful. There are many drag racers that use multiple smaller motors. It really depends on your end goal.

there is a third issue about vehicle dynamics at low speed causing excessive wheel hop. I think the only way to reckon the correct wheel geometry is through experimentation.

Wheel hop can be reduced through torque management algorithms, and proper bracing and suspension design/upgrades.

Here's the stuff I would like some advice on: Never attempted regenerative breaking with a bldc motor and controller of the hobbyist type...doubt many have either. I also know many prefer these motors to be very (overly) simple, lacking hall sensors and encoders etc... and the controllers themselves may prove unsuitable. Anyone have any thoughts into this sort of arrangement?

To do regen, you'll need a controller capable of regen. That's the simplest way I can put it. I believe the "Any AC Motor" project handles dual motors and regen. All you would need is a power stage and the "brains" from that project.

For your application, you'll need some type of rotor feedback whether sensored or sensorless.

Check out this thread:

http://www.diyelectriccar.com/forums/showthread.php/any-ac-motor-any-inverter-umc-152234.html

 

[no_reluctance]

Thanks,


I am in agreement, I have subscribed to that thread and PM'd eldis for availability. What a great project and a nice guy for being so open with it all and not becoming cagey or anti competitive, and again for wanting to advance the state of the DIY community without only seeing dollar signs as most with the talent and ability do.


There's a similar but not the same project here where I think I wanted to emulate to get a good start. I believe I've got the math formulas worked out for what speeds to drive the motors for an arbitrary curve (turn)

http://hackaday.com/2014/07/12/independent-wheel-drive-rc-car/


I am still struggling to find a good source for sensored RC motors. they don't seem to be very common, likewise ESC's with outputs instead of just inputs seem also less common. anyone have any recommendations on this? Even if it isn't RC sized, I'd probably be interested.



Oh, also, do you have any pictures of winding pitch etc? I'd love to chew on that for a while. And with regard to stacking motors, apart from having a common shaft, is there a way (thinking an adapter) to affix ouput shafts through motors without pressing the shaft and putting two rotors on a single shaft? or can you just buy them this way? I was looking at EV drive guys for their hardware in what might be the full build...only thing keeping me back was price. they seem to have a good offering in the US. (though I'd love to hear what the negative reviews have to say).

as for wheel hop and wastefulness...the wheel hop is because I'm going to be trying to steer this thing like a tank. (though with some nuance) and the tires themselves will experience off axis drag (like pushing the car sideways with the wheels spinning or doing donuts etc). Suspension dynamics here I don't think will help me unfortunately, just better wheel geometry since I want to see if this will work without conventional ackerman steering.

For why it would be wasteful, Its just that I'm not sure I can afford a full size motor, only several smaller ones in the same volume..but that is by definition a waste of volume optimization.


Sorry if this post is kinda incoherent, I don't want to have a huge nested quote that grows and grows...etc.

 

[Sonikaccord]

I'm not too versed in RC land. Really, I only learned this motor stuff a few months ago. I picked up a few books off the web and just started reading.

http://www.emetor.org/glossary/
This website is pretty good, it has nice explanations and pictures as well.

Ok, I see how you are maneuvering now. Are you using a wishbone or Mac strut system?

If the power density is the same, then it would be wasteful, I agree. But if it is the same volume and you can increase the output power per volume with the smaller motors, then you have a win. Plus you get built-in redundancy.

As far as adapters, I'm not sure of any inexpensive way to go about that. Maybe machined couplers.

 

[no_reluctance]

https://www.youtube.com/watch?v=ET1HEyqEQJQ


This project is very similar to what I wanted to do. I'll probably emulate quite a bit of it to start but I have different goals in mind than the creator of the linked project. He seems to want maximum power for a four wheel drive setup. I'm more interested in differential steering.


Does anyone have any experience in stuffing hall effect sensors onto a motor for rpm feedback? I understand you need a bit of logic in there because its just counting pole crossings but does anyone know any cautionary tails about analog inputs into the micro-controller and or axial orientation to get the best pickup?

 

[Sunking]

I am a RC pilot junky and I fly a lot of electrics. i have been down the same path of thinking about using RC motors. So take the advice of someone who has gone before you and already learned the expensive lesson you get for free.

Don't RC motor because they are not designed to do what you want. Besides being made from the lowest quality materials the big issues are:

1. RC motors are made to run full out for only very short periods of time of 5 to 8 minutes. And even a hard run of 5 minutes can burn the motor up.

2. They are very low torque and extremely high RPM.

3. Will burn up in just a matter of a few seconds when stalled or under a heavy loads at low RPM like accelerating from a standing start.

4. Operate at way to low of a voltage to get any serious power. RC motors are designed to operate at voltages of low as 3.7 volts for micro aircraft, up to around 35 volts for large helicopter motors.

Hope that helps and prevents you from making a expensive mistake.

 

[Karter2]

Lots of experience over on endless sphere for running RC motors with hall sensors, and at high (100v) voltage.
.. But as Sunking said, not recommended for heavy vehicles or extended use.
There are plenty of brushless motors in the 5-10kW range and even hub motors that could be found with those outputs, so I don't know why you would mess with lots of toy motors and failure prone ESCs when you could use those.
PS.... You could try out your ideas on a small scale RC equipped model before investing the time and resources in a full size project ?

 

[no_reluctance]

Oh yeah, thanks guys, its great stuff and...for a few reasons, I'm hoping to be talked out of the RC approach.. it seems wasteful for a given volume to have 6 undersized motors vs 1 "vehicle load" sized motor...needless waste of windings, shafts...gearing complexity...etc.

To the point that RC motors are cheap/cheaply made. I would ask though: compared to what? a DC starter motor in a car is about as cheap as it gets to make an electric motor, yet they have enough torque to motor the entire engine on the coldest day of the year. an RC motor (thinking here bldc) while completely different type of motor is much higher quality. many of them source the best winding varnish for high temps and are high speed balanced. even most of the so-called golden motors from China are not high speed balanced. so...while still not up to the task in cars I think there's a bit of a qualifier to say an RC is a cheaply made motor. some are definitely as nice as you can buy for that size. Bar none. AS for the burning, that's just a matter of current in the windings. If you get a stall the drive signal just slips passed the rotor and you get no motion, it doesn't increase heat or current...etc.. not to my understanding. I'd be happy to get put in my place by the pro's though.

I'd like to talk to a mechanical engineer about loading because It is my suspicion that the way the shafts are sized and the bearings etc... I suspect that even the best 2 or 3 cubic inch motor would snap a shaft or crush a roller/ball due to load mismatch.


All that said. I have no love for RC motors that would preclude my going with another size. Is there a smaller say 5-40kw motor out there that has built in hall sensor or encoder...possibly a harness for motor telemetry? (amps, RPM as a digital value vs analog etc..) It'd save me a fair bit of digging.

Also absolutely yes, scale model comes first. There's a critical parameter about wheel size + wheel base and axle geometry which affects steering immensely. I have to solve or at least conclude the limits of torque steer performance there. or at least, find some decent documentation on it. Skid steer is not really an accurate or relevant data-set because it is A) not involving 4 independent wheels and B) usually off road so the friction for everything is all different.

By my work so far. there is a minimum radius and forward/backward speed by which the wheels begin to skid and...but it doesn't have to be perfect if its more capable etc..

Again, Thanks for the input guys I hope to put rubber to the road with this someday. (even if its only just my DIY e-vehicle..rather than the prodigal future-wagon I am contemplating today)

 

[no_reluctance]

I am a RC pilot junky and I fly a lot of electrics. i have been down the same path of thinking about using RC motors. So take the advice of someone who has gone before you and already learned the expensive lesson you get for free.

Don't RC motor because they are not designed to do what you want. Besides being made from the lowest quality materials the big issues are:

1. RC motors are made to run full out for only very short periods of time of 5 to 8 minutes. And even a hard run of 5 minutes can burn the motor up.

2. They are very low torque and extremely high RPM.

3. Will burn up in just a matter of a few seconds when stalled or under a heavy loads at low RPM like accelerating from a standing start.

4. Operate at way to low of a voltage to get any serious power. RC motors are designed to operate at voltages of low as 3.7 volts for micro aircraft, up to around 35 volts for large helicopter motors.

Hope that helps and prevents you from making a expensive mistake.

If you will do me the kindness of hearing how I'd address these completely legitimate concerns.. ( no sarcasm intended. meant to acknowledge your expertise on the subject.)

1. If you must frequently operate at full load for much higher multiples than 5-8 minutes then you (the designer) has/have poorly sized the output of motor groups to the load. (not a problem for RC's, huge problem for transit and steady state application. If you operate them at normal load (or better with liquid cooling) and normal heating/cooling. I don't see why load is a problem except that obviously 1 motor the size of a coffee cup won't move a vehicle, and even if it could, it would fail when on a long hill. totally granted. mechanically, though, there might be issues transmitting 5 horsepower backwards and forwards through a shaft of only 1/8" and bearings made for a 10kg object.

2. torque vs RPM is a longstanding problem in vehicles. It was solved with gearing. you figure out a good vehicle load envelope and a gear ratio to match the torque to cope with the load, and a reasonable top speed, which depends on a few more qualities. though it limits your top speed to trade rpm for torque, this is still a feature on normal vehicles. Tesla's are geared final drive something like 9-1. most cars are at least 5 to one if you multiply transmission to diff ratio

3. I was not aware BLDC motors consumed more current at a stall, rather the danger is in over-currenting them at any time. also a stall as we would think of it wouldn't happen, you'd just get overshoot or slip. for this reason all electric cars with Regen braking should have some sort of mechanical brake for hills, so you don't have the precarious problem of balancing a true stall condition on a hill for indefinite amounts of time.

4. the voltage and power density (because of the voltage) are absolutely a huge problem. You hit the nail on the head with that one. RC's have it because they can't haul the batteries or fancy electronics for boost converters. huge negative in terms of power to weight and amps needed to be sure.

All told, I probably wont go with RC's for the full concept but maybe the 1/2 size attempt (go-kart size or similar). But I find its interesting to consider none-the-less

 

[Karter2]

As I mentioned, and I suspect you already know,..there are many choices of "BLDC" motors in the 5 - 50 kW range that would fit your application.
So why would you still even consider the excessive complication required to employ multiple small RC motors to do the job of one single motor ?
If you are thinking cost advantage, I doubt that would be the result overall.
Apart from the additional drive and mounting hardware, multiple controllers, etc ....you would have to chose a "Quality" RC motor (Astro, Graupner..etc) , AND significantly de rate the output for traction drive loading to ensure reliability.
If cost is a major issue, why not use brushed motor/ controller systems for development until you are ready to "productionise" the final design.

 

[ken will]

So why would you still even consider the excessive complication required to employ multiple small RC motors to do the job of one single motor ?

I believe most DIY type people will give the answer:

"Because nobody else has done it!"

 

[piotrsko]

Dunno, there was a guy trying to build an electric B36 peacemaker........

And the kid down the street has an octo quad heavy lifter

 

[no_reluctance]

Price isn't an overwhelming concern. I mean, I probably can't afford to import Brusa or emrax motors, but I'm still kinda leaning towards brush-less DC or AC motors for efficiency, maintenance and cool factor. I was thinking seimens or possibly a solution from a boutique vendor like evdrive. Has anyone used their motors/controllers before that could comment on how "open" the control schemes and motor TM is? I'm not opposed to series DC motors either. but I do wonder how tightly RPM can be controlled. I have a few qualms with DC motors...but the price is hard to beat. I guess I'd just be happy with any motor, RC or not that had a sensor package as part of the motor rather than thinking about how I am going to weatherize and strain relieve an added sensor and cabling set...but so far that looks like my only option. can anyone point me to a good build thread where that was done for closed loop control that didn't involve recycling the OEM motor or controller?