Nuvinci hub on a Twist/Lite

Krow

Pedelecer
Mar 7, 2007
25
0
Seattle, WA
I just had a Nuvinci hub installed on my Giant Lite. It replaces the Shimano Nexus-4.

The first improvement is the added gear range. The Nexus-4 has a total range of 184%. The Nuvinci has a total range of 350% with a low end gearing of 1:0.5. This gives me a much better climbing gear for the long hills around here (1 mile up at 7-8% grade) and a hope of climbing the steep ones (12%) with my son in tow on the back. The Nexus-8 would have provided a similar improvement.

The other change is more subtle and I'm still learning how to us it. The "Continuously Variable Planetary" system allows for very fine resolution control of the gearing ratio. The controller is a twist grip. The grip will twist a full 360 degrees to go through the whole range of the hub. There are no stops or clicks on the controller - it's more like a dimmer switch that smoothly travels through the gear ratios.

I'm still experimenting with how this added resolution impacts the behaviour of the Panasonic motor and controller. As most of you know, the Panasonic controller responds to the torque sensors on the cranks. The controller limits the motor's contribution at low-torque (to limit top speed) and at high-torque (to protect the motor?). With the Nexus-4 hub, I was developing a sense of when the motor would come in at each gear.

With the Nuvinci, I can more easily find a point where I spin the cranks at a high cadence. When I dial the gearing up from that point, I can hear when the motor starts to kick in - at first just during the power stroke portion of the crank rotation. As I increase gearing further, the motor stays on throughout the crank rotation and is working more. I'm not sure I understand what the "sweet spot" is for this motor/controller.

I haven't come up with a new strategy for how to use this finer control. I think I should be able to spin the cranks with less assistance in all conditions and thus increase the effective range of the battery. I don't know by how much yet. Alternatively, I could select a gearing that makes the motor contribute at it's maximum efficiency.

I'll continue the report when I have more experience on the new configuration,

keith.
 

flecc

Member
Oct 25, 2006
52,763
30,349
Interesting Keith, thanks for sharing this information. It will obviously take some getting used to with the Panasonic system on the Twist, but basically you've highlighted the right approach already, letting the cadence drop when you need more assistance, and raising it to economise on current.

The question mark is regarding the sensor on the motor output shaft gear. This is certainly the sensor for top speed cutoff, but conceivably may have the additional function of measuring the rotation rate, as there's a very large number of sensor triggering elements making continuous precise measure possible.

These are shown in the fourth photo down be the top red arrow on this webpage.
.
 

Krow

Pedelecer
Mar 7, 2007
25
0
Seattle, WA
first week with the Nuvinci hub

I've completed a full week of commuting and riding with the new hub for a total of about 120 miles. This hub is making a big difference to my riding experience.

I find it very easy to adjust the "gearing" to ride just over the point where the motor assist kicks in. The Nuvinci shift control is exactly as advertised - "like having a dimmer switch on your bike". I find that I can move the gearing ratio to meet conditions so easily that I don't ever hesitate to change as i ride along. And, even small changes have an impact on how much the motor kicks in.

Flecc, I'm still not entirely sure I've deduced the controller's algorithm, but it seems like the primary input is pressure on the pedals. The motor is ramping up at roughly the same pedal pressure at a wide range of different cadences. With the higher top gear ratio, I get motor contribution even at 23 mph (going down a slight hill).

I also believe that I can find a more efficient gear ratio for climbing. The added gearing range means I can drop lower than before. Since it's pretty easy to shift, even going up hill, I can fiddle with the ratio while going up the hill to find the right combination of work from the motor and from me.

The overall effect is to amplify the already great Giant Lite experience. It really does feel like a bicycle that amplifies your effort when needed and gets out of the way when you want to just ride the bike.

I'm riding with a lot less assist now during my regular commute. The motor kicks in mostly on starts to make me snappier off the blocks and then again up the big killer hills we have around here. As soon as I'm up to speed, I can give the shifter a slight twist to take the motor out of the picture - or add it back if I'm facing a headwind.

This has had a big impact on battery range. I ride the 15 mile round trip to work and use about 20-30% of the charge. With the Shimano Nexus-4 I was using about 50-60% of the charge. I haven't run the battery full out to see what the maximum range is, but I forsee some longer trips in my future to check that out.

For me, this has been a very successful experiment. The combination of crank drive, pedal-sensing controller and fine resolution shifting makes my e-bike experience more efficient and more fun!

cheers,

keith.
 

flecc

Member
Oct 25, 2006
52,763
30,349
Many thanks for that very full feedback Keith.

That "pressure" sensing is a bit of a conundrum though, and I agree, many of us feel convinced that's how it works by our riding experience. The problem is, there is no internal pressure sensor on this unit! No strain gauge or anything approximating to one. The pedal shaft where it sits inside the sensor drum is very crudely scuffed with deliberate scoring, but not in any precise way, but it's far too thick a steel shaft to have rotational torsion. That also makes it unlikely that the sensor could pick up a bending moment, but perhaps not completely impossible.

I might "take the bull by the horns" sometime and break open a sensor drum unit to see what's inside, as it's the only part hidden from me at the moment.
.
 

Tim

Esteemed Pedelecer
Nov 1, 2006
770
78
London
I've completed a full week of commuting and riding with the new hub for a total of about 120 miles. This hub is making a big difference to my riding experience.

I find it very easy to adjust the "gearing" to ride just over the point where the motor assist kicks in. The Nuvinci shift control is exactly as advertised - "like having a dimmer switch on your bike". I find that I can move the gearing ratio to meet conditions so easily that I don't ever hesitate to change as i ride along. And, even small changes have an impact on how much the motor kicks in.
Very interesting, I've been very much a big fan of the NuVinci concept for a while and it's great to hear you've been using one. So, how much does one of those cost then?
 

Dell

Finding my (electric) wheels
Apr 3, 2007
16
0
This reminds me of an old Philips washing machine pulley although it worked in a totally different way.
It used large steel balls inside a `V` pulley and as it turned faster the balls pushed outward causing the pulley to close, so the belt was forced to run further out so changing the gear ratio.
I think Daf used a similar technique for there variomatic transmission.
 

Krow

Pedelecer
Mar 7, 2007
25
0
Seattle, WA
price for Nuvinci hub

Tim, I bought it from an online cycle shop in the US. The hub came built into a 26"x1.75" wheel (a perfect fit for the Giant Lite/Twist). I paid $550 for it and then had it installed at a local bike shop.

The funny thing is this wheel probably came from Seattle Bike Supply but I bought it from a company in Massachusetts, on the opposite coast. So this poor hub has crossed the country at least twice already. Even with heavy use, it'll be a while until I catch up to that.

keith.
 

Krow

Pedelecer
Mar 7, 2007
25
0
Seattle, WA
pressure sensing on the Panasonic controller

Flecc, that is a mystery. I've been thinking hard about the control mechanism on the Panasonic motor. As you said, it seems that the control is driven by how much torque I'm applying to the pedals.

OTOH, that may be an illusion. Maybe it's strictly a function of cadence. Since we usually shift to bring the cadence to something we can turn comfortably, it may be that motor comes in at (say) 60 rpm and that happens to be roughly the same "pressure" at different gears and speeds.

But, that may not fully explain how the Giant seems to jump off the starting line as soon as i start pushing. hmmmm, i'd love to sniff at the code in that control box - or read enough Japanese to look at the original specs.

keith.
 

flecc

Member
Oct 25, 2006
52,763
30,349
Yes, the internals all point to cadence Keith. The pedelec drum has a multi lobe ring against one end, and it may also sense rotation from that scored surface, maybe accounting for the three wire rather than two wire connection.

On the motor drive gears final output shaft there's a also a multi element rotation sensor feeding directly onto the mainboard, so the control design seems entirely centred on rotation for feedback of what's going on. That makes it possible that it's sensing subtle relationships between the two though.
.
 

Krow

Pedelecer
Mar 7, 2007
25
0
Seattle, WA
decoding the Panasonic controller

I think there has to be another input to the controller's behaviour, Flecc. I was experimenting a bit during my commute today.

When going downhill, you can try any cadence you want and the motor won't engage. When stopped at a light, any pressure on the pedal will cause the motor to engage.

I wonder if there's a sensor on the driving gear that detects resisting load on the motor. When there's no resistance (like going downhill) the motor won't engage. When there's lots of resistance (stopped) the motor engages easily. Cadence seems to be an additional input.

Anything in the wiring that would confirm or deny this hypothesis?

keith.
 

flecc

Member
Oct 25, 2006
52,763
30,349
All true Keith, but there's nothing I can detect, and no wiring appropriate to that. As mentioned, the only possibility I've seen is that of the pedelec sensor drum sensing some bending moment in the pedal shaft at the point where it's been deliberately roughened. The problem with that theory is that it would require extraordinary sensitivity to detect, but that sensor drum is cheaply made and fully pressed steel encased. Not an environment for the detection of minute electrical changes.

I can see that sometime or other I'll have to strip a unit completely, degrease everything thoroughly and minutely dissect every part and it's functions. I'm too busy for that at present and for some while, but I may be able to get round to it at some future time.
.