Mid Drive with Internally Geared Hubs

[LI-ghtcycle]

I can't speak to the N360 NuVinci Hub, but I have used the original heavier built N171B in this bike:



(link to build in Sig)

With 72V (18s) of 45 - 90c Turnigy Nanotech Lipo, and although cruising I use around 250 - 500W, I climbed a 5% grade hill going 25 MPH + and pulled 1400W continuous for about 1/2 mile with no issues. I haven't weighed the bike, but I know it has to be at least 80-90lbs, I weigh 180 lbs, so pretty impressed with the Cro Motor and Lyen Controller, nothing even gets warm.

I've read where gas bikes are putting crazy amounts of torque to the N171B and things only start slipping at 90 ft/lbs torque with no damage, just darkening the color of the fluid inside, so I imagine that the N360 should still hold up to the average E-Bike, we see them on so many mid-drive "turn key" E-Bikes, I am sure they would hold up to around 500W.

 

[Joe T.]

My N360 failed on my bike with a totally stock GNG with LR upgrades. It was working fine for a few months. Then it started to "thump" in low range. It still worked but the issue was getting worse. They sent me a new one and I have about 180 miles on it now.

The cause of my failure was not apparent but my theory was slow jarring roots on some of the trails I ride. I am now cautious in these types of trails. Otherwise I like the N360 a lot.

 

[LI-ghtcycle]

My N360 failed on my bike with a totally stock GNG with LR upgrades. It was working fine for a few months. Then it started to "thump" in low range. It still worked but the issue was getting worse. They sent me a new one and I have about 180 miles on it now.

The cause of my failure was not apparent but my theory was slow jarring roots on some of the trails I ride. I am now cautious in these types of trails. Otherwise I like the N360 a lot.

Thanks for posting Joe T.

I am curious what the NuVinci says about how much power their N360 should handle?

IIRC, it was originally somewhere in the 350W constant range?

 

[Joe T.]

Thanks for posting Joe T.

I am curious what the NuVinci says about how much power their N360 should handle?

IIRC, it was originally somewhere in the 350W constant range?

I was a little vague on my application. I told them I was climbing steep hills with my mountain bike. I felt I left the door open for more discussion if they were interested. I was not about to give them an excuse to get out of a warranty. I think the NuVinci concept could make a great home in an Ebike. The weight seems to be an issue for the spandex wearing bike crowd. I don’t see it as an issue with the added electric power of an Ebike.

 

[Chalo]

The NuVinci hubs have two different specified limits that both apply: There's a gross power limit, and an input torque limit.

If I remember correctly, the gross power limit on the N360 is at least 2000W. But if you exceed the maximum input torque, you're apt to break the thing even if the power is well within that amount.

 

[Miles]

These were the limits given in the N171 Developer Kit:

Torque, Power, and Speed Limits*

Maximum Sustained (Pulse-free) Input Torque 65 Nm (48 lb-ft)
Maximum Instantaneous (Spike) Input Torque 130 Nm (96 lb-ft)
Maximum Sustained Input Power 5 kW (7 hp)
Maximum Recommended CVP Speed Input: 1000 RPM / Output: 1000 RPM
*Operation at these limits is not recommended for extended durations, and durability is highly dependent on duty cycle. Higher RPM, torque and power configurations can be designed for production applications.

Not sure how the N360 compares...

 

[Joe T.]

Not sure how the N360 compares...

The instructions want a maximum of 2-1 ratio as in a 36 drive into an 18 wheel cog. A lower ratio, 1-1 would over torque the drive I assume. I have a 36 into 18 setup. I suspect I over torqued the NuVinci while powering through a rooted up section. I have not been gentile on my replacement N360 with the exception of slow rooted trails. I have climbed the same hills as before.

A friend has broken 2 N360s. Both times it was in near stall condition while going slow up bumpy trails.

 

[spinningmagnets]

Time ha sshown that, the more power you have, the fewer gears you need to be happy. There is a huge difference in the construction of the common 8 and 11-speeds, compared to the classic Sturmey Archer AW design. The more I read about them, the more I understand how they are MUCH stronger than the others.

Second gear is the strongest. Two large pawls (of a "rocker" design) are allowed to move outwards and engage the hub shell directly (Nexus uses four smaller pawls, but they operate the same way). In second gear the input driver does not send power through the planetary gear set at all.





Many other hub designs have separate sun gears that spin freely on the shaft, and then a "dog" rises up or slides over to grab splines at the center of the gears, in order to lock or unlock them. This means the forces concentrate a lot of leverage onto the parts that are near the axle. In second gear (above) the connection takes place much farther from the axle.



For third gear, the four radial-arm clutch is pushed all the way inwards by a coil-spring (as the shift-cable is relaxed), and the four-armed clutch grabs the protruding pins that act as two parts, the clutch dogs, and also the axles of the four planet gears. In third gear, the planets are driven around a fixed sun gear, and it is the ring gear inside the shell that provides the overdrive output.





For first gear, the clutch is pulled all the way outwards, so it drives the planetary shell (which holds the ring gear), and the power output is taken from the pins of the planet gears. Sun gear is fixed, ring gear is driven, and the four planet-pins drive the output.

 

[Chalo]

Not sure how the N360 compares...

The instructions want a maximum of 2-1 ratio as in a 36 drive into an 18 wheel cog. A lower ratio, 1-1 would over torque the drive I assume.

The recommendations presume human power. When you add motor power to human power, you have to increase the primary gear ratio to keep from feeding in too much torque.

For my Rohloff hub bike, I chose a primary ratio 10 percent higher than the minimum ratio specified for tandem bicycles, because I'm heavier than many tandem teams and I don't want to break my expensive hub.

 

[spinningmagnets]

One way to increase the power capability of a given hub is to increase its RPM's in relation to the wheel RPM's, which can be accomplished by using the IGH as a jackshaft.

One of the Tangent Ascent ebikes Uses an IGH as a jackshaft.

Here is a pic of the Italjet #59. The BB-to-IGH ratio looks to be somewhere around 4:3, and the final drive chain appears to be 1:1. If you increased the BB-to-IGH ratio to around 8:3, and the final drive to 1:2...then the forces acting on the IGH should be cut in half (or inversely, you can input twice the power without breakage?). The rear wheel RPM's and pedal-cadence would stay the same, but the IGH would spin twice as fast.



IGH's that are advertised as being made for a trike, pedicab, or quadricycle often come with dual sprockets (Nexus, Sturmey, etc). One on the freewheel driver, and one attached to the hub shell. http://www.sturmey-archer.com/en/products/rear-hubs

 

[cjh]

Spinningmagnets, do you think the SRAM Automatix 2-speed automatic hub is strong enough to use as a mid-drive jackshaft? What do you think its limit might be?

 

[Miles]

Spinningmagnets, do you think the SRAM Automatix 2-speed automatic hub is strong enough to use as a mid-drive jackshaft? What do you think its limit might be?

They are direct drive and overdrive so, not well suited for motor drives, for most peoples' needs....

http://www.sheldonbrown.com/automatix.html

 

[spinningmagnets]

Yeah...what he said

 

[cjh]

Thanks guys, looks like the 3-speed is the answer then.

I have never done this before so how do I do the calculations to work out what gearing I need where, when using a 3-speed IGH as a jackshaft?

 

[spinningmagnets]

The higher the top-speed, the worse the hill-climbing. First, choose the lowest top-speed you can be happy with, then decide on a tire diameter. The Sturmey classic gearing is a 33% underdrive/overdrive for 1st and 3rd (the Nexus is 36%?).

On a Sturmey, if the speed of the "direct drive" second gear is "X", then the first-gear is X times 0.75 (tooth-count ratio of 3/4), and third-gear is X times 1.33 (4/3)...as an example, if a pedicab wanted to emphasize torque instead of top speed, X might be 12-MPH, so...first gear would be 12 X 0.75 = 9-MPH, and third would be 16-MPH (for the Nexus, X is direct drive, third is X times 1.36, and first is X times 0.77).

9, 12, 16-MPH for the Sturmey, the Nexus would be approximately:
8, 12, 17-MPH

Without the weight of pedicab passengers on a conventional ebike, you can go for a higher speed. The California legal speed limit for an un-licensed PAS ebike is 28-MPH, so...the goal for the Sturmey and Nexus would be:

15, 21, 28-MPH for the 33% Sturmey, and the 36% Nexus would be roughly:
13, 20, 28-MPH

 

[cjh]

Thanks spinningmagnets. If I want to use a Cyclone and run 24" wheels, I know I need 420rpm at the wheel to run 30mph. How do I work out the sprocket sizes through the system?

EDIT: If I understand correctly I would need to be spinning the IGH at 840rpm in 3rd gear to reduce load on the IGH?

 

[cycleops612]

Terrific post. Fascinating.

 

[tommie]

hi guys, i`m in the process of installing an Alfine 8 speed on my BBS02 750watt conversion, are there any gears i should go easy on the power in?

 

[cycleops612]

Shame there isnt a gear shift sensor which tells the controller to slowly turn chain with ~zero force for next, say, second, to allow smooth changes, then resume throttle power gently.

 

[spinningmagnets]

With enough sudden power, you can break anything, so "ease on" the throttle. tommie, just be aware that for the Alfine-8...5th gear is the strongest.

It is a 2-speed coupled with a 4-speed. The 2-speed is either direct drive in high range, or a 1.53:1 reduction in low range.

 

[spinningmagnets]

The first step in designing a mid drive system is determining the top-speed you want. Then choose a tire diameter. Here is a chart to use those two data points and get an RPM result. Then you can calculate input/output sprocket sizes.

"Wheel diameter to speed/RPM chart"
https://endless-sphere.com/forums/viewtopic.php?f=28&t=16114

 

[brucemetras]

Shame there isnt a gear shift sensor which tells the controller to slowly turn chain with ~zero force for next, say, second, to allow smooth changes, then resume throttle power gently.

You could use a commercially available gear sensor and program (assuming Bafang mid-drive) more or less delay and ramp up to suit your needs..

http://www.lectriccycles.com/gear-sensor/

 

[cycleops612]

Shame there isnt a gear shift sensor which tells the controller to slowly turn chain with ~zero force for next, say, second, to allow smooth changes, then resume throttle power gently.

You could use a commercially available gear sensor and program (assuming Bafang mid-drive) more or less delay and ramp up to suit your needs..

http://www.lectriccycles.com/gear-sensor/

come to think of it, many achieve similar by twitching the brake lever as they change - the brake sensor pauses the motor.

 

[WoodlandHills]

Or just backpedal for 1/2 inch to kill the power.

 

[brucemetras]

Most of the time, I just get my cadence close to the motor no load rpm and shift normally.. if you're running an inline amp meter, you can also glance at it and shift when amp draw is low (like 1 or 2 amps).. and shift normally .. but most of my riding is onroad, so a bit easier to keep cadence up..