Using RC motors on E-bikes [Archive]
- Thread starter dirty_d
- Start date
- Status
That can't be right... 32 mohm is more like it...
Jeremy Harris
100 MW
That resistance is a typo, several motors on the Hobby City site have winding resistances listed in ohms, when they should be millohms.
On a different point, has anyone looked at adding Hall sensors to one of these big RC outrunners?
My thoughts are running along the following path: Brushless RC motors are as cheap as chips and very light for their power. Brushless RC motor controllers, that will reliably drive a motor at high current, low speed and high torque seem both expensive and relatively rare. Brushless ebike controllers that will drive lots of power, but need Hall sensors, are readily available and cheap as chips. Ergo, combine a cheap RC motor with a cheap ebike brushless controller for the best of both worlds.
It seems feasible to me, but depends very much on whether or not the magnets in a brushless outrunner lend themselves to a 60 degree or 120 degree sensor arrangement. Hall sensors are cheap and should be easy enough to fit into the bigger motors, provided that they have an amenable magnet arrangement. Failing that, it might be possible to add an external position sense arrangement to give the right sort of signals to an ebike controller.
Has anyone tried this? Does it vaguely make sense?
Jeremy
On a different point, has anyone looked at adding Hall sensors to one of these big RC outrunners?
My thoughts are running along the following path: Brushless RC motors are as cheap as chips and very light for their power. Brushless RC motor controllers, that will reliably drive a motor at high current, low speed and high torque seem both expensive and relatively rare. Brushless ebike controllers that will drive lots of power, but need Hall sensors, are readily available and cheap as chips. Ergo, combine a cheap RC motor with a cheap ebike brushless controller for the best of both worlds.
It seems feasible to me, but depends very much on whether or not the magnets in a brushless outrunner lend themselves to a 60 degree or 120 degree sensor arrangement. Hall sensors are cheap and should be easy enough to fit into the bigger motors, provided that they have an amenable magnet arrangement. Failing that, it might be possible to add an external position sense arrangement to give the right sort of signals to an ebike controller.
Has anyone tried this? Does it vaguely make sense?
Jeremy
johnrobholmes
10 MW
Typical outrunners have 14 magnets, 7 pairs. It would be easier to attach an encoder to the backside of it I think. I know it has been done once or twice, but there is not much need for sensored outrunners in the RC world- with exception to the rock crawlers. So far nobody has foot the bill to get sensored outrunners built for a market.
safe said:
32 milliohms seems more likely though, in this case :wink:
Whether a particular resistance is good or not depends on the rest of the specification, right?
Certainly, there are plenty of motors with winding resistances of over 0.5 ohms...
safe
1 GW
- Joined
- Dec 22, 2006
- Messages
- 5,681
The worst of the Unite motors calculate out to about 0.5 ohms. Their better motors are closer to 0.2-0.3 ohms.Miles said:
That's still quite high compared to these RC motor that rate in the milliohm range. The PMG132 motorcycle motor comes in at 0.025 ohms by comparison. That's 25 milliohms if I have my decimal point in the right spot. :?
And yeah... 0.0032 should have been 0.032... oops...
Wait, I've now officially made an error, so from now on that means everything I say must be wrong... isn't that how it works?
Jeremy Harris
100 MW
johnrobholmes said:
Very interesting, thanks.
Working this idea through a bit more, if I were to:
- Couple a pair of big outrunners together on a common shaft (seems easy enough to do, the shafts seem to just be straight ground bar).
- Ensure that the motor stators and rotors are both aligned.
- Fit three Hall sensors and a trigger magnet, to give 120 degree synch pulses.
- Connect the stator windings of both motors together in parallel.
- Hook the resulting combination motor to a suitable high power sensor-type controller.
I should, with luck, get a compound motor that's good for an awful lot of maximum power, with no worries about motor/controller synch, start up problems etc. I've no doubt that such a beast would want some active cooling, but that should be easy enough to sort out with a small fan.
Getting a controller that will handle the current would mean modifying something like one of those new Infineon controllers, by adding more FETs, but that shouldn't be too hard. A nice bank of IRFB4110s in there should do the job.
Using a couple of those big HXT motors I could get a compound motor that would put out about 12 to 13kW peak and weigh around 3kg or so. Pretty impressive numbers and not too expensive either.
Jeremy
dirty_d
10 kW
you would need more than one magnet for it to work good, that would be fine if the motor didnt change rpm, but on a vehicle you would need the same number of trigger magnets as on the rotor. ive thought about doing this, im not even sure if the hall sensors are fast enough to be used on these high commutation speed motors.
lawsonuw
1 kW
I'm waiting for a sensorless controller that has a encoder that can be attached to the motor's shaft. The controller can then spin the motor up sensorless and use the sensorless commutation to learn what encoder positions correspond to which phases. With an absolute encoder this would work from startup every time after training. With an incremental encoder it would work once the controller saw an index pulse. (i.e. all starts except for the first start of the motor would have full torque) Justin's controller ( mentioned http://endless-sphere.com/forums/viewtopic.php?f=2&t=5652 ) is close.
Marty
Marty
paultrafalgar
10 kW
eP said:
eP I downloaded your spreadsheet. I have no column or row headings. Is that correct? Hard to understand without any :wink:
johnrobholmes
10 MW
Sounds totally feasible to me. The hard part would be aligning the stators and magnets. Once you get the rotating bells on a common shaft (and aligned) it would be a matter of having a mounting plate for one motor that would allow the face to move in relation to the other motor so that the stators could be aligned. Then attach an encoder to the back end.
paultrafalgar said:
columns are mph 1,5 etc up to 50
rows are slope % 0, 1 etc up 18
Load datas are from :http://endless-sphere.com/forums/viewtopic.php?f=3&t=3981&p=58768&hilit=watts#p58768
I hope that help :wink:
edit
That spreadsheet is for advanced users mostly. Anyone can modyfy it , add some descriptions or more functions.
The spreadsheet show us the third gear for efficient motors is not useful and give us very limited efficiency gain only at flat at moderate speed (light load).
So two gears cover all conditions at decent efficiency.
Fumesucker
100 W
Since Castle Creations has a low cost line of low voltage controllers out (Thunderbird) I find it hard to believe that it is the "sensorless" part which is driving up the costs. Assuming of course that the CC low voltage controller does as good a job as their more expensive versions.
After all, the "sensorless" portion of the controller is mostly software, which once it is written can be duplicated easily enough any number of times.
After all, the "sensorless" portion of the controller is mostly software, which once it is written can be duplicated easily enough any number of times.
Fumesucker
100 W
dirty_d said:
If hall sensors aren't fast enough, you could always use optical, I've done enough with optical stuff to have no doubt they are plenty fast for controlling a motor.
Jeremy Harris
100 MW
Having checked, it looks like Hall sensors are fine at the sort of speeds these big outrunners work at. There are smaller inrunners using Hall sensors that run at very much higher speeds.
If I can find a way to fit sensors inside the stator, at the correct locations, then that would seem to be the best way to go. I woun't be able to see if this is possible until I actually get my hands on one of these big motors and take it apart, though.
If one of Knuckles/Keywins Infineon bike controllers can be adapted to drive a big RC motor, then that seems to be the best of both worlds - it'd certainly be cheap.
Somewhere here there is a thread with Hall timing diagrams on, I think I'll have to dig it out and see where I need to stick the sensors to get 120 degree phasing.
As for why HV RC controllers are so much more expensive I just don't know. The FETs are a bit more expensive for HV operation, but that doesn't come close to explaining the price ratio.
Jeremy
If I can find a way to fit sensors inside the stator, at the correct locations, then that would seem to be the best way to go. I woun't be able to see if this is possible until I actually get my hands on one of these big motors and take it apart, though.
If one of Knuckles/Keywins Infineon bike controllers can be adapted to drive a big RC motor, then that seems to be the best of both worlds - it'd certainly be cheap.
Somewhere here there is a thread with Hall timing diagrams on, I think I'll have to dig it out and see where I need to stick the sensors to get 120 degree phasing.
As for why HV RC controllers are so much more expensive I just don't know. The FETs are a bit more expensive for HV operation, but that doesn't come close to explaining the price ratio.
Jeremy
Jeremy Harris said:
HV RC controllers are expensive for the sake of marketing reasons IMO.
They are targeted for niche expensive models which need also expensive motors and batteries.
In EV worlds we need much cheaper energy batteries, cheaper motors and more controllers (the same extremly powerful) per vehicle unit.
@MitchJi
[quote ="MitchJi"]The chances of a sufficient volume of ebikes using two ESC's to cause a price drop highly unlikely.
Probably more likely (eventually) is increased volume of RC motors on Ebikes will cause a reduction in the prices of big escs.[/quote]
If e-bike folks will really want the same powerful and much cheaper ESC so soon or later such goodies become to reality i'm sure.
If they will be the same reliable as branded now so RC guys will want them too i'm sure.
As a result prices should drop. If e-bike folks will accept high current prices and will not be interested in alternative (white box) solutions then price drop is highly unlikely i agree.
The similiar things are valid for batteries too. If we want cheaper batteries we must evaluate them and to choice the most econo friendly option.
The difference is: good batteries cannot be homemaded but ESC we can do (if we really want them).
Jeremy, I had the same idea about using hall sensors in BLDS sensor less motors for modelers and then saw your post here in this thread.
Hall sensors are cheap and I don't believe you can synchronize controller and motor better than using them, especially when taking off and under a heavy load!
So lets stick them in and see.
My next project (after finishing my cd welder) will be:
E-dahon matrix 2008 bike.
And guess what I'll use :
Tower pro 5330-10T (your find)
Dahon matrix 2008 26 inch folder, only about 11.5 kg (goi it)
266T:12T HDT belt drivetrain (22:1 ratio) using a big freewheeling pulley utilizing 266T GEBE ring (will be interesting making so big freewheeling pulley) 1000g
my cloned crystalite controller (got it, incl. cables 800g)
13s3p A123 ( got it ,3.4kg incl. case)...43V
Speed should be about 40km/h
power about 2.4KW peak .........60Ax40V
total weight incl. batteries possibly only 17kg
The main chalenge will be making that huge lightweight freewheel.It will kind of follow the curvature of the wheel, be angled so it fits inside the regular frame. Something like a big Chinese hat with a 266teeth rim and freewheel inside. From the side it might look like a rear carbon wheel, if made from carbon...
Don't know if the pull of the belt slightly off the mounting freewheel plane will cause any problems.
I don't think that anybody has achieved a 22:1 ratio in 1 stage yet.
The advantage is obvious. The best efficiency and simplicity ! Possibility of using smaller and cheaper motors.
Hall sensors are cheap and I don't believe you can synchronize controller and motor better than using them, especially when taking off and under a heavy load!
So lets stick them in and see.
My next project (after finishing my cd welder) will be:
E-dahon matrix 2008 bike.
And guess what I'll use :
Tower pro 5330-10T (your find)
Dahon matrix 2008 26 inch folder, only about 11.5 kg (goi it)
266T:12T HDT belt drivetrain (22:1 ratio) using a big freewheeling pulley utilizing 266T GEBE ring (will be interesting making so big freewheeling pulley) 1000g
my cloned crystalite controller (got it, incl. cables 800g)
13s3p A123 ( got it ,3.4kg incl. case)...43V
Speed should be about 40km/h
power about 2.4KW peak .........60Ax40V
total weight incl. batteries possibly only 17kg
The main chalenge will be making that huge lightweight freewheel.It will kind of follow the curvature of the wheel, be angled so it fits inside the regular frame. Something like a big Chinese hat with a 266teeth rim and freewheel inside. From the side it might look like a rear carbon wheel, if made from carbon...
Don't know if the pull of the belt slightly off the mounting freewheel plane will cause any problems.
I don't think that anybody has achieved a 22:1 ratio in 1 stage yet.
The advantage is obvious. The best efficiency and simplicity ! Possibility of using smaller and cheaper motors.
recumpence
1 GW
With such a large pulley, you would not need teeth on it. Friction should be enough with a large diameter like that.
Matt
Matt
TylerDurden
100 GW
The Whizzer uses a spoke-mounted pulley:
http://woodstockwhizzer.com/whizzer_parts.htm
http://woodstockwhizzer.com/whizzer_parts.htm
paultrafalgar
10 kW
That's an interesting concept, Tyler. It's hard to see from the pic exactly how that inner rim/pulley is attached to the spokes. I suppose you might tack-weld the spokes to the edge of the rim/pulley but you'd have to be a pretty good welder, I think. Any other suggestions for an easier method?
Zoot Katz
100 kW
paultrafalgar said:
The current crop of 49cc bicycle conversions use backing plates and some rubber washers to bolt the sprocket through the spaces between spokes.
http://www.zippycycles.com/instruction_manual.html
I think that their setup is fine for a motor that is pulling but when coasting the small wheel will need to rotate at about 7000rpm. And at this speed the clutch bearing will put quite some resistance.
Not good when batteries/gas run out. I would use the ring and cut out the mounting part and glue it onto a big pulley with a freewheel. That is the idea.
Not good when batteries/gas run out. I would use the ring and cut out the mounting part and glue it onto a big pulley with a freewheel. That is the idea.
HAL9000v2.0
10 kW
TylerDurden said:The Whizzer uses a spoke-mounted pulley:
http://woodstockwhizzer.com/whizzer_parts.htm
Nice system, just give me idea that cheap pulley can be done by attaching 20" narrow rim on 26" wheel and use it with flat belt, it can be screwed to spokes with similar system like one holding magnet for cycloputer...
- Status
Similar threads
