Using RC motors on E-bikes [Archive]
- Thread starter dirty_d
- Start date
- Status
Jeremy Harris
100 MW
Many thanks for the link, Miles, I can see many hours of interesting browsing ahead....................
I could run both motors via over-run clutches, but I like the idea of a simple drive. One other benefit is that I can most probably use commercial off the shelf PPG prop drive parts to build the reduction drive if I use a single shaft, which would make the construction much easier.
From a brief read of that link, it looks as if my idea of locking the shafts together such that the phases are properly synchronised will be fine. I see there's even a neat trick to get the shafts electrically aligned described there, by putting DC through one phase.
The next challenge is to find a suitable controller, at a price that makes the risk of destroying it in an experiment not too big an issue.
Meanwhile I'd better get back to the electric RD50MX build - so many projects, so little time......................
Jeremy
[edited to add: Thanks for the second link, Miles, that's exactly what I had in mind.]
I could run both motors via over-run clutches, but I like the idea of a simple drive. One other benefit is that I can most probably use commercial off the shelf PPG prop drive parts to build the reduction drive if I use a single shaft, which would make the construction much easier.
From a brief read of that link, it looks as if my idea of locking the shafts together such that the phases are properly synchronised will be fine. I see there's even a neat trick to get the shafts electrically aligned described there, by putting DC through one phase.
The next challenge is to find a suitable controller, at a price that makes the risk of destroying it in an experiment not too big an issue.
Meanwhile I'd better get back to the electric RD50MX build - so many projects, so little time......................
Jeremy
[edited to add: Thanks for the second link, Miles, that's exactly what I had in mind.]
Drunkskunk
100 GW
A few comments on the twin motor idea. First, it works fine sometimes, then sometimes it doesn't. since the controller is looking for the back EMF partly as a timing sourse, having a small variance between the two will have a cumulitive effect, and knock the timing out for both of the motors. when it works, it works fine. One way to cure the problem is to lock the motors as precisly as posable. If you use a custom shaft, and put both motors on the same outpout shaft, you can align them more precisly, and vibration free.
You also need 1 realy big controller to run 2 motors, which can be a lot more expensive than 2 smaller controllers.
The other point is many of the PPGs are using 1 motor, and there are a number of motors out there that will be more than enough power.
You also need 1 realy big controller to run 2 motors, which can be a lot more expensive than 2 smaller controllers.
The other point is many of the PPGs are using 1 motor, and there are a number of motors out there that will be more than enough power.
Drunkskunk
100 GW
Miles said:
Yeah, I've been bribed. I must have one now!
1.9Kg weight, 15Kw power
in the last 3 years brushless motors have been increasinging in power weight almost exponentialy.
Jeremy Harris
100 MW
15kW is OK, I'm running with 20hp at the moment (an H&E R120 two stroke). I wouldn't want to go with less than about 15hp (~11kW), as the take performance and climb rate would be poor.
The great advantage of using two smaller motors is that the cost is a great deal lower, which makes them viable as a hobby experiment. That big Predator 37 costs nearly one thousand Euros (not far off $1600), whereas two of those cheap 6kw motors mentioned earlier would only set me back about two hundred Euros - that's a big difference.
Jeremy
The great advantage of using two smaller motors is that the cost is a great deal lower, which makes them viable as a hobby experiment. That big Predator 37 costs nearly one thousand Euros (not far off $1600), whereas two of those cheap 6kw motors mentioned earlier would only set me back about two hundred Euros - that's a big difference.
Jeremy
dirty_d
10 kW
http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=4691
theres a 100A 50V controller for $95.00. maybe there is something cheaper though. ive been thinking about the motor current limiting. it would be hard to put shunts on the motor phases because they are constantly switching polarity, you would probably need some expensive op-amps to convert the voltage across the shunts to a voltage relative to ground. you could use an allegro current sensor but i think they stopped making them, or i just cant find them, they also have crappy accuracy at different tempuratures. so i figured you could put a shunt on the batteries ground, and use that signal, since youre using an MCU its easy to do the math to get the motor current, if it senses 20A through the batteries ground, and its giving the controller a 50% throttle signal, then the motor current will be about 20 / .50 = 40A. i dunno, maybe there is a better way to do it directly from the motor phases.
theres a 100A 50V controller for $95.00. maybe there is something cheaper though. ive been thinking about the motor current limiting. it would be hard to put shunts on the motor phases because they are constantly switching polarity, you would probably need some expensive op-amps to convert the voltage across the shunts to a voltage relative to ground. you could use an allegro current sensor but i think they stopped making them, or i just cant find them, they also have crappy accuracy at different tempuratures. so i figured you could put a shunt on the batteries ground, and use that signal, since youre using an MCU its easy to do the math to get the motor current, if it senses 20A through the batteries ground, and its giving the controller a 50% throttle signal, then the motor current will be about 20 / .50 = 40A. i dunno, maybe there is a better way to do it directly from the motor phases.
recumpence
1 GW
I can tell you from experience that you cannot effectively link two motors together with one controller. I have tried it with my RC helis. I also did alot of research into it a couple years back. Even the double AXI requires two ESCs to work. 
But, the up side is, two small controllers are only slightly more expensive than one big one and you get redundancy (if one fries, the other will still run).
I have cure the belt skip issue with one wide 15mm belt. However, my efficiency went down. It did not go down much. But, a small amount. I do not have numbers for you because my logger bit the dist from overamping it. I have a 150 amp logger coming at teh end of this month. The up side is, the whole system is running smoother and quieter than ever.
Anyway, yes there are cumulative losses with a double reduction like mine. That is undeniable. I have looked into doing a single reduction using a readily available 12 tooth motor pulley and a 72 tooth secondary pulley driving a freewheel to a left side chain to the disc brake flange on the left side of the hub. That seems to be the best solution to this. The only reason I did not go that route was space constraints under my seat. The big 72 tooth pulley would not fit.
At any rate, I am not an electronics expert. I am a self taught mechanical designer/problem solver/engineer of sorts. If it is mechanical, I can probably deisng, debug and refine it. All the technical jargon regarding motor IO, efficiency curves and the like are foreign to me. The best I can do is build it, run it, and measure AH useage over a give distance to calculate rage per charge.
Oh, I have found a problem with the AXI motors (I mention this in my build page on the web site). The AXI motors have a strange trait. They put off a strange back EMF issue from about 2800 RPM through 3200 RPM. In that range, the back EMF is sketchy and the ESC gets confused easily. So, through that RPM, I must back off the throttle so the motor doesn't "Sing" and lose power there. I just back off for a second and get back on it as it transitions through that RPM range. I spoke with Castle Creations about it and they said this is the only motor that does that. They said they could program the software around it, but they will not just for one brand motor. So, for now, this is an annoying (though admittedly minor) issue.
Other than that, I have the bike pretty well refined at this point. I am glad I built it. But, my next setup will definately use a single stage reduction to increase efficiency.
Sorry for the long post!
Matt
But, the up side is, two small controllers are only slightly more expensive than one big one and you get redundancy (if one fries, the other will still run).
I have cure the belt skip issue with one wide 15mm belt. However, my efficiency went down. It did not go down much. But, a small amount. I do not have numbers for you because my logger bit the dist from overamping it. I have a 150 amp logger coming at teh end of this month.
The up side is, the whole system is running smoother and quieter than ever.Anyway, yes there are cumulative losses with a double reduction like mine. That is undeniable. I have looked into doing a single reduction using a readily available 12 tooth motor pulley and a 72 tooth secondary pulley driving a freewheel to a left side chain to the disc brake flange on the left side of the hub. That seems to be the best solution to this. The only reason I did not go that route was space constraints under my seat. The big 72 tooth pulley would not fit.
At any rate, I am not an electronics expert. I am a self taught mechanical designer/problem solver/engineer of sorts. If it is mechanical, I can probably deisng, debug and refine it. All the technical jargon regarding motor IO, efficiency curves and the like are foreign to me. The best I can do is build it, run it, and measure AH useage over a give distance to calculate rage per charge.
Oh, I have found a problem with the AXI motors (I mention this in my build page on the web site). The AXI motors have a strange trait. They put off a strange back EMF issue from about 2800 RPM through 3200 RPM. In that range, the back EMF is sketchy and the ESC gets confused easily. So, through that RPM, I must back off the throttle so the motor doesn't "Sing" and lose power there. I just back off for a second and get back on it as it transitions through that RPM range. I spoke with Castle Creations about it and they said this is the only motor that does that. They said they could program the software around it, but they will not just for one brand motor. So, for now, this is an annoying (though admittedly minor) issue.
Other than that, I have the bike pretty well refined at this point. I am glad I built it. But, my next setup will definately use a single stage reduction to increase efficiency.
Sorry for the long post!
Matt
recumpence
1 GW
Hey "D" are you the one I spoke with the yesterday?
Matt
Matt
Matt, just to be clear - Your efficiency went down with the idlers on or off? If you can manage without them, I would have thought that would just about compensate, no?
Also, I think you should say 3 stage and 2 stage reduction or there's nothing left for me to call mine
Also, I think you should say 3 stage and 2 stage reduction or there's nothing left for me to call mine
recumpence
1 GW
My efficiency went down wen I increased belt width from 9mm to 15mm. It was terrible (huge motor heat) with the idler. So, I kept the 15mm wide belt and removed the idler. This increased my efficiency quite a bit (motor heat dropped enormously). Then I reduced belt tension. I still have no belt skip, but now the motor runs much cooler and quieter (and smoother) that it iever has.
My overall reduction is one primary belt, to a 1/4 inch chain drive secondary, then into a BMX chain to the rear wheel. My overall reduction is 3 stage from motor to wheel. It is two stage internally in the power unit.
Sorry if it sounds confusing. I agree it needed clarification.
Matt
My overall reduction is one primary belt, to a 1/4 inch chain drive secondary, then into a BMX chain to the rear wheel. My overall reduction is 3 stage from motor to wheel. It is two stage internally in the power unit.
Sorry if it sounds confusing. I agree it needed clarification.
Matt
recumpence said:
Would there be some kind of resonance (electrical and/or mechanical) in the range between 2800 and 3200 rpm?
What is the frequency of the BEMF in that range?
Wonder if you had used an scope to see the BEMF in that range?
What do you mean "sketchy"?
recumpence
1 GW
I am not an electronics engineer. I am just repeating what Castle Creations told me.
They said it is only an issue with AXI mtors.
Not sure other than that.
Matt
They said it is only an issue with AXI mtors.
Not sure other than that.
Matt
dirty_d
10 kW
Matt, do you know anything about that beefy HXT motor? that one has a 12mm shaft so the smallest pulley i could find for it was 24t, and thats with a 10mm bore that would need to be enlarged. if i went with a 72t pullet after that i would have a 3:1 reduction, then i could use a 3.5:1 reduction on the chain for 10.5:1 total, that would put the bike at 40mph at 5500rpm(4300W/90% efficiency), no-load speed is 6240rpm at 48V. does that sound like it would be ok?
d,
You could certainly go down to a 20t (the hub for that is still over 20mm). Less than that, you might have to make your own hub for it, or use one of the fancy split ones. Sticking with 24t you'll be ok for 2kW, though (at half speed) with 20t you're down to about 1.5kW (15mm GT2 belt).
You could certainly go down to a 20t (the hub for that is still over 20mm). Less than that, you might have to make your own hub for it, or use one of the fancy split ones. Sticking with 24t you'll be ok for 2kW, though (at half speed) with 20t you're down to about 1.5kW (15mm GT2 belt).
recumpence
1 GW
dirty_d said:
I could make a very small tooth count pulley for you. I use standard pulleys and modify them for my use. I would bet you could go down to 14 tooth. That would completely eliminate the hub. But, I could drill and tap two set screw holes between teeth. I do that with my own pulleys.
Let mw know if you would like that done.
As long as I have a good 1/8 inch material for set screw threads we are good. 8)
Matt
recumpence
1 GW
Miles said:
Man, every time I make something for myself, it turns into a business.
Matt
dirty_d
10 kW
Matt, how much would one like that cost? im not building this thing right now, but when i do ill send you a message. are you using the standard HTD pulleys and belts or the GT2 ones? i thought i read somewhere that HTD and GT2 pulleys are the same only the belts are different, is this true?
Miles, what do you mean limited to 2000W and 1500W, you mean because the belt will shred?
Miles, what do you mean limited to 2000W and 1500W, you mean because the belt will shred?
recumpence
1 GW
dirty_d said:
This is where it gets weird.
I have been told by a number of engineers including a pulley manufacturer (I spoke with the pulley machinist myself) and a belt manufacturer and both told me HTD and GT are different profile for the same pitch (in my case, 5mm). However, I am currently running a HTD belt with a GT pulley and they appear identical. I held the belt and pulley (flangeless pulley) together and looked to see light visible between the belt and pulley teeth to see where the belt is engaging and where it isn't. I did this with both HTD and GT belts on the same pulley and they appear to the naked eye to be identical pitch and profile. But, that may just be visual, not functional. However, the torque this belt will take is significant and it is running quite efficiently. So, I would say it this way, order the same belts and pulleys if you can. If you need to mix and patch HTD and GT, from my experience over the last few days of testing, they do indeed work together.
Does that make any sence?
I may be missing something, but again, this is my personal experience.
Oh, I think $25 of machining time plus whatever the cost of the pulley is would be fine. That would cover 45 minutes of pulley modding time. So, if the pulley is $10, that would eqal a total of $35 for the modified pulley.
Matt
recumpence
1 GW
Miles said:
Thanks for the heads up. I was wondering what the difference was. I cannot see it from looking at it. It must be very minimal difference.
At any rate, no-one would tell me wat teh specific difference was.
Thanks Miles. 8)
Matt
- Status
Similar threads
