NuVinci CVT Vs. Nexus 8 speed

[Rassy]

After reading LIght-cycle's review of his first NuVinci application I started thinking about replacing my Nexus. I've had the trike for over 2 years, with the mid-motor for the last year and a half. The Nexus has performed pretty good, but it does not like to shift under power. So for every shift that means stop pedaling and get off the throttle for a few moments. Taking off from a stop sign it seems like I have to keep regaining speed that is lost during the shift. It's even worse when down shifting on a hill, and I normally drop two gears at a time to make up for the loss of speed while shifting. Remember, I'm an old guy in my 70's so not quick reacting like you young guys under 60.

This week all my parts finallly arrived. The double wall rim got back ordered for a month. Then the spokes I had JRH make for me were about 2mm too long because the ERD of the rim was less than specked. I ended up using the spokes on a cheap single wall rim I already had, and still had to add a cheater washer under the head of the nipple. Plan to ride it this way for a while, but will order a new set of spokes from JRH and relace the NuVinci in the better rim. By the way, John's spokes are great and they arrived within a week of ordering.

I'm not using the developer's kit, but the standard manual shift version N171B. Haven't ridden it yet because the shifter on the right side axle was not compatible with my chain tensioner, so the chain hangs almost to the ground right now. Tomorrow's project will be to shorten the chain and whatever else is needed.

Surprising though I can run the motor WOT with the wheel up with no chain issue. So I collected some amperage data just before removing the Nexus and then repeated today with the NuVinci. If the readings seem high to you, remember this isn't the motor's unloaded WOT amps, since in addition to turning the wheel through the IGH it is also pulling a tripple length chain.

All tests were with the same battery with no charging done. The voltage stayed real near 53 1/2 volts during each test with the motor turning at about 270 RPM.

Nexus in 1st gear: 1.15 amps
Nexus in 8th gear: 1.5 amps

Nuvinci in lowest ratio: 1.0 amps
Nuvinci in highest ratio: 1.65 amps

When I get some real rides in I'll report back. One thing that will take some getting used to is the amount of turning required to shift from low to high. I didn't measure it but probably about two full turns, whereas the Nexus only required about 1/4 turn from low to high. On the other hand it should be real easy to fine tune the gear ratio for best efficiency.

 

[johnrobholmes]

Can't wait to see how it compares!


The next time I tear down my spoke machine and change die sets, I will be sure to add an extra 2 or 3 mm of length to the rolling area. It will take some grinding and modifying of the machine, but having just a few more threads would be nice I think 8)

 

[Rassy]

John, I have zero complaint on the spokes. Even if I had the rim in hand before ordering I probably would have ordered wrong. I learned enough to make it worth it.

I used the single wall rim because the eyelets in the double wall made it harder, plus I couldn't get the washers in the double wall.

 

[mwkeefer]

+1 on the can't wait - I have Nexus 8 speed on a bike, 2 3 speed SA Inter 3 Silent custom builds and a Nuvinci CVT sitting on a shelf waiting to be built into a wheel, your tests should help me determine if the weight offset will be worth it - the CVT weighs nearly as much (if not more) than a small, nylon geared ebike hub motor so it's up in the air if the trade off will be worth it on my platforms but your tests will be very helpful in pointing my future actions P

TIA!

Regards,
Mike

 

[Warren]

Guys,

I have considered a NuVinci, because of the automatic shifting feature. But for manual shifting, up to 1500 watts, nothing beats a derailleur for weight, efficiency, and price. If you are not running a cargo bike, a single chainring, and a short cage derailleur, shifting an 11-34 cassette, works flawlessly. I switched to the short cage derailleur, after deciding E-power was reliable enough that I wouldn't need my small, bailout chainring. It controls chain tension better, and makes for quicker shifts. On 350-500 watts you can run wide open all the time, and just shift up and down under load. Above that power you would be better off backing off on the throttle momentarily, while still pedaling, during the shift.

I've got 1197 miles on mine so far. I've lubed the chains twice in that distance, and check the chains for wear with a wear indicator gauge. About two thirds of their life still to go. Shifts flawlessly too.

http://endless-sphere.com/forums/viewtopic.php?f=28&t=23259

 

[Rassy]

Okay, I took a little test ride, and I'm not happy with the NuVinci at all.

Pulled out of the driveway and turned up the hill which goes to a 15% grade within about 50 feet.

First thing I found was that I couldn't turn the shifter under power on anything over about 10% grade. I could neither up shift nor down shift until I let off on the throttle, and then the shifter turned easily. I was not expecting this to be an issue. Has anyone else experienced this problem? What about with the auto shifter, does it shift smoothly under power on a 15% grade?

I'm really bummed out about this, because being able to shift under power was the number one issue with the Nexus which I was hoping to resolve with the Nuvinci. Unless I've got something wrong the Nexus will soon be back on the trike.

Another issue cropped up, that I call freewheel slip. When I first put the mid-drive on the trike the Shimano freewheel I used on the motor started slipping frequently on 15% grades. I had to replace it after about 30 miles. Even though I have a track cog on the motor now, there would still be an occasional slip in low gear on a 20% grade or on a lessor grade if I didn't shift down. I always figured this was from the internal freewheel in the Nexus, which is one reason I baby it and never even try to shift under power. If I wasn't unhappy with the shifting on the NuVinci I would remove the external freewheel, which I think is slipping cogs, and just let the NuVinci internal freewheel do it's thing. But right now that would just prove whether or not I am right about where the slip is coming from, and I really think I'd rather have the Nexus than the NuVinci.

Right now I plan to just let it be for a while and see what thoughts come in from the forum. Is there any way to tell for sure if it's the freewheel that slips?

 

[johnrobholmes]

When a freewheel slips it normally makes some big racket.

How much power are you running?

 

[Rassy]

I'm using the controller I got from you along with the Bafang several years ago. It tops out at about 23 amps, so with the 48 volt system something a little over 1100 Watts.

John, if you haven't cut those new spokes yet will you hold off. I'll send you a PM and check if I can send a message through Holmes Hobbies. If they're already done, no problem.

 

[johnrobholmes]

Depends on when the order was placed. I haven't cut any in a few days because of some previous backorders, so if you placed the order yesterday or late wednesday they aren't rolled.

 

[Rassy]

JRH said:

When a freewheel slips it normally makes some big racket.

Yep, it is a loud clank. The last trip up the hill to my house a year and a half ago with the bad Shimano on the 9C sounded like an old Ford that had thrown a rod. Same sound, but just one at a time now.

What I don't know is what would the sound be like if the chain was jumping a cog? I don't think this is what is happening, but just wondering how it would differ since I can't see that area when riding. The 9C has a good looking 16T track cog and the Nuvinci and Nexus both have 18T cogs. There is plenty of chain wrap on all of them.

 

[Rassy]

I said:

There is plenty of chain wrap on all of them.

I was thinking about this issue and realized I might be able to duplicate the problem in the shop.

Hoisted the rear of the trike so the rear wheel was off the ground.

First test I put the NuVinci in high gear and with full throttle applied the roller brake. I was able to smoothly stall the motor.

Next test I put the NuVinci in low gear and with full throttle applied the roller brake. And just before stalling, bam. I did this test several times, and even sitting right next to the motor I couldn't see the chain slip, but that's where the noise came from and the motor gave a little shake, probably releasing some tension on the motor brackets.

The Watts-Up showed 1288 Watts Max and 27 Amps Max after the tests. I never see over 21 or 22 amps after any of my rides so it was pulling hard. The fact the trike wasn't rolling with my weight on it probably affected the test results a bit.

My next plan is to lengthen the chain and move the idler that follows the motor sproket to substantially increase the wrap on the motor sproket.

I have committments for the next couple of days but will report the results when I can. I hope this is the cause, since I have blamed it on several different freewheels, etc.

 

[dumbass]

I've had a 171 manual shifter for several years. Nuvinci told me it would take about 500 miles to wear before it would shift under load. I can tell you now that it will never happen. I can generally down shift under load but upshifting under anything but the lightest load can not be done.

Acording to what Nuvinci told me the 171 does not have a freewheel the way we bikers would normally think of a freewheel. They advised me to always use an external freewheel.

Bob

 

[Rassy]

Thanks for the comments Bob. I have decided to go ahead and try one of the developer kits to get away from the manual shifting issues. My guess is some conditions will still require backing off on the pedals and/or throttle.

I passed on the auto shifting initially because I have briefly been on a couple of those auto shift bikes that move the rear deraileur using centrifical force from weights built into the rear wheel. Didn't like the automatic shifting at all, even when expecting it.

My increased wrapping on the motor sprocket did not correct the slippage problem, so I am back to blaming the freewheel. I've ordered an 18 tooth track cog to replace the freewheel, but have already seen that getting the freewheel off will be a problem. I've ordered a tool to help, but it will still be tough. May not get on to this until after the first of the year.

This thread was meant to compare the Nexus 8 speed and the NuVinci, so here are a couple of initial observations:

Neither of them like to shift under power, but the result is way different. On the Nexus, you can easily shift to another gear, but if under power it won't always shift immediately, in which case it will "klunk" when it does complete the shift. On the NuVinci you simply can't turn the twist shifter until the power is backed off, but then when you can shift it is smooth.

The Nexus shifter twists very easy between the eight gears, and the amount of twist between 1st gear and 8th gear is just a fraction of a full turn. The Nuvinci is smooth but takes a tighter grip to turn the twist shifter. From lowest ratio to highest ratio takes a couple of full turns. Just working it from low to high about a dozen times made my wrist sore. This is why I'm going to try the auto shift kit.

I ordered the Nexus with Bob "nuts", which is all that is needed to hook on a Bob trailer. To hook a Bob trailer to the trike with the NuVinci would mean making some sort of custom bracket for each side.

And of course there is the size and weight. The NuVinci is bigger and heavier than some geared hub motors. The Nexus is relatively small in diameter, and I think weighs between one and two pounds.

As mentioned above I'm still trying to figure out the cause of the "clunk". If it does turn out to be the external freewheel on the NuVinci, then I would conclude the Nexus internal freewheel doesn't like over 1100 Watts. But right now I don't know that I'll reinstall the Nexus to try to prove this.

 

[Miles]

The Nexus is relatively small in diameter, and I think weighs between one and two pounds.

Between one and two kilograms....

1.68 kg, to be precise - which is 3.7 pounds.

 

[liveforphysics]

If I remember correctly, doesn't the nuvinci throw away like ~20% of the power you feed it?

If you have an 85% motor feeding an 80% hub with a 97% efficient controller, a full third of your battery just goes into heating drivetrain components, and that would be ina perfect battery with no sag.

 

[dumbass]

If I remember correctly, doesn't the nuvinci throw away like ~20% of the power you feed it?

If you have an 85% motor feeding an 80% hub with a 97% efficient controller, a full third of your battery just goes into heating drivetrain components, and that would be ina perfect battery with no sag.

Without question the Nuvinci eats eff. for lunch. But the fact that you can always better fit your ratio to your needs in my opinion has made up for most of my losses. The biggest complaint I have about the 171 is the weight. I trasport my bikes on a car rack (a big heavy car rack). And picking up a heavy framed bike was bad enough. Then I added the 171 (manual) and now I get a workout just loading and unloading the bikes. I did a weigh in before and after adding the 171 and the bike gained 5 lbs over a 7 speed shifter.

For anyone that may consider using a Nuvinci manual shifter......don't force the shift. Bad things will happen to your cables and the shifter on the handlebar.

Bob

 

[grindz145]

yeah CVTs are not cool for bikes where you care about efficiency. They make sense for ICE scooters, becuase they allow the system to stay simple, but otherwise... not worth it.

 

[amberwolf]

If I remember correctly, doesn't the nuvinci throw away like ~20% of the power you feed it?

Depends on how *much* power you're feeding it. The less power, the less that's lost. I've forgotten the username, but someone has a dual motor setup driving a NV that reports about a third of the power lost, or more. IIRC that's at least 1KW input, so I expect as you go up toward the limits of the NV171 (5KW?) it gets worse.


What I want to do with mine, since it will be part of the middrive setup and not in the wheel, is to set it up so that 1:1 is the default, and then make a locking pin I can put thru the motor input sprocket (which will probably be on the splines) into the flange of the hub, physically locking the hub into 1:1 to totally bypass it, at least for some efficiency testing, and possibly for actual road use for when I need power efficiency vs gearing changes, if it turns out to make a significant difference.

 

[Miles]

If I remember correctly, doesn't the nuvinci throw away like ~20% of the power you feed it?

Depends on how *much* power you're feeding it. The less power, the less that's lost.

Less in absolute terms but more as a percentage, I would have thought. Efficiency usually goes up with input torque - all the data I've seen on geared hubs shows that.

 

[amberwolf]

Interesting. In my conclusion above, I am making assumptions based on the little data that's been reported by others using the NV hub powered by a motor; most don't report any efficiency info. What little I have seen (which I wish I had saved each time I saw it, as I don't know where any of it is on the web other than some of it is here on ES in various threads) indicated that people running lower power saw higher efficiencies than those running higher power.

I assumed that this was because of slip in the traction fluid/ball interface, where the power ended up just going into heating the fluid and mechanical parts. This wouldn't happen in other IGHs, because they are using gears, so the losses are in different places, and tend to be fixed losses, whereas with friction setups the losses would I think tend to get higher with higher input power, due to more slippage.

Again, these are intuitive feelings, rather than a factual thesis, so I could be completely wrong.

Need to get my bike finished, then take it to Liveforphysics and have him dyno it, preferably not to death. :lol: Then we can get efficiency curves for the NV itself.

 

[Miles]

I assumed that this was because of slip in the traction fluid/ball interface, where the power ended up just going into heating the fluid and mechanical parts. This wouldn't happen in other IGHs, because they are using gears, so the losses are in different places, and tend to be fixed losses, whereas with friction setups the losses would I think tend to get higher with higher input power, due to more slippage.

Possibly. There must be a substantial pre-load on them, though.

We really need to do testing for various loads and speeds at different ratios across the range.

 

[Ypedal]

My Nuvi sheds plenty of heat driven at 4000w .... takes abuse well, but not efficient.. i have to quantify that somehow... got rims for my 3 speed SA hub, need to order spokes..

 

[grindz145]

My Nuvi sheds plenty of heat driven at 4000w .... takes abuse well, but not efficient.. i have to quantify that somehow... got rims for my 3 speed SA hub, need to order spokes..

What do you drive it with? Cyclone?

 

[StudEbiker]

It may not be the most efficient, but it does make for a very nice ride. I am still getting mine dialed in, but it just gets better and better. Of course my application is an attempt to have a reliable commuter that climbs hills well. I think it is within my grasp to get this using the NV dev kit. I absolutely love not having to shift. I am now getting 30-35wh/mi depending on how much I pedal, and I think I can do better if I get a larger front chainring so I don't have to spin so fast to help the motor.

 

[Ypedal]

My Nuvi sheds plenty of heat driven at 4000w .... takes abuse well, but not efficient.. i have to quantify that somehow... got rims for my 3 speed SA hub, need to order spokes..

What do you drive it with? Cyclone?

Seriously ?... dude.. where you been all these years ? lol

Page 4, about to start the 5th !
http://ypedal.com/RC/RC4.htm

Astro 3210 ! :wink: