Newbi - Add-a-motor DIY trike project.

[Hawk521]

Sun EZ-3 HD USX recumbent trike - Adding electric power.

The trike design led me to a plan to utilize the left rear axle for the motor
driven wheel and leave the right rear axle/wheel as the peddle driven wheel.
Each wheel is locked to its respective axle with a good heavy duty key, and
both axles are carried by ball bearings suitable for a heavy duty load. This
approach might seem odd but it does keep things simple.

Thus I began designing and acquiring off-the-shelf hardware that I intend to use
for this project:

1. Ordered an extended axle to replace the non-drive left axle.
This is simply a right side axle to be installed on the left side. It
provides about 2 inches of mount room for a final drive pulley.

2. Ordered a heavy duty freewheel and hub adapter to mount the motor drive
freewheel on the inside end of the above axle. Plan is to afix the final drive
pulley to this freewheel thus making it simple to peddle the trike w/o
motor assist. Nothing in the drive/pulley/belt assembly will need to rotate
when being peddled w/o motor assist.

3. Ordered a jackshaft assembly (solid aluminum) w/ball bearing shaft
from a go-kart parts supplier. Need this to reduce the overall drive
rpm to something suitable for directly driving the left rear axle. On
my trike's 20" wheels, 100 rpm of the rear axle equals ~6 mph.

4. Ordered 4 pulleys. HTD 5 mm pitch x 25mm width (2 @ 24 teeth, 2 @ 80 teeth)
Plan is to utilize all belt drive in two reduction stages. Hopefully this
will result in a fairly quiet drive system. My goal is that only the motor
will be audible - and even that will hopefully not be obnoxious.

Note: The pulleys I chose should yield approx 18 mph at 3000 rpm at the motor.
And if needed I can switch out a pulley or two to move that sweet spot
up or down the rpm range.

5. Will order the drive belts after I have finalized the drive components layout
and select a motor.


All the above leads me to the motor / controller / battery choices. Truly the
big ticket items! I have a vision of what I want, but could use some advice on
how to best achieve it.

a) I'd love to get a 3kw to 5kw BLDC motor that could handle 24v/36v/48v, yet
initially I would run the motor at 24v. I cannot fathom ever going past 48v as my
need for speed expired about 15 years ago. Am hopeful that my needs can be
met at the 24 volt level. Essentially I am willing to buy a bigger motor than
I need initially, in the hope that should I need more power out of it, I will
have the option to do so by adding more battery, volts, amps, etc. Also, I would
think that oversizing the motor a bit will prevent some overheating issues!?

b) Along with above motor I would like to purchase a controller that allows:
- use at 24 volts (and optionally 36 or 48 volts if needed in the future)
- provide reasonably smooth throttle (speed) control so the bike motor doesn't
act like it is simply ON or OFF. I'd like to be able to transition to the motor
without having it hammer the freewheel assembly too harshly.

c) For batteries my ultimate will be a LiFePO4 battery pack with 10-15 Ah or more.
But initially I would like to test with a pair of SLA batteries that are cheap.
(Or is there good reason to NOT try the SLA batteries at all?) Who knows, if I
don't utilize the motor too much I might be by for a while with SLA batteries.

I'm up for any advice or council on this project. Realizing that many of you have
gone down this road a time or two I will value your input. Do my requirements
make sense? Any obvious choices I should consider?

Thanks.

-hawk521

 

[mlt34]

Awesome trike. Sounds like a fun project, but I think you have some more research to do.

First of all, do you know what type of motor you plan to use? Looks like you'll have to rig up something custom - you obviously can't use a standard hubmotor. The right side drive you are replacing on the left side is a good start, but does it freewheel in the correct direction?

Next, 3-5kw is some serious power. Assuming you are on 24V, even 3 kw requires well over 100A. There's no way you'll be able to pull that from a standard 10-15AH LiFePO4. If you want to go LiFePO4, you'll need at least 50AH to even get close to keeping it in the safe range. To better accommodate that current, either SLAs or Lipos are more appropriate.

for the controller, I highly recommend Lyen's controllers. He can build you a great quality controller that can run on 24/36/48V. I've never had a problem with all of my Lyen controllers.

Any hall sensor throttle should be good for you. The acceleration is pretty smooth, though it will vary a bit from motor to motor.

Hope some of that helps.

 

[Rassy]

I had an earlier model of that trike before is was beefed up a little and renamed as HD. I used various 2 wheel pusher trailers. My Daughter-In-Law has the trike now with a front hub motor instead of a pusher trailer since she doesn't have the type of hills I had to deal with.

Anyway, my question is, how do you plan to deal with the cantilevered axle? Do you plan to mount the jack shaft and motor on the same angle as the axle?

Sounds like you have a good solution for the freewheel. I never took any of the drive system apart and don't remember where the original right side drive wheel's freewheel is located, but if it is at the inboard gear cluster and not out at the wheel hub your system should operate properly.

Good luck with the project, I'll enjoy watching your progress.

 

[Hawk521]

More research is underway. My post is part of an ongoing search for the best result I can build. I am a mechanical engineer - so some parts of this are fairly straight up in my mind. But I have very limited knowledge of ebikes and BLDC motors and controllers. Am learning as I go.

I plan to use a BLDC motor along the lines of one that Golden Motor sells as a 5kw motor. But that one is bigger than I really need. I've seen a 3kw motor for ~$400, but even that (as you point out) may be overkill for my needs. Am still researching motors and controllers.

Yes the freewheel attaches and freewheels in the proper orientation. The hub it threads onto fits tightly to the drive shaft and is locked in place by a 1/4" square key. The freewheel threads onto the hub properly. I have a machinist lined up to help mate the freewheel flange with the final drive pulley, and hope we can mount the freewheel in center the pulley to avoid any sideways twisting torque on the freewheel. I figure the freewheel bearings will hold up best if the belt load is centered on the freewheel bearing. I bought the heavy duty freewheel from Sick Bikes for $70. It is pretty robust compared to most common freewheels. Am hopeful it is up to the job.

As I said - I'm not necessarily committed to the 3-5 kw motor range. But I do hope to build this with enough motor that should I find the urge to upgrade the batteries with higher voltage/amp capability I won't be overtaxing the motor. I'd readily settle for a solidly built, continuous rated 1 to 1.5 kw motor. Just would prefer not to try to push smaller motors beyond their rated capacity.

Am quite open to suggestions regarding batteries. Obviously I'll be shopping for the best I can afford.

Surely a 3kw motor does not draw 100A all the time? I do not see ever actually needing the full 3kw of power (much as I didn't need the 400 HP of my Trans Am). Cannot a controller limit the current? And thus the actual power being delivered to the motor? Or are there significant trade-offs that I'm missing here? You might say that the "Tim the Tool Man Taylor" in me wants to over-spec the motor / controller enough that I can have options for future battery enhancements to the project.

I'll certainly do some reading re: Lyen's controllers. "They can run on 24/36/48" - Does that mean I can readily change batteries for different voltages voltages? Or does the controller have to be built to just one of those specs?

Is the option for a "hall sensor throttle" something specific to only certain controllers? From a practical viewpoint I could make a twist grip or lever type throttle work. If a lever type exists that is spring loaded to return to 'off' that'd be a feature I might want.

Your thoughts were helpful. As I have not yet actually begun the final assembly, I'm appreciative of all input. Especially from folks who have some working knowledge and experience with these things that I lack.

-hawk521

 

[Hawk521]

Anyway, my question is, how do you plan to deal with the cantilevered axle? Do you plan to mount the jack shaft and motor on the same angle as the axle?

Sounds like you have a good solution for the freewheel. I never took any of the drive system apart and don't remember where the original right side drive wheel's freewheel is located, but if it is at the inboard gear cluster and not out at the wheel hub your system should operate properly.

I chose to use a belt drive partly because I think it will give me some options regarding the cantilevered axle. It appears to be only about 5-7 degrees off horizontal. Will either let there be a slight twist in the final belt connection, or more likely I'll just fabricate a mount that secures the entire motor/jackshaft assembly in alignment with the left drive shaft. I'm waiting on parts to arrive so I can begin to mock up my final design options. Have not yet ordered the HTD belts as the final dimensions are not yet set.

The right side freewheel is actually integral to the inboard gear cluster. And the entire right side axle rotates the right drive wheel. The net result of my approach is that each wheel can freewheel independent of what is happening to the other wheel. If I can build the final motor/jackshaft assembly and battery pack so that each can be removed without too much work I can preserve an option to occassionally ride the trike without the extra weight of those features - everything but the final drive pulley should be easily disconnected from the trike.

I'm having fun with the conceptual and design process. Hopefully it will come together one day.

-hawk521

 

[mlt34]

More research is underway. My post is part of an ongoing search for the best result I can build. I am a mechanical engineer - so some parts of this are fairly straight up in my mind. But I have very limited knowledge of ebikes and BLDC motors and controllers. Am learning as I go.

I plan to use a BLDC motor along the lines of one that Golden Motor sells as a 5kw motor. But that one is bigger than I really need. I've seen a 3kw motor for ~$400, but even that (as you point out) may be overkill for my needs. Am still researching motors and controllers.

Yes the freewheel attaches and freewheels in the proper orientation. The hub it threads onto fits tightly to the drive shaft and is locked in place by a 1/4" square key. The freewheel threads onto the hub properly. I have a machinist lined up to help mate the freewheel flange with the final drive pulley, and hope we can mount the freewheel in center the pulley to avoid any sideways twisting torque on the freewheel. I figure the freewheel bearings will hold up best if the belt load is centered on the freewheel bearing. I bought the heavy duty freewheel from Sick Bikes for $70. It is pretty robust compared to most common freewheels. Am hopeful it is up to the job.

As I said - I'm not necessarily committed to the 3-5 kw motor range. But I do hope to build this with enough motor that should I find the urge to upgrade the batteries with higher voltage/amp capability I won't be overtaxing the motor. I'd readily settle for a solidly built, continuous rated 1 to 1.5 kw motor. Just would prefer not to try to push smaller motors beyond their rated capacity.

Am quite open to suggestions regarding batteries. Obviously I'll be shopping for the best I can afford.

Surely a 3kw motor does not draw 100A all the time? I do not see ever actually needing the full 3kw of power (much as I didn't need the 400 HP of my Trans Am). Cannot a controller limit the current? And thus the actual power being delivered to the motor? Or are there significant trade-offs that I'm missing here? You might say that the "Tim the Tool Man Taylor" in me wants to over-spec the motor / controller enough that I can have options for future battery enhancements to the project.

I'll certainly do some reading re: Lyen's controllers. "They can run on 24/36/48" - Does that mean I can readily change batteries for different voltages voltages? Or does the controller have to be built to just one of those specs?

Is the option for a "hall sensor throttle" something specific to only certain controllers? From a practical viewpoint I could make a twist grip or lever type throttle work. If a lever type exists that is spring loaded to return to 'off' that'd be a feature I might want.

Your thoughts were helpful. As I have not yet actually begun the final assembly, I'm appreciative of all input. Especially from folks who have some working knowledge and experience with these things that I lack.

-hawk521

You're right, a 3kw motor wouldn't pull 3kw all the time. Mostly just on steep hills or when accelerating from a stop. Still, you want to make sure your batteries are suited to handle that current when the time comes that it does draw that much. SLAs can handle the high current draw but will wear out quickly and you'll probably need 50 - 70 lbs of lead to get any usuable range with that current draw. Most lifepo4 batteries are only rated for 2C discharge so a 15Ah pack couldn't handle much more than 30 A current draw. Consider looking into RC lipos but begin researching all the safety issues and practices - they can be dangerous and should be respected for what they are - highly volatile bombs that can also provide useful electrical energy storage.

Also you'll need to think about how much load a bicycle chain can readily support. People usually put out around 100 watts of pedal power. On one of my early chain drives, I could power it on over 1000 watts, but if I wasn't careful to let off the throttle when shifting gears, I'd break the chain.

Correct, Lyen controllers can be run on a range of battery voltages. Sometimes you just have to open or close a connector on it to tell it which battery voltage you are using.

The ebike throttles nowadays are almost always hall sensor types, and it is so standard that they don't even mark it. There a potentiometer throttles too, but just buying any "ebike throttle" should be fine, they are all basically the same in the inside (hall sensor and magnet).

I hope this answers some questions. I'm a MechE as well, though I've found that in the world of ebikes, a year of experience with them is worth a lot more than a 4-year engineering degree

 

[Hawk521]

...
Also you'll need to think about how much load a bicycle chain can readily support. People usually put out around 100 watts of pedal power. On one of my early chain drives, I could power it on over 1000 watts, but if I wasn't careful to let off the throttle when shifting gears, I'd break the chain.

Not sure how this applies to my install. The peddle chain drive does not play any role in my motor drive plans. In this trike, the right rear wheel is driven via the 21 speed chain/deraileur system. Nothing changes about that. The motor drive will be applied solely to the left rear wheel and will have its own freewheel so as to not limit the ability to peddle the bike w/o motor assist. With the motor on, I am not sure whether 'torque steer' (from the left drive only wheel) will be a serious concern or not. Only time will tell. But with my hefty weight holding the front end securely to the road I think this won't be a problem.

My research continues for a suitable motor/controller for this project. Last evening I got side-tracked looking at one of the really large RC motors. Price was great. Performance impressive. But the more I read the more I saw stories of folks using these motors having bearing issues and in some cases the motors seemed to be simply coming apart when run very hard. Maybe I can find something similar that is built like a proverbial tank!? I just hope it doesn't come with a Pentagon sized price tag! lol

The quest goes on...

-hawk521

 

[mlt34]

Oops, you're right. My mistake. I forgot you're powering the left wheel independent. Derrr.

Have you seen this motor. I haven't used it myself but I've seen people power it up to 1000 watts successfully. It could probably take more. http://www.bmsbattery.com/central-motor/321-ebike-central-motor.html

With those RC motors, you'll want to be sure you find one with skirt bearings. Otherwise you'll have serious issues like you saw others having.

 

[Rassy]

Hawk, I don't know how steep of a hill you have had your trike on, but during pedal only tests when I had my Sun USX trike I had trouble with the single rear wheel drive on grades between 15 and 20%. Even at my weight of 200 pounds the front wheel would slide a little to the left with each power stroke which was being applied to the right rear wheel. Trying to get going from a dead stop was impossible since the front of the trike would just slide to the left. When you apply your electric assist to the left rear wheel you might experience the same issue. However, several factors may prove to make this a non issue. You may not have such steep grades to contend with, but if you do the smooth application of electrical power (as opposed to intermittent power strokes) might be more controllable. Also, you will essentially have two wheel drive which should make it possible for you to smoothly apply power to both wheels when starting up from a dead stop on a steep grade.

Just some random thoughts a little off topic, but I would be interested to know how it all works out. Some old threads talked about setting up trikes with both wheels driving for use on snow and ice, but the Sun USX trike was never a candidate because the cambered rear axles couldn't be replaced with a single straight axle. Your approach might provide a reasonable alternative.

 

[Hawk521]

Have you seen this motor. I haven't used it myself but I've seen people power it up to 1000 watts successfully. It could probably take more. http://www.bmsbattery.com/central-motor/321-ebike-central-motor.html

With those RC motors, you'll want to be sure you find one with skirt bearings. Otherwise you'll have serious issues like you saw others having.

If this motor could sustain 1000 watts for say 8-10 minutes at a time I might could use this. Will do some research to see how well it has been holding up for folks. Does this one have the 'skirt bearings' you mentioned?

Question: Can I use this motor (or any typical 1kw BLDC motor) with a controller rated for 24v to 72v? Or does it depend on the controller? Even if I opt to use a smaller motor such as this one, I'd like to be able to choose a controller that would make it possible to migrate to a bigger motor in the future if needed. While it makes sense to somewhat match the motor and controller capabilities, I am hoping someone makes a controller flexible enough to handle say anything from 24v to 72v and 750 watts to 2000 (or 3000?) watts. If I can do this it makes it simpler to upgrade to more motor power when/if the time comes.

I'd like to start my testing with a simple 24v setup. Get some cockpit time with that and maybe step it up to a 36v configuration at some point. Then maybe even 48v. That's about as high in voltage as I have any aspirations. But if every change necessitates dumping the old motor AND controller it just sort of doubles up on the expense.

-hawk521

 

[Hawk521]

Hawk, I don't know how steep of a hill you have had your trike on, but during pedal only tests when I had my Sun USX trike I had trouble with the single rear wheel drive on grades between 15 and 20%. Even at my weight of 200 pounds the front wheel would slide a little to the left with each power stroke which was being applied to the right rear wheel.

I have noted that the steering can be slightly affected by a strong power stroke on the peddles in low gear. But never enough to actually slide the front tire. Perhaps my 285 pounds keeps it planted on the road.

The roads here are probably not more than 5-7% grades. But some of them are LLLLOOONNNGGG climbs. One in particular is about a quarter mile climb. Total altitude change is about 80 feet according to my car's GPS (gps altitude is not terribly accurate, but it averages around a change of 80 on this hill).

Tonight I'll be studying BLDC motors in the 1kw-3kw range - hoping to find just the right one. Probably a rating between 100kv and 200kv could work for me.

My pulleys arrived today. They look impressive. The smallest two and the largest are made of iron! The other large one is aluminum. I wish the biggest one was aluminum, but alas, I'll see if I cannot talk one of my buddies with a lathe into taking a few pounds off that iron monster. I haven't weighed it, but I'm guessing the 80 tooth iron sprocket that will mount on the axle probably weighs 8 to 10 pounds as it exists. And at least 75% of that is simply not needed. Whether it will be practical to shave that much off of it remains to be determined. Sure wish I had my own lathe to play with!

-hawk521

 

[mlt34]

The Lyen controllers can be run on a range of voltages. I had one of his 6 fet controllers that could be run on 24-60 something volts. I used it on 24, 30 and 36 volts for some testing of different battery setups. I'm not sure the ultimate peak power it can take, but I'd guess it is at least 1500W if not more.

The skirt bearings I mentioned are for the RC motors. The cheaper ones have only a bearing at the place where the output shaft exits the body of the motor. That causes problems when you try loading it perpendicularly to the axis of the axle. For RC plane props its fine, since loading is just torsional, but ebike loading would be perpendicular to the long axis of the axle and that is a problem for motors without skirt bearings.

The motor I linked to I haven't actually used myself, so I'm not sure what it looks like on the inside and what its bearing situation is.

 

[mlt34]

btw, it appears that motor I linked to may be the same motor used in this ebike, which apparently is a company that is either a fraud or became one when they couldn't meet production. http://endless-sphere.com/forums/viewtopic.php?f=3&t=42010

 

[Hawk521]

Found some interesting reading on the Astro 32xx line of RC motors last night. The 3215 might just be ideal for my needs - everything about it (except price) appears to fit my needs. I guess you gotta pay to play these days, but $600 for a motor (plus I'll need a controller and battery pack) is beyond my project budget.

1. I would appreciate thoughts on this motor (or its 3210 & 3220 cousins) from folks who have experience with them. Are they worth the investment? And am I thinking wrong when I consider buying a motor that is rated beyond anything I'll ever ask of it? It just seems to me that running a 2kw to 3kw motor with nominal <1kw demand would surely result in fewer motor issues over time. Or is that just wishful thinking? Or is it bad design? Do I need to run a given motor near it's optimal design power to get good efficiency?

2. Are there less expensive, but similar capability, impressive motors that I ought to be considering?

I read about Matt Shumaker building a big brother (3230) motor which is not only more powerful, but has better bearings, cooling, and looks cool with the fins molded into the housing. But I did not find whether this motor ever came to market, nor what it would cost? Is there an underground network that sells these? <insert secret handshake here>

One concern I do have is putting 'too much' motor on the trike. With the left rear wheel drive, I suspect a motor like Matt's 3230 would probably lift the front wheel and spin the trike around like a top if I dropped the hammer on the throttle. Would certainly be entertaining for bystanders, but I'm not sure I'd want to be the one riding that bronco!

Too much fun. The search for a proper motor and controller goes on.

And while I'm asking - is it practical to use LiPO batteries? I have a vague understanding that they can be dangerous and subject to overheating and causing a fire. What sort of circumstances cause this type of catastrophic failure? Does it happen while a load is being put on the batteries? Or is it a matter of overcharging them (or charging too fast?)
The LiPO technology does appear less costly than some alternatives. Is it a risk that can be managed safely?

I do apologize if I'm asking questions that I might be able to ferret out answers to by reading forums for the next few months (or years). There is a LOT of stuff out there! And I am searching and reading as much as time permits.

-hawk521

 

[Hawk521]

Motor and controller are on order.

At Edward Lyen's suggestion I took a look at the http://www.thesuperkids.com/15wabmcbrmos.html 1500 watt motor. Quite likely all the motor I will ever need. And I am betting that it is lighter than the 11kg (24 lbs) Golden 5kw motor. While the Golden motor would have been quite a toy to play with, I think this more practical motor will better suit my actual needs, and the more modest performance might even save me from myself. I do know from experience that having way too much motor in a vehicle is a temptation that I have not always been able to resist. So I ordered this 1500 watt motor on Monday.

Also ordered the 18 FET controller that Lyen suggested. The plan is beginning to come together.

Now attention turns to finding a suitable battery pack. Have raised my goal a bit and decided to use a 48 volt pack. I have been internet shopping. LiPO, LiFePO4, Li-Ion, and ??? So many options. Is LiPO really as dangerous as it was once felt to be? I'm not keen on burning my house down, but the LiPo batteries are tempting for cost and energy density. I can see that this decision warrants some due diligence so I'll do more research.

Slowly but surely I'm gaining on the project. Hoping I can get it done w/o busting the budget too severely.

-hawk521