Merging the R/C crowd with EVers

oofnik

100 W
Joined
May 1, 2008
Messages
264
Location
Israel
Hi everyone,
I'm researching to build my first e-bike out of a Mongoose DX 3.3. I see a lot of people using relatively archaic e-bike parts for lots and lots of $$$. I'm a mechanical engineering student with little money and lots of faith in new technology :D and I think we can do better in 2008. So instead of settling on some of the most common options like the Crystalyte controllers that use through-hole MOSFETS and ICs, I have been looking at the other popular hobby application of PM brushless motors: R/C planes, cars, boats, and helis. R/C brushless motor systems are designed to be ultra high power-weight ratio. Motors that are capable of producing 1500w are measured in GRAMS. :shock: Similarly, the controllers used in R/Cs typically all use very lightweight components (all SMD) and sophisticated controller electronics (almost all are BEMF). It seems to me that the combination of a high power R/C electronic speed controller, abbreviated in their world as ESC, plus a high output BLDC motor from their territory would be a giant leap forward from the expensive controller + hub motor setup so common on these forums. I don't mean to insult anybody of course, but I feel like the technology a lot of folks are using to build their e-bikes is aging and needs some radical updates. This thread was partly inspired by the E-cumbent project, by far my favorite e-bike project I've seen. Now of course we don't all have unlimited access to expensive CNC machinery to cut our own mounts and sprockets and stuff but we DO all have access to these relatively cheap and powerful R/C parts! Check out this controller for example. All SMD components, capable of 50V @ 100A, programmable, and it all weighs 125g.

What's keeping more and more EV and e-bikers from using these awesome, readily available R/C parts? Please forgive me as I am very new to the whole hobbyist EV scene, and I hope I'm not overlooking something very obvious to all of you... but I think it's time the R/C folks and the EV folks start talking to each other and share more ideas. We use almost the exact same technologies, just in different shapes and sizes! :wink:
 
oofnik said:
What's keeping more and more EV and e-bikers from using these awesome, readily available R/C parts?
We've been somewhat distracted by good batteries that won't set our butts on fire. If that weren't the case, we'd be beating the bushes harder for lighter and more efficient motors & controllers.

It might be safe to say that more RCers are fabricators and more ebikers are handy with tools and can assemble kits. If some enterprising people created a kit that used RC technology for bikes, convergence would accelerate.

:D
 
yes ok you just saw that ...we did too at the same time you did and we agree it should go lighter in every technology ...when you put out the money you realize it is changing fast enough ...every year i spend more than i imagine i would of... i started with 36 volts sla 4 years ago then the same summer added one 12 to be 48 volts year after 36 volts nimh 450.00...the li-ion was way out of price and way less popular and hard to get ...now this year i spent again on a123 and my nimh are still good ...and if they get a controller similar in size and weigt of the rc i will probably buy it ...

in the end are we really saving on gas ...loll
 
i guess you could use a comparison between a Mack truck and an F1 racecar they both use an engine rated for about the same horsepower. the F1 uses the newest cutting edge technology and is tiny and lightweight. The F1 will run for about 5 hours. The Mack Truck uses fairly modern but not quite so cutting edge technology and it will run at that power for thousands of hours. The F1 will rev to almost 15K rpm, the Mack to only a couple of thousand.

most modern bldc controllers use the same micro chips. on the bike and EV controllers all of the parts possible will be surface mount. they are cheaper both from a materials point and assembly point. if they could get away with tiny smt FETs they would. makes loads of economic cents. and manufacturers are all for a penny saved.....

the only through hole components that i have seen on most modern controllers are the output FETs and some large filter capacitors that are imposible to make as SMT. the reason they use the through hole part is because it has a BIG BOLT HOLE and a LARGE METAL SURFACE to make it as easy as possible to hook up a BIG BAD HEATSINK needed to efficiently dissipate the heat generated by the FETs constantly transitioning from off to on. heat generation is a lot greater on a bike. the RC model engine pushes around a model that weighs a few pounds, not the 200+++ pounds we expext the bike to push. that makes a big difference on the parasitic losses to bearings and other places. the F1 comparison works here as well. the F1 does great at moving 1500lbs or so. but would it even budge the 45,000lbs of a tractor trailer.

by the way this heavy load that the e-bike motors push generates a lot of big spikes of BEMF and is the reason for why the bike controllers have those HUGE LOW ESR filter capacitors that at this time are technically impossible to make surface mountable.

BUT then again i could be wrong. over on another forum, either yahoo's power-assist or e-motor-assisst someone is trying just this. he has a big RC out-runner and a ESC rated for 100A. plus a ton of custom machining to create a 2 or 3 stage belt reduction drive to get the 10,000 rpm motor down to the 300rpm that he needs for the bike. lucky for him he has access to some nice CNC machinery to make his own parts. but if he had to pay for the machining, it would be way cheaper to buy a car. the motor and ESC + BEC and servo tester (that he uses as a throttle control) combined cost almost as much as a conventional hub motor and controller.

he had originally intended (i think) not to make it cheaper but to prove it could be done. he also wanted a smaller package than the typical e-bike uses. well, after seeing the very pretty machining of his reduction drive. i can say it is impressive to look at, but just as big if not bigger than the equivalent bike part. and if he had to pay for machining his bill would be in the thousands.

rick
 
I've only known RC motors to be quite small and light. I would've naturally gravitated to RC motors had I known that they are that powerful and inexpensive! Wow, now I can retrofit two of them bad boys on my wife's revive! Torque motor at the rear and speed motor at the front! Like Deafsooter's set up. Cylindrical motors would require fabbing mounting brackets but it would eliminate the need to fab torque arms. Add some A123 cells and my wife will smoke me!

Man, we got loads of RC hobby shops here in Japan....

This was my vision, lighter motors, and powerful batteries! 8)

Damn, you got my wheels turning once more, thank you!
 
I think oofnik has a valid point. The technology employed in most of the controllers I've seen is pretty last century. There is always room for improvement. One thing about EV's is we don't need to worry about a few grams of weight, so we can build things heavier. Heavier means it takes longer to burn up.

The RC market is many times bigger than the EV market and can support better R&D. Until ebikes and scooters become more mainstream, I suspect it will stay this way for quite a while. That's why we have to re-engineer stuff here :twisted:
 
The RC market is many times bigger than the EV market and can support better R&D. Until ebikes and scooters become more mainstream, I suspect it will stay this way for quite a while. That's why we have to re-engineer stuff here :twisted:
\

This is the most true thing yet said in this thread. There are no ebike magazines, or competitions, even real organizations. (I am a card-carrying member of the AMA. The closest equivalent in the ebike world is... membership to this forum. :| ) Most vendors catering to the "hobbyist ebiker" are pretty tiny operations, like ebikes.ca.

The most important thing we can do, as ebikers and pioneers, is to spread it. To improve the technology in the long run, we need to expand the market for fun, relatively inexpensive, socially-responsible personal vehicles.
 
The trouble is that, although you can save weight by running a smaller motor faster for a given power output, and assuming that you can gear it down to the required speed with reasonable efficiency, you are reducing the motors ability to handle/dissipate heat by reducing its mass (even though the surface area to volume ratio increases). That's ok if you stay close to the peak efficiency, but if not, it may quickly overheat. So, depending on the case, you might be "condemned" to travel at top speed if you have a fixed ratio between motor speed and wheel speed. So, variable gearing becomes more necessary as the motor mass is reduced, right?

Sure, there's potential in using high speed motors but there's also a whole set of problems. You need to find the best compromise for a given design objective.
 
Out of interest, has anyone tried hooking one of these sensorless RC controllers up to a bike hub motor?

The TongXin sensorless controllers are known to be a bit unreliable, so maybe one of these might make a good replacement, if it works.

Jeremy
 
oofnik, A year ago I was exactly where you are. It was raining here, and windy, and I had spare parts from some planes, and thought "Why the hell can't I?"

Well a year ago the problem was heat. There was no way to keep the motors available last spring cool enough, and there were virtualy no controlers that could handle partial throttle for long periods in voltages over 36V.

A year has made a huge diffrence.


But there are still some technicle hurdles. 300 hours on a RC motor is a full life. Thats a bout what the bearings can handle before expiring. The RPMs are also high. a bike wheel needs to cruse at 200 to 300 rpm. Most RC motors run 10-40 thousand RPM. Gear reduction on that scale kills efficancy, but also causes some other problems.

I've got a 400 watt Eflite sitting in a wrecked airframe thats taunts me daily. the total weight of the motorand controller was under 4 ounces. For under 10 pounds of equipment and batteries, I could build it into a 40 mile range bike with plenty of power, Comparable to my 500 watt C-lyte Hubmotor.... if there was a way to deal with its 35,000 RPM at 11 volts.
 
I'm glad I got this discussion going. :D
You guys have brought up some points I didn't give much thought to such as bearing life, gear reduction complexity / efficiency, heat dissipation ability, etc. Which brings me to my next idea:

Jeremy Harris said:
Out of interest, has anyone tried hooking one of these sensorless RC controllers up to a bike hub motor?
Best of both worlds? Maybe?
The only thing I can think of as to why this wouldn't work is that the R/C controllers are designed to operate motors at 40,000 rpm and may not particularly like being run (or even work at all) at such low speeds. But a lot of the latest generation models are fully programmable, including features like adjustable battery cut-off voltage, throttle delay, electronic braking and governing, all in the same tiny package. The advantages with controlling with a programmable integrated DSP as opposed to discrete PWM, op-amp and regulator chips are seemingly endless. So who has tried powering our big bad hub and chain drive motors with high power R/C ESC? This is my favorite idea thus far for my project.
 
Not all the RC motors are such high speeds - the larger ones work between 4000 and 10,000 rpm.

I've been testing the Castle Creations HV 85 controller, which is one step down from the one that Matt used on his recumbent. It senses direction in a fraction of a second and works fine at 0 to 5000 rpm speeds. I've no experience of using it on a direct drive hub motor, though.
 
Miles,

This sounds like it might work well as a fairly tough controller for a geared brushless hub motor. The major advantage seems to be that these RC controllers are programmable, so can be tweaked to give best performance.

Although I'm now busy building an electric motorcycle, my Tongxin powered recumbent is still in regular use. The one niggle I have is that the controller will do what so many have done and go "phut" before too long, so having an alternative would make a lot of sense. I've already designed, built and tested a throttle interface, using a small microcontroller, for the motorcycle, as I wanted to be able to fine tune the throttle response. It would be simple to change the code so that instead of outputting a control voltage it delivered a PWM signal to control an RC controller. I wouldn't even need to change the PCB, as the present design uses PWM to generate the analogue output voltage.

Which motor have you tried it on?

Jeremy

PS: I assume you're the same Miles from the UK Pedelec forum?
 
Hi Jeremy,

Yes, the same. It's getting a bit testy on Pedelecs, these days.... :mrgreen:

I should think it will work ok with the Tongxin - light rotor and freewheel. Not so sure about the heavy direct drive motors, though...

I've been using it with a 1kg motor from Astro Flight - only on my test bench, so far.

If you're on a budget you might investigate this: http://www.unitedhobbies.com/UNITEDHOBBIES/store/uh_viewItem.asp?idProduct=4691

Looking forward to seeing your motorcycle conversion.
 
More than on person has tried running a bike motor with a sensorless RC controller. The posts are burried in the archives somewhere....

Seems to me, one case it seemed to work OK. Another user fried the controller. The inertia of a bike is several orders of magnitude greater than your typical airplane propeller.
 
our bikes in general do not use clutches like ice engines use. ice engines are running before we couple the load that we are trying to move. our ebikes have the load connected while the motor is off and stationary. we need then to start up and move the load moving in the right direction instantly.

Most of the problem of a sensorless controller is to start the motor moving with enough power to keep it going. Picking a couple of coils to energize, hope that the rotor will orient to that position and then start a commutation sequence. the first few pulses of that commutation sequence are delivered automatically in a predetermined sequence to get the rotor moving fast enough to generate some BEMF. until the motor does start to produce BEMF the commutation is more or less hit and miss. depending on the load the rotor may or may not have had the time to move to the correct position before the next phase pulses arrive. start up torque is quite low. in an RC model not much mass is being moved to start. in the case of a propeller (plane, helicopter or boat) the connection is loose the inertial mass is almost non-existant and a considerable low speed movement can be made at start up with virtually no load. the chances that a rotor will move and orient itself properly at the start are pretty much guaranteed. this is not the case with a bike where the load is a large inertial mass solidly attached to the motor shaft. the chances that the rotor will kick to the correct orientation for reliable starts are considerably reduced.

some will say what about an RC car? isn't it just like a bike. well almost. but remember they are using a motor and a controller capable of the same power as a bike o start up a model that weighs a fraction of what the loaded bike weighs. all other things being equal the rotor in a 5lb rc car has a much better chance to orient than one in a 200+++lb bike plus rider.

this is where a sensored motor has a huge advantage. the position of the rotor is known at any moment. the correct phase pulses can be delivered to continue the rotor movement. starting torque is much higher and much more reliable.

on the other hand once the motor is running the BEMF generated using sensorless techniques can be used to measure instantaneous motor performance. that information can be used to adjust the timing pulses to tune motor performance. this is not possible with Hall sensors. Hall sensors only report motor position.

so here is the issue. we need a sensored design for a reliable start up. we still need the hall sensors at low rpm where the motor not revolving fast enough to produce a reliable BEMF signal. but once we have a reliable BEMF the sensorless design has the advantage. choose your BEMF feedback control method Trapezoidal? Vectored? some new combination? none of the controllers that we can afford to buy uses both of these techniques in the same box. they are designed for one or the other but not both.

now if you know someone who works in a secret government lab developing drives for high speed torpedoes. we need some samples of their controllers. baring that, someone will have to sit down and design one. Any Volunteers, anyone... anyone.....

rick

when i was a kid i used to ask myself almost the same question. i used to wonder why i had to push around that big heavy lawnmower powered by a briggs & Stratton 2 Hp motor. why could they not use something like the Enya .60 my model plane used and produced almost the same power.
 
Knuckles said:
Heck let's add data logging and a wireless interface for a nintendo DS lite. :idea:

Don't forget the touch screen. It could be used to enter data while the top shows predicted performance graphs overlayed on the current actual performance graph.

That and you could make it do the Mario jumping sound whenever it detects the front wheel leaving the ground. It should play the star music if you have a boost pack 8).
 
fechter said:
The inertia of a bike is several orders of magnitude greater than your typical airplane propeller.
wow that sorta nails it eh? How about decoupling things with an air gap?
Use these RC motors to run a "prop" as enclosed turbine, air as transmission fluid, and direct drive to the wheel... hmmmmm.... 24" wheel at ummm... some horrible efficiency losses/slippage... even at 5% of prop (turbine?) speed of say 10,000rpm? So, 35mph? Ya have to keep tip speeds below about 700mph (approaching speed of sound)

seen on the web re RC props:
"For Master Airscrew props, we suggest the
following formula: RPM Operating Limit = 160,000 divided by Diameter in inches. In
our .40 size engine example, a 10" diameter g/f nylon prop has an operating limit of
16,000 rpm, well above the requirement of a .40 engine. "

Combine RC and EV...? Simple answer here:
http://www.youtube.com/watch?v=UnntlyUiNFU
:lol:
locK
 
Yah. What he said. :twisted: Appears to involve some old-school lithium chemistry
l
 
Don't Ya wish ya had one of those for those tailgater's in rush hour traffic?

rick
 
Back
Top