Using RC motors on E-bikes [Archive]

[recumpence]

http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=5594

Here is the thread John suggested.

Matt

 

[John in CR]

...I have about 16 years RC under my belt now, all electric. ...

John,
Here's a question for you and Matt. Given the simplicity of these RC motors, is it reasonable to assume that some are already available to the public at quite reasonable prices that could be used on a daily commuter bike and if we limit the output to say 50% of their maximum potential, that we can get the kind of dependability that we should all expect from an electric motor? Hub motors are a decent approach because of their simplicity, but their weight being in a lightweight spoked wheel and their wholly inadequate connection to the bike create real and often dangerous problems. If you keep an electric motor adequately ventilated, clean, and dry; and you don't overpower them or their controller, then it's reasonable to expect a service life of a decade or more. Can this ever be an expectation from these small yet powerful motors, or are the stresses simply too great? The first thing the experienced car EV guys ask to help determine whether a motor is suitable for EV use is "how much does it weigh and what are its dimensions?". Performance is great and I love it as much as the next guy, but at what cost? More than anything, I just want something that will always work.
JohninCR

 

[recumpence]

In an word......... YES

That comes not from my bike experience, but primarily from my AXI outrunner that began in a twin rotor helicopter, then went into a scratch built RC monster truck, now it runs my CNC. That motor has hundreds of hours of running time on it at about 40% to 50% of its rated output and it still runs like new. Also, I have controllers from years ago that still work fine.

Unfortunately, I have sort of tainted many people's view of RC motors by trying to push the very edge of physics on what I can pull from 13 pounds of E-gear. Try not to use my bike as a typical example. I am pushing my system very hard. Set up a bike at 2KW or lower, gear it for 30mph and enjoy a long service life.

For me, I am trying to push the edge. I am bound to fry components.

Matt

 

[John in CR]

Thanks Matt,

I forgot you had lots of hours running your AXI in other uses before the bike. It does bring up a question that I tried to ask before, but probably improperly. If you look at the graph that started this thread, the area under the efficiency curve is almost a right triangle, with only good efficiency near maximum output. What do we need to do to run one a good efficiency, but only at say half of its max rated output?

John

 

[recumpence]

Proper gearing would be my best advice. Also, a good current limiter is a must for these motors for a number of reasons;

#1 To protect the driveline.

#2 To protect the controller.

#3 Most controllers are going to try to maintain a consistant motor RPM (to a degree) as the bike goes up hill. A current limiter will wliminate that tendency.

#4 To achieve best efficiency.

Matt

 

[johnrobholmes]

Give me a ring, we can talk. It is pretty easy to use a tester though. My first setup was using a mamba max and servo tester, you just have to know where neutral is. You just want to disable the brake and reverse on the controller.

 

[recumpence]

Yup, I kind of figured that. But I wasn't too sure.

I have already programmed the ESC with my Castle Link. Also, I have run a few different controllers with a servo tester. However, every one of them were airplane controllers. A car or boat controller looks for different parameters.

However, I have run an HV110 wth a pistol grip radio without a problem.

Oh, I also wanted to talk to you about Neu 22 series motors.

I will call you after the family and I get back from church. :wink:

Matt

 

[johnrobholmes]

Pretty chart that I misinterpreted on first look:

I am not big on the current limiting idea. My counterthoughts:

1. Throttle ramping can protects a gearbox. It will limit amp draw by design without actually putting a cap on amp draw.

2. The controller should be able to handle the motor in a stall. The motor should always be the weakest link. (IMO)

3. Motor's will try to hold a constant rpm since they will get an average applied voltage at a given throttle setting (fixed duty cycle). As the load increases a PM motor will automatically ask for more amperage to keep spinning at the same speed. It will bog down some under load.


I am not totally against the idea, but feel that there are better ways to execute the final goals that are wanted. One hill climbing gear would be a great way to limit current on a climb.

 

[safe]

I am not big on the current limiting idea.

On a motor that seeks to have higher power I have to agree. Current limiting really only works best when you have gears (and I know you are tired of hearing about it) and so you should focus on building an initial geardown system that is strong enough to handle the full torque that is produced. Current limiting produces a linear power curve... great for people trying to figure out the perfect time to shift gears, but lousy for fixed geared people.

Low end torque is the "pseudo gearing" of the unlimited electric motor... you want to keep that if you go single speed...

 

[johnrobholmes]

Having gears limits amp draw by design, less torque load on the motor when accelerating through the bottom half of the RPM range. I would be even less inclined to want amp limiting with gears so I can bang through them as fast as possible! :lol:

 

[dirty_d]

johnrobholmes, nope youre just looking at the graph wrong, the left side of the graph is no-load at 1200rpm and the right side fo the graph starts at 950rpm, thats just a very small portion of the motors performance. the part where the efficiency rapidly drops is the real no-load area, not just before it where efficiency peaks. all permanent magnet motors have that same curve. for this curve the efficinecy peak is at 35A.

\youre looking at the torque(0-160), the current only goes up to 59A on this portion of the graph.


heres a proper graph of that motor...

 

[safe]

Having gears limits amp draw by design, less torque load on the motor when accelerating through the bottom half of the RPM range. I would be even less inclined to want amp limiting with gears so I can bang through them as fast as possible! :lol:

But the actual horsepower is best up top, even though it feels like it's better down low... that's the whole "EV Grin" thing messing with your head, the fact that torque peaks early while horsepower peaks late tricks your mind into using the wrong gear. I've talked about this endlessly elsewhere, so we're sort of retracing areas of knowledge already covered. Armature Current Limiting simply "guides" the rider into a better gear selection because it focuses the rider experience towards where the real power is. The best power and the "EV Grin" are two separate things. The best power is higher in the powerband.

 

[safe]

...here's a proper graph of that motor...

I know you understand Dirty_D, but it should be said for others that the actual powerband shape is effected by the controller limit and it's limiting techniques. You really need to look into the specific controller induced powerbands to get into a deeper discussion about motor behavior. That chart is valid, but no one would use an unlimited controller because the motor would overheat too quickly. Anyway... another long discussion to avoid I guess... suffice to say the RC motor deals with all the same powerband issues that any other permanent magnet motor does.

 

[johnrobholmes]

....but no one would use an unlimited controller because the motor would overheat too quickly.....

Say what? A properly geared/ volted motor will not overheat if it is right for the job- period. I want the motor to be the weak link, not the ESC. The chart he posted is using a motor controller for the data, you can't run a BLDC motor without one.

Maximum power (horsepower, wattage, mice ear/inch, whatever) is generated at 1/2 of the no load speed, it does not peak late. Maximum efficiency is generated at 1/10th no load speed, but you will only be able to squeeze maybe 1/3 of the total available power from the motor at that speed. The "EV grin" occurs where power is greatest indeed, no trickery.


Thanks for the chart Dirty_D, I should have caught on that my graph didn't go to stall amperage. My brain is safe, the worldview of PM motor properties did not in fact change.

 

[safe]

Maximum power (horsepower, wattage, mice ear/inch, whatever) is generated at 1/2 of the no load speed, it does not peak late. Maximum efficiency is generated at 1/10th no load speed, but you will only be able to squeeze maybe 1/3 of the total available power from the motor at that speed. The "EV grin" occurs where power is greatest indeed, no trickery.

The "trick" is to use a motor that is significantly larger than needed and run it at near it's maximum efficiency. That's the whole point of all this efficiency experimentation. If you try to actually use a motor in it's middle rpm range like that your efficiency will be terrible. This is a chart that shows the "real world" of how a regular limited controller works on a typical "real world" bike controller verses how something like armature current limiting would work. The "EV Grin" on the bikes we really ride is caused by the inflection point of acceleration... once that early "surge" fades your mind responds by thinking the power is no longer rising, which is false. The reasoning for using the current limiting is that it focuses the mind on the peak power FOR A LIMITED CONTROLLER so that you tend to shift at the more advantageous times. Bringing the armature current limiting into it might be making it more complex, but this is a really good chart once you know what is really taking place.

...you might be able to teach me something here though, I was under the impression that RC controllers also did standard battery side current limiting... is this NOT correct? If they do current limiting then it will behave like in the chart I presented. If they do not do current limiting then it will be like Dirty_D's chart. (the "raw motor" chart)

You need to be able to instantly recognize charts for:

"Raw" Unlimited Motors

Standard "Battery Side" Current Limited Motors

Armature (Motor) Current Limited Motors

...because they are the three most common powerbands that are going to be used and thought about. (after you get used to staring at these charts for a long time they sort of sink into your head and you can spot them easily)

However, just to make things more complicated there are the programmable controllers and their "Ramp Rates" which make things even more complex. :?

 

[johnrobholmes]

.. once that early "surge" fades your mind responds by thinking the power is no longer rising, which is false....

.... I was under the impression that RC controllers also did standard battery side current limiting... is this NOT correct? If they do current limiting then it will behave like in the chart I presented. If they do not do current limiting then it will be like Dirty_D's chart. (the "raw motor" chart)

When the surge fades the power is indeed falling, while efficiency is rising. Maybe you just have your terms mixed up here. I understand that you are just trying to get peak efficiency, and choosing a larger motor is one good way to do it. Another good way is to gear the motor so that top speed on flat ground is close to no load. Another good way is multiple gears OR slower acceleration (which is what your amp limiting accomplishes).


Some RC controllers (Castle) have crude amp limiting, but it is not very precise. The amp limiters work most during startup to keep cheap batteries happy. Cheap controllers do not have this. A more effective amp limiter is the throttle ramp control, which is just a throttle damper. It prevents too much acceleration by simply slowing down the throttle increases. It would not limit amp draw on hills however, limiting amp draw on a hill would just slow the vehicle down. What I don't get on your chart is how a motor can have more power with current limiting. The battery voltage doesn't change. To limit amp draw the average applied voltage must be lower, so your peak power will be lower as well unless you are raising your voltage to compensate.

 

[safe]

What I don't get on your chart is how a motor can have more power with current limiting.

Apparently RC motors operate in a totally different world than regular ebike motors and controllers. For pretty much all the ebikes they use standard battery side limiting which will set a hard limit of something like 40 amps so that the battery doesn't get damaged from excessive discharge and the motor doesn't overheat. If what you say is true and all the RC motors do is some crude "damage control" style limiting then we are talking about toally different animals. On ebikes the armature current limiting can be used with a higher baseline current limit (maybe 50 amps rather than 40 amps)... that's how it gets more power because it's being allowed to peak higher, but since it compensates with less low end torque you don't get the heating problems normally associated with a higher peak current limit.

NOW I understand why Recumpence needed to build a slipper clutch... these RC motors don't really have current limits on them so when you get an "EV Grin" it's something directly tied to the "Raw" motor powerband.

It sounds to me (as some other thread had already pointed out) that what is needed is a real true-to-life controller that can be configured with real world current limits and other luxury items that the ebike world is used to... in a sense the RC motors are sophisticated in some ways (the motor) and crude and primitive in others. (the controller)

 

[johnrobholmes]

I see now your graph is comparing ACL to BCL. Now it makes sense.

RC motor controllers mostly don't limit current. It would be absolutely silly to do so, since we want the highest power to weight ratio possible. As I said, Castle controllers (ground based) have what is called "punch control" that limits the initial burst so that cheaper batteries won't cause problems. Basically it is still a throttle damper, but on a smaller and faster scale.

 

[safe]

RC motor controllers mostly don't limit current. It would be absolutely silly to do so, since we want the highest power to weight ratio possible.

But wait... ebikes are under load most of the time if you are like me and shooting for racing type speeds. RC motors are designed to be loaded for an instant and then mostly run at their designed rpm the rest of the time. In the "real world" an ebike is never going the same speed for more than a second. (at least I am almost always speeding up or slowing down with the only exception being my top speed where I'm just holding it as long as I can)

Sacrificing 50% of your energy to wasted heat (which is what happens with an unlimited motor) is insane for an ebike where it's all about efficiency.

If you took a motor running at near it's efficiency peak (and used gears) and raced it over something like a 10 mile course against a motor that tried to do everything with "brute force" low efficiency power and both bikes were allowed the same battery it doesn't take much to realize the more efficient machine will win because every bit of energy is used to produce forward motion.

All that ineffiency means that you are throwing energy away in the form of heat... your 0-100 meter times might be great, but you would not win a long race.

 

[recumpence]

Good discussion.

RC motors are (from a simplistic perspective) unlimited in their power consumption and power output (this is not absolutly true, but I say it to make a point). So, with my bike as an example, my tiny motor will pull EXTREMELY hard and just keep asking for more. I have seen over 6,000 watts pulled on accelleration. That is hard on all components. That is why I want current limiting.

Now, I have also seen how a slight gearing increase can drastically increase load and failure of components.

When I approach a hill at typically 25mph in my neighborhood, I leave my throttle stationary. As the bike climbs the hill, the speed only slows a tiny bit (typically 3 MPH or so up a very steep hill). The approach watts are typically around 400 or 450. As the bike climbs the hill, the ESC trys to sustain speed. As it does the wattage goes through the roof! I have seen 2,400 watts as the bike merely sustains speed up that hill. Having 6,000 watts on tap means I can accellerate (pretty hard) up that same hill. But, obviously, that is super hard on equipment.

Now, if I am pushing for maximum efficiency, I will approach that same hill at maybe 30mph instead and gradually lower my throttle to sustain a relatively reasonable wattage as the bike climbs. Then as I crest the hill, I allow it to accellerate again. This maximizes efficiency. That, in a nutshell, is the prime reason I would love a current limiter. It would also eliminate the need for my clutch.

Matt

 

[johnrobholmes]

RC motors are designed to be loaded for an instant and then mostly run at their designed rpm the rest of the time.

Safe,

I certainly agree with your post. Efficiency is really what we should shoot for. As for the motor manufacturers, they test for use that would be similar to planes. This is where we come in and have to push the motors and find out what works and what doesn't. Manufacturers are too busy making me a motor right now, and it will be our job to sort out the quality and "size" of a motor. Payload, gearing, terrain, and throttle hand all change the equation. Overshooting the needs is always the safe way.


Matt,

I now understand the good use for the ACL. Don't know why I had not of thought of that. Possibly because I haven't ridden your bike? I do the same thing on hills with my throttle too, but my drivetrain isn't under as much stress to NEED ACL

What would be useful is to have an ACL button. I want to be able to scooooot if I need too, but approaching hills could use an autopilot for the climb limiter. Almost a built in stall you could say.

 

[recumpence]

Exactly.

What I want is a pot that can adjust my current from a low of maybe 10 amps to an infinate maximum (effectively bypassing the limiter).

That is what I am after.

Matt

 

[johnrobholmes]

The design on this "limiting box" would need to be very adjustable in our most useful range, which I think would be from 10 to 100 amps or so. A linear range in the useful area, with a quick jump off the deep end into nothing.


This whole last page should really be in the ACL thread, but it pertains to an "issue" that R/C controllers have where the average applied voltage doesn't change under load.

 

[Pete]

I have not gone for a dig back through the thread, apologies if this has been covered already. I thought the Cycle Analyist 2.1? beta code has variable current limiting that can be adjusted on the fly.

 

[Miles]

I think I would favour 2 adjustable ranges with a push button to access the higher amp limit.

Yes, it looks like we could use the CA, Pete: http://endless-sphere.com/forums/viewtopic.php?f=4&t=3466