High power RC motor and drive unit production

[CNCAddict]

That's awesome they're supplying the parts without haggling on the phone. I don't see the 100t pulley though. Only the 60 and 80, can you post a linky?

[Miles]

The 100t is from the Strida bike (http://www.strida.us/), it's not a C-Drive part, but it has the same BCD (104mm). There is also an earlier version of the pulley, which fits on a standard 18t sprocket.

I used it on my first build:

[MitchJi]

Hi David,

Looks like the ixi uses components from the C-Drive system....
http://www.cycledrive.com/home.html

So, now we have a choice of 60t, 80t and 100t pulleys....

Hi Miles,

They list two belt length's, 1200mm and 1280mm. Are compatible belts available in other length's?

Thanks!

Mitch

[Miles]

I should imagine they're just standard Gates PolyChain spec., so any 8M PolyChain (or even PowerGrip/HTD) belt could be used. The PolyChain belts now use carbon fibre reinforcement (as David said) and not the Kevlar listed on the C-Drive site (it's stronger and stretches less, but is a bit less forgiving....).

[CNCAddict]

Miles...I thought I would be the first to use that setup...aargh Anyhow, do you have a link to more info on your bike. I'm interested in the pulley adapter since I haven't yet designed this part. Oh, yeah it's for sure a standard Polychain 8mm tooth profile. The best prices on these belts is by far http://www.limct.com. You can get a ton of different lengths.

Edit: OK, here it is viewtopic.php?f=6&t=5313&st=0&sk=t&sd=a

[Miles]

Miles...I thought I would be the first to use that setup...aargh

David,

I thought you were going to make an adapter for the disc brake mount?

[See, also: viewtopic.php?f=28&t=7305]

[skydog]

There was Solar car out of MIT, very light weight maybe 250lbs. with batteries, etc.
One of the very cool things about it was the rear sprocket (it was a trike with 2 front wheels) that drove the rear wheel was also the rotor for the disc brake. I may have a photo of it somewhere. Nice weight and space reduction.

[CNCAddict]

I thought you were going to make an adapter for the disc brake mount?

My personal EV won't have any pedals. Just the belt going to the right side...I just want to make the setup pretty standard so I can use any top quality freewheel hub since I expect they won't last long with the huge amount of torque going to it. I've ordered most of the motor parts...and will begin carving up metal next week Still tweaking the stator and magnets....and I'm afraid this thing is gonna have HUGE cogging torque on the order of 1Nm. Prolly gonna make it a lowsy low speed EV motor. Everything I do to decrease cogging also screws up the motor performance. We'll see if that turns into a big problem, but I suspect it will be ok.

[Miles]

Ah, Ok.

The Mk III Strida pulley can be used with a standard 18t bike chain sprocket as its hub - it's not the greatest piece of engineering design I've seen, though - probably not up to your requirements.

I'm looking forward to seeing this motor...

[swbluto]

I thought you were going to make an adapter for the disc brake mount?

My personal EV won't have any pedals. Just the belt going to the right side...I just want to make the setup pretty standard so I can use any top quality freewheel hub since I expect they won't last long with the huge amount of torque going to it. I've ordered most of the motor parts...and will begin carving up metal next week Still tweaking the stator and magnets....and I'm afraid this thing is gonna have HUGE cogging torque on the order of 1Nm. Prolly gonna make it a lowsy low speed EV motor. Everything I do to decrease cogging also screws up the motor performance. We'll see if that turns into a big problem, but I suspect it will be ok.

Are you using any simulation software to greatly ease your efforts? I've seen some around, but I imagine they cost thousands given how specialized it is.

[CNCAddict]

I'm using FEMM which is free. But I have custom lua scripts to help me out. Plus, none of it really helps unless it's possible to accurately interpret the results....so your mileage may vary

P.S. Castle is still having problems with the SHV250 so now the ETA is a big question mark once again. I'm seriously thinking of starting an open brushless design group to get this going. OSMC has never really warmed up to the brushless idea, and there is a giant hole in the market for a reliable, tweakable controller that is adaptable to a range of voltages and currents. I'm developing some nice motors...but they won't be worth much without getting some serious current and voltage to them

[recumpence]

Well, I booked the entire day today for drive unit prototyping. But, my material hadn't arrived yet. So, I made something else for my bike (I will post about that later).

Anyway, it is getting dark and, WOW, UPS finally brought my materials! Of course, I have a full day of work tomorrow. So, Friday and Saturday will be devoted to machining.

David, I would love to get a 40 series motor from you. Let me know how much.

Matt

[GGoodrum]

Matt, is there another thread that descibes the details of this new single/dual reduction system you are talking about for this new motor? Did you ever post any pics/drawings of this new drive? I'm having a hard time visualizing what this will look like.

-- Gary

[recumpence]

It is hard to describe and hard to draw to show. Basically, the drive unit will consist of mechanical tubing holding the jackshaft bearings and those tubes will be held together with double clamps that look like front fork tripple clamps. The entire unit will be very adjustable and configurable for various layouts.

I am really excited to post pics of it. I will try as hard as I can to get a prototype machined ASAP.

Matt

[MitchJi]

P.S. Castle is still having problems with the SHV250 so now the ETA is a big question mark once again. I'm seriously thinking of starting an open brushless design group to get this going. OSMC has never really warmed up to the brushless idea, and there is a giant hole in the market for a reliable, tweakable controller that is adaptable to a range of voltages and currents. I'm developing some nice motors...but they won't be worth much without getting some serious current and voltage to them

Hi David,

Does this one sound good?:
http://endless-sphere.com/forums/viewtopic.php?f=1&t=6944&start=0&st=0&sk=t&sd=a

've just found this forum and would like to introduce myself in my first post.

I work for a company in Exeter, UK called Green Energy Technologies Ltd
We are just getting into the electric vehicle market with some brand new technologies. Our first product is a high efficiency Power/Motor Controller that we call Split-Pi. It converts up and down, so you can have 0 to 2 times battery voltage supplied to your motor and it is bidirectional allowing current to flow from battery to motor to accellerate your vehicle or from motor to battery for regenerative braking. It also has a nominal efficiency of 95% and peak efficiency of 99% at full load. The output voltage is PURE DC, not PWM like most other controllers. PWM controllers also loose power in motor heating, but with pure DC, motor heating is dramatically reduced. Quick specs are 0-60V +-25A (1.5kW)

We have found many applications for this particular product, I wanted to know what you all thought of it for electric vehicle applications and stir up some interest. We were currently thinking of small bike scale for this particular product. We have higher power (10kW) products in development for large bike and small car scale EVs and are also planning to produce a complete electric/full-series hybrid drive train solution using some new and novel technologies

The prices you see (£695) are for our current model of Split-Pi which is over engineered for a bike controller, we are making a smaller, lighter and cheaper version specifically for the bike market around the £250 price. I know this is still much more expensive than controllers currently on the market, but it has several key features that make it better than other things out there:
* Up and Down voltage conversion.
You could have a 36V battery pack and Split-Pi will supply 0 to 60V to the motor.
When it is down converting, Split-Pi will supply 0-36V to the motor and more current for torque
When it is up converting, Split-Pi will supply 36V-60V to the motor for extra speed
The output can be controlled smoothly from 0 - 60V

* Bidirectional for regenerative braking
Split-Pi allows the motor to recharge the batteries when braking
It will continue regenerative braking until the motor comes to a complete stop

* 95-99% Efficiency even under regenerative braking

* Pure DC output (not chopped up PWM with varying time-space ration)
keeps motors cooler and results in higher efficiencies overall


Split-Pi currently works directly with brushed DC motors, an additional commutator is required for Brushless DC multiphase motors.

Split-Pi completely replaces the PWM controllers you use at the moment. It supplies a variable DC output that can be digitally controlled using a 2wire communication protocol or it can be controlled with a variable voltage input, i.e. from a twist grip on a bike handle bar

We envisage this product occupying the high spec end of the bike market where people spend over £1000 ($1500) on a bike, the sort of people that like to have the very best in new technology.

[deecanio]

Hi Mitch,

Jozzer was telling me about the split-pi controllers a few weeks back but the only problem we see is the 1.5kw limit at the moment, could be very interesting for the future but as you know i'm into the here and now side of things


Cheers,

D

[GGoodrum]

It is hard to describe and hard to draw to show. Basically, the drive unit will consist of mechanical tubing holding the jackshaft bearings and those tubes will be held together with double clamps that look like front fork tripple clamps. The entire unit will be very adjustable and configurable for various layouts.

I am really excited to post pics of it. I will try as hard as I can to get a prototype machined ASAP.

Matt

I'm still confused, but as I understand it, you will still have both single and two-stage versions of this, right? I'm thinking this would be a good fit for my NeuMotor 2215-3Y, with the P62 planetary gearbox on it. The kV of this motor is 480, and the GB provides a 6.75:1 reduction. Here's what it looks like:



I want to use this with a Nexus 3-speed rear hub, which has ratios of about 1.33:1, 1:1 and .75:1, so I figure I need another reduction of around 6:1 in order to get the speeds and torque I'm looking to get. What I'm thinking is putting your new widget in "series" with the existing pedal drive, using two oneways/freewheels, with two sprockets, on your output jackshaft, one that drives the rear hub, and a second one that connects to the front chainring. That way I can still use the pedal drive, with the 3-speed hub, just like stock setup.

I seem to remember you tried a Power Jazz controller, but had some issue with it. I have one of those that I was planning on using, but I'll pick up an HV110 and do the cap mod., if you think that's a better option.

-- Gary

[recumpence]

Hey Gary,

Yes, for your application, you will need a chain drive as a belt would not take the torque in that situation. I can set you up with a 6 to 1 .25 pitch chain drive. I would need your motor in hand to do the machining needed to mate it with the mount plate.

Also, what is the output shaft diameter (maybe 6mm)?

Oh, the guys on RC Groups are looking for you. The keep mentioning it "Where is GGoodrum?"

I think myself and Askman are about the only ones who know what is going on.

Matt

[MitchJi]

I want to use this with a Nexus 3-speed rear hub, which has ratios of about 1.33:1, 1:1 and .75:1, so I figure I need another reduction of around 6:1 in order to get the speeds and torque I'm looking to get. What I'm thinking is putting your new widget in "series" with the existing pedal drive, using two oneways/freewheels, with two sprockets, on your output jackshaft, one that drives the rear hub, and a second one that connects to the front chainring. That way I can still use the pedal drive, with the 3-speed hub, just like stock setup.

-- Gary

Hi Gary,

I plan to do exactly the same thing ("with two sprockets, on your output jackshaft, one that drives the rear hub, and a second one that connects to the front chainring") but using Davids motor. I was planning to ask Matt what he would do to will help accommodate that as soon as I gather a little more information.

If two of us use the same setup it and purchase them at the same time that might help?

I think what we need on the output jackshaft is:

1. The sprocket that connects to the front crank needs to freewheel so that the motor doesn't drive the pedals.

2. The sprocket that connects to the rear hub needs to be fixed because if it freewheels pedaling won't apply any force to the rear sprocket.

3. The output shaft (or both sprockets as a unit) need to freewheel if you want to pedal with the motor off and not have that force go back through the gears.

There are at least two ways to accomplish number three above. I'm not sure which is easiest, I think number 1.
1. Both the sprocket and freewheel can mount on the shaft and the shaft can freewheel (freewheel or oneway on the input side of the final shaft).
2. The guts from the right side of a freehub can be mounted on the output shaft. This whole assembly will freewheel. Then a fixed sprocket and freewheel can be mounted on the freehub.

Number one might be easier but would require a freewheel on the input side of the final shaft.

Option two would not require any modification to his drive except that it would be helpful if the final shaft could be modified to accept the freehub.

[CNCAddict]

For those interested, I've finished up the motor design for the first 2 prototypes. I'm building one to test myself, and sending the other one to Matt. I've lengthened the stator some to final dimensions of 6050. The weight will be pretty high at 3.5lb but it should also allow about 10lbft of torque continuous at the higher rpms (more cooling). For a 15:1 reduction that's 150lbft of torque at the rear wheel. If I do a 3 turn it will have a Kv of 155rpm/v and will need ~200A to get 10lbft. A 2 turn will be around 230rpm/v and need ~300A to do the same. I'm talking with Kelly about some future controllers, and they apparently have a sensorless in the works. 72V and 100A (not sure continuous or peak). I'll initially test this motor out with my spin200, and we'll see how it goes. Anyhow, I should have all the parts in a few weeks and the motors completed a few days after

Matt, the 40mm motors will be done really soon. All the production parts are done and they should be wound up in the next few days

[GGoodrum]

Hey Gary,

Yes, for your application, you will need a chain drive as a belt would not take the torque in that situation. I can set you up with a 6 to 1 .25 pitch chain drive. I would need your motor in hand to do the machining needed to mate it with the mount plate.

Also, what is the output shaft diameter (maybe 6mm)?

Oh, the guys on RC Groups are looking for you. The keep mentioning it "Where is GGoodrum?"

I think myself and Askman are about the only ones who know what is going on.

Matt

Yes, I definitely think the chain drive would be better. I remember trying to add a belt drive "extra" stage to a big Ion helicotper, so that we could use a Neu 1521 motor, but even with the Kevlar-reinforced belts, they would shread with the power we wer trying to put through them (5kW+, or so...). We originally tried gears, but even doubling them up, the nylon gears would melt. With the Kevlar HTD 9mm belts, I think we could get 2-3 flights before they came apart.

I barely have enough time to spend over in this world, so I haven't been on RCGroups in awhile. When I get to the point that I can give up my aerospace consulting work, maybe I'll have time to do both, but not now, unfortunately. Maybe you and Andrey can let those who are interested know this. I've also talked with Fred Bronk a bit.

The output shaft of the P62 gearbox is actually 14mm. It is a real beast. I think the output shaft diameter of the motor, going into the GB is 8mm. What about using the PowerJazz, worth a try? If you PM me your address, I'll send you the 2215/P62, for whatever 'fit" checks you need to do.

I'm still trying to decide which bike I'm going to install this on. I have a 16" Dahon Curve D3, which has a 3-speed S-A hub (similar to the Nexus 3...), but I'm thinking I will try putting one of the new BMC geared hubmotors on this. I also have a 20" folding bike that currently has a Cyclone 1000W setup. Other than being a bit noisy, this performs quite well, even on 48V (16s4p a123s...), so I'm probably going to leave this alone. With the power this Neu-based setup will be capable of, I'm now thinking the best choice might be to replace the front-mounted 5304 on my wife's Townie. This would solve several problems. this Townie came with a Nexus-8 hub in back, which also has an internal hub brake. The front wheel also has a hub brake, and the fork doesn't have mounts for standard V-brakes, or for disc brakes. I had the 5304 mounted on the front, and used some coiled wire/retifier-based "ebraking" as the front brakes. This worked okay, but the pushbuttons I was using kept failing. I was also woried about putting this much power (72V/50-60A...) through the stock aluminum forks. I then decided to replace the front fork with a heavy-duty steel mountain bike fork, but it took awhile for my local bike shop to find one that would fit the Townie headset. They finally did, but now there is an issue with getting a V-brake setup that would fit properly with the big Fat Frank tires that came on this bike. There is a clearance problem, something to do with where the V-brake mount is on this fork. I don't know, but at this point, I think what I want to do is put this back stock, with the standard fork and hub brake setup, and install this Neu-based setup, with your new widget, in a fashion like wht Mitch and I have been discussing, which is basically in series with the pedal drive. There is definitely lots more room to work with, on the Townie, and even on a 48V/16s a123-based setup, I think this will have tons of power. Having the 8-speeds on the Nexus will provide even more options, I think, although I doubt that any more than 2-3 gears would ever be used/needed. With 8-speeds you could have stump-pulling low-end torque and still have a 50 mph top end, all on a 48V setup. To me, this is really getting closer to my own personal "holy grail". I love the 72V/80-90A performance you can get from an x5-based setup, but I really want to standardize all my setups on 16s/48V, and make use of these wonderful planetary hubs to get both good low-end performance without sacrificing a "decent" top speed capability.

Mitch, I will have to think about the whole freewheel/oneway issue. I'm having a hard time picturing what you are describing, as I'm not completely up on what is readily available. I know that the rear-type freewheels seem to have big diameters, so some sort of adapter would be needed to make it work for a "jackshaft". but I know that SDP-SI has lots of one-way options that might work just as well. Marrying these up with sprockets is where I get lost. I get that we need a one-way/freewheel for the one that is driven by the pedal. This would eliminate having to have one in the crankset, like the Cyclone setup uses. I also see that it would need to be locked to the second sproket that drives the rear hub. I think then all we need is one more freewheel/oneway, somewhere in the reduction box that Matt is doing, right? Is that what you meant by the jackshaft input? I guess another way to look at it is that you have a direct sprocket (i.e. -- no freewheel/oneway...) that drives the chain to the rear hub (which has a freewheel already...), and then there are freewheel/oneways for both the motor sprocket and for the pedal sprocket. How this all comes together is what I don't yet understand.

-- Gary

[recumpence]

Hey Gary,

14mm should be doable. I can lathe the sprocket open to accomodate it.

I may make the face plate of the drive unit "Blank" to be machined for each specific application. That may be the easiest option.

Your bike will have CRAZY torque with that motor reduced 6 to 1. My guess is you will have near chain snapping torque.

I had issues with my PowerJazz. But, that may have just been a bad controller. One other guy here (don't remember who) is using one without problems.

Anyway, Dave and I went over some production details for the reduction unit. I may go straight to production without any prototyping because my design looks pretty much bullet-proof.

I just ordered $600 of materials (aluminum bar stock in multiple thicknesses, end mills, pulleys, collars, chain, sprockets, etc, etc, etc. Meanwhile Dave is programming the HAAS and the Trak DPM for production.

CNCadict,

Let me know how much cashiola to shoot your way for a 40 series motor and the prototype 60 series. Or we can work out a swap.

Oh, hey, can you wind one of the 40 series with a 250kv?

Matt

[GGoodrum]

A 6:1 reduction works for the Townie, as well, and the #25 chain drive sounds like it will certainly handle the power. Let me know what you need, both money-wise, and if you want be to send you the motor. I'm anxious to see some pics, as I'd like to start visualizing how to mount this setup.

I also have a Hacker A60-18L. This one would need a two-stage widget, I think. If the NM works out, I might do one of my other folding bikes with the Hacker. I notice they now also have an A80 series. These seem to be sized more like what David is doing, but probably not as beefy in the right areas, and a lot pricier. they have some new controllers as well, but these too are uber-expensive.

[recumpence]

When I get drives done we can discuss what is needed for your situation. It would be a bit custom, but not too bad.

I have decided to have the motor mount face plates for the reduction machined without any motor mount details. That area will be left blank. That way I can machine each one for the specific motor desired.

I have had quite a few custom requests. I don't mind doing custom stuff, within reason. This drive will be relatively modular. That will allow me to do a few custom bits, but still use the main foundation components. Each setup seems to have its own set of challenges.

I will probably stock a few of David's 40 series and a couple 60 series motors. Actually, it would be great to have a couple drives built up with one of each motor (in a common KV) ready to go for anyone who orders. Then maybe I can have a few different mount plates and what-not that the customer can choose from that will best match their application.

I am trying to keep this from becoming too complicated. But, any time you deal with custom or short run "Boutique" stuff, complexity becomes innevitable.

When we get closer to that point (in a few weeks or so) I can PM you my number and we can talk directly.

Oh, CNCaddict and I spoke at length today. He is setting me up with a E-bike spec 40 series (8mm shaft and 4 or 5 turn 250 to 300kv) motor along with the one-off 60 series to test. I will build a high efficiency setup with the 40 series and a high power setup with the 60 series to test.

Matt

[MitchJi]

Mitch, I will have to think about the whole freewheel/oneway issue. I'm having a hard time picturing what you are describing

1. I get that we need a one-way/freewheel for the one that is driven by the pedal. This would eliminate having to have one in the crankset, like the Cyclone setup uses.

2. I also see that it would need to be locked to the second sproket that drives the rear hub.

3. I think then all we need is one more freewheel/oneway, somewhere in the reduction box that Matt is doing, right? Is that what you meant by the jackshaft input?

4. I guess another way to look at it is that you have a direct sprocket (i.e. -- no freewheel/oneway...) that drives the chain to the rear hub (which has a freewheel already...) and then there are freewheel/oneways for both the motor sprocket and for the pedal sprocket.

5. How this all comes together is what I don't yet understand. I'm not completely up on what is readily available. I know that the rear-type freewheels seem to have big diameters, so some sort of adapter would be needed to make it work for a "jackshaft". but I know that SDP-SI has lots of one-way options that might work just as well. Marrying these up with sprockets is where I get lost.

-- Gary

Hi Gary,

You got it.

1. Exactly.
2. The sprocket to the rear hub is actually locked to the output shaft which is driven both by the motor and the pedals. The two freewheels pevent the motor and pedal sprockets from driving each other.
3. Yes!
4. Yes!
5. Me too I know what the functional requirements are but I'm not sure the best way to do this. Marcus has been doing this for years. I will try to ask him what he would suggest sometime next week and post his response. Matt and Miles are also experts so we should be able to work out a very good solution.

Matt already plans to have a oneway or freewheel on the output side of his final shaft. If he mounts the oneway or freewheel on the input end of our shafts instead of the output end that would cover most of what we need and I don't think it would be much extra work for him. It would require a larger sprocket than he would normally be using with the freewheel or oneway. If he does this we only need to mount a fixed sprocket and a oneway or freewheel on the on the output side and our drives are complete .

I got the impression that Matt thinks freewheel are stronger and have less friction than oneways so we might be better off with freewheels. If Matt goes with oneways and a 1/2" final shaft another approach, if he doesn't mind, would be for Matt to make a couple units with a 5/8" final output shaft. This would make it easier for use to use freewheel(s) instead of oneways because there are adaptors available for adapting freewheels to 5/8" shafts.

A serial drive to the right side definitely complicates the drive but I don't think its a major complication and once we get that taken care of the tradeoff is we don't need to do any modifications to the bikes. We only need to mount the motor/gearbox and connect two chains. I think if we get it worked out at least a few more people will be interested if for no other reason than the ease of installation.

BTW I asked if the "standard 6-bolt prop flange" output of the A80 Hacker is a problem for Ebike use and Richard responded as follows:

You should look for one that has the shaft coming out of the mount side. If you use a belt or chain, there will be a lot of radial load and you don't want the sprocket to be too far away from the bearings.