The E-cumbent - A project by Matt Shumaker

We usually express things here in terms of Watt Hours in the battery and peak Watts from the motor. The efficiency ends up being the difference between what the battery provides and what the motor delivers.
Click to expand...

Indeed. Since so many folks use different voltages, it's better to state measurements in watts and watt-hours. Also, the most basic and universal efficiency measurement is Wh/mile. Now much energy, measured at the battery, did it take to go one mile?
 
Hi Recumpence,

Your build is looking awesome - Illinois you say? shame, i'd love you to do some work on my bike :)
The setup you made on the cnc/mills is absolutely awesome - im in total shock at the skill here as i along with many others have to build in a modular fashion.
I dont really understand how such a small motor can give such amzing torque and handle such power when we have colossal motors (hub) in comparison? also the size of the controller gear is simply amazing :shock:
Im just about to drop some more money into my own bike but this whole thread has me thinking that maybe i should hold on and see how yours pans out :) it is astounding that your build seems to undercut the weight and overdo the power on every aspect of the bike - i guess i as well as lots of others here are wondering if we could transfer to this kind of kit.
I wouldn't know as i've only been here a year and a half or so but do the more experienced guys here see potential for a serious leap for all our bikes with the tech Matt is running??


Cheers


D
 
Ben said:
Convert to metric I say. Imperial is dead.
Click to expand...

When US traffic law and law enforcement switches to metric, I will use kilometers. Until then, miles.

Though, things could be much worse. We could all be using the FFF system.

Last Wednesday, I was a little liberal with the throttle while running some errands. Turns out the 5303 at 84v in a 20" wheel has some go. I did 1.19e5 furlong/fortnight in a 1.08e5 zone! :wink:

Oh, a resident of Illinois? Which part?
 
deecanio said:
I dont really understand how such a small motor can give such amzing torque and handle such power when we have colossal motors (hub) in comparison? also the size of the controller gear is simply amazing :shock:
Im just about to drop some more money into my own bike but this whole thread has me thinking that maybe i should hold on and see how yours pans out :) it is astounding that your build seems to undercut the weight and overdo the power on every aspect of the bike - i guess i as well as lots of others here are wondering if we could transfer to this kind of kit.
I wouldn't know as i've only been here a year and a half or so but do the more experienced guys here see potential for a serious leap for all our bikes with the tech Matt is running??
Click to expand...

The torque from the motor is very small - that's why it has to be run so fast and then geared down.

Some neat machining aside, there's nothing particularly groundbreaking in Matt's set-up. What we need is some detailed data on efficiency - not just a general Wh/Mile. Then we can find the best compromise between motor weight and drive train efficiency/complexity.
 
Miles said:
lazarus2405 said:
When US traffic law and law enforcement switches to metric, I will use kilometers. Until then, miles.
Click to expand...

Oh, I thought this was an international forum :)
Click to expand...

I meant with respect to my personal usage. But yes, I do agree that using real units to measure efficiency is indeed better.

What we need is some detailed data on efficiency - not just a general Wh/Mile.
Click to expand...

Agreed. But Wh/mile is better than the very limited data Matt has posted so far. "x amps" doesn't tell us a single thing! :wink:
 
lazarus2405 said:
I meant with respect to my personal usage. But yes, I do agree that using real units to measure efficiency is indeed better.
Click to expand...

I was only joking :) We're still half metrificated here in UK [Kilos and MPH :roll: ]

Agreed. But Wh/mile is better than the very limited data Matt has posted so far. "x amps" doesn't tell us a single thing! :wink:
Click to expand...

Yes, I'm with you :wink:
 
The Efficiency of Gears

In order to take 10,000 rpms down to the 100 or so rpms that the pedals use you need to step down the speed with gearing. The efficiency ends up being a direct way to judge the precision of his machining techniques. If his precision is really excellent and he used high precision bearings (which I'm sure he did) then it should be possible to get this geared down without much in the way of losses. A good gear to gear transfer is in the 98% or better range. Running a brushless motor at higher speeds is also something that tends to increase efficiency.

Taking a step back... what this might mean is that "we" (the ebike public) might want to think about these high rpm motors and doing this kind of step down idea ourselves. If you can step down a 10,000 rpm motor by 25 you are down to 400 rpms which is in the ball park for most of the things we do. Once you get to 400 rpms output then you can use the chain and gearing to go that last step.

My Unite motors have built in geardowns that are about 9 to 1 and so when I overvolt the motor up to 4000 rpm to 6000 rpm the output ends up being in the 450 - 650 range which is managable.

Geardowns are inevitable of you want to integrate pedaling... the hub motor is "nice" but weighs 25 lbs in order to do it's job. The small high speed motor with geardown can be as light weight as 5-10 lbs.

:arrow: What I want to see is a 20 to 1 planetary geardown which would be the ideal if done with incredible precision.

You can buy this stuff right now, but they tend to be built for industrial purposes without regard for weight. We need a lightweight ebike version of the same idea.

10.jpg


Multispeed rear hubs are essentially planetary geardowns (often multiple layers of them) so the idea is there and it's just a matter of someone to apply the effort to develop the product. High precision and small light weight parts means that you need to destroy a few before you can find that lightest weight possibility to do the job.

http://www.cyclone-usa.com/store.php?crn=203

This is the standard gearbox that is used on the 360 watt and 500 watt motors. It provides a 9.55:1 reduction.

$124.67

300_large_image.gif
 
safe said:
In order to take 10,000 rpms down to the 100 or so rpms that the pedals use you need to step down the speed with gearing. The efficiency ends up being a direct way to judge the precision of his machining techniques. If his precision is really excellent and he used high precision bearings (which I'm sure he did) then it should be possible to get this geared down without much in the way of losses.
Click to expand...

safe, I really don't understand this...
 
Miles said:
safe, I really don't understand this...
Click to expand...
The "core" idea of his design was to take a high speed (10,000 rpm) RC airplane motor and step it down with gearing to a speed that can be used for an ebike.

This can be a problem if your precision is poor because gears that are slightly misaligned tend to produce friction and that means losses.

If he used high enough precision it's possible to drop the speed that the final chain sees to a level that is managable.

This is one of my Unite motors with built in geardown:

106115.gif
 
Miles said:
But he's not using gears he's using a 2 stage belt reduction!
Click to expand...
Ooops! My mistake then... I thought he was using gears. Well, chains and belts can also be very efficient too. The ideal is the planetary gear setup though.

I think the main thing is the idea of harnessing the 10,000 rpm motors... something that people had dismissed as impractical before. That's the "breakthrough" that he's made here.

Hmmmm... belts huh... :oops:

motor11_sm.jpg


done7.jpg
 
So the precsion on his design comes to how well aligned the axles are and how well the precision bearings are set into the frame as well as the losses due to belt tension and misalignement.

I found these estimates:

Helical gear pair - 95% to 98%
Planetary gear train - 94% to 97%
Bevel gear pair - 95%
Belts - 93 to 95%
V-Belts - Less than 93%
Worm gear pair - 55% to 88%

http://promot.cres.gr/promot_plone/pumps/overview/motor-and-transmission
 
safe said:
So the precsion on his design comes to how well aligned the axles are and how well the precision bearings are set into the frame as well as the losses due to belt tension and misalignement.
Click to expand...

As all the machining is to a single structure, that shouldn't be much of a factor, and precision bearings won't make a huge difference to efficiency, I would have thought. I think the diameter of the idler wheels is a bit small, perhaps? The back bending will certainly reduce the life of the belt.

I should imagine that the efficiency losses on the whole drivetrain (not including the motor) are between 10% and 15%, but it would be good to have some actual measurements.
 
Doing the Math

:arrow: If the transmission has two geardown pulley's and one chain that's:

93%
93%
98%

So if we start with 1000 watts of power into the motor and we get 900 watts out (90% motor efficiency) then we get:

900 * 0.93 * 0.93 * 0.98 = 763 watts at the rear wheel.

:arrow: Compare that to:

Cheap Unite Motor with single geardown and chain:

95%
98%

So if we start with 1000 watts of power into the motor and we get 800 watts out (80% motor efficiency) then we get:

800 * 0.95 * 0.98 = 745 watts at the rear wheel.

...in the end you lose a lot of that extra motor efficiency because of the extra geardown losses. :cry:
 
You compromise to get a lighter weight system.

We need some actual measurements, so that we can see which is the best compromise for a given design objective.
 
Miles said:
We need some actual measurements, so that we can see which is the best compromise for a given design objective.
Click to expand...
I think he's pumping out so much power to begin with that even with losses it's a lot of power. Didn't he say something like 3000 watts of peak power?

3000 * 0.90 * 0.93 * 0.93 * 0.98 = 2300 watts at the rear wheel.

:arrow: Which is still three horsepower! :shock:
 
The Motor's Secret

:?: What is it that makes these motors so good?

It's the same thing that makes the Etek, PMG132 and PMG080 motorcycle motors so good... they have lower internal resistances.

Most hub motors and cheapo Unite motors have internal resistances of:

0.30 ohms

...but the better motors have:

0.030 ohms

...or somewhere near that, a factor of 10 times better. Since all the calculations for efficiency boil down to internal resistance it's this lower resistance that makes it better.

That's the "secret".

So ideally a motor could be built using the same logic of the PMG132 types and not require such high rpm's.

What's needed is better motors for our specific needs... :?
 
As a non-engineer I have two questions:
1. How can a motor that small, running a 10,000 rpm have a long life? In RC it's running for minutes without much load (?), whereas driving an ebike for thousands of miles...?
2. Could the cost of fabricating that fabulous powertrain cage be reduced by precision casting it?
 
You must log in or register to reply here.

Similar threads

E
    • Like
    • Cheers
rideable garden model e-train project thread
Replies
8
Views
2,048
slaphappygamer
S
F
E-bike project
Replies
5
Views
964
amberwolf
amberwolf
Josh K.
Josh K. Precision E Bikes Service station, Electric Bike Review
Replies
1
Views
1,749
Josh K.
Josh K.
M
Can I use my 3-phase e-bike motor as a generator with rectifier?
Replies
5
Views
1,479
crossbreak
crossbreak
R
Various E-bike Motor Performance Comparisons recorded with Komoot, Strava and Garmin
Replies
4
Views
2,306
docw009
docw009
Back
Top