Nuvinci developer kits for cheap

[dumbass]

Yeah, I worried about this as well, but Tommy, at FFR Trikes says they have been using them on their setups with 3220s, and they've had no issues. I'm also thinking about adding one of Matt's adjustable torque limiters on the big pulley, so that should help limit the stress on the Oddessey 13T freewheel. BTW, it looks pretty beefy, to me, with 1/8" sprocket teeth instead of 3/32". Not sure of the innards, though, but we'll see. I've got a question into Staton, to see if they might have a larger splined sprocket laying around. If that happens, I'll just use a 16T ENO on the jackshaft in place of the 13T Oddessey. If 27T is it, I'll give the Oddessey a whirl.

-- Gary

Gary, What motor setup you are planning to use and how you are going to reduce the RPMs. I'm asking because as mentioned I have a very simular setup already. I am planning to replace my drive with a new motor this winter that is rated for (I think) 100rpm/volt and I am planning on a 48v or 60v system. The motor/controller is rated to 72v. Like you I was planning on using a belt for the first reduction and go with chain for the second (final) reduction. But I really would like to try a belt for the second reduction as well because of the possible noise of chain. Do you think the smaller chain (#25) is going to be strong enough? Bob

 

[drewjet]

I got my hub this week, and will have my rim by Monday, and hopefully laced up by mid week so that I can start testing. I went over to the Fallbrook Nuvince forum to hopefully get some answers, but so far it is awful quiet over there. See if anybody here can answer.

1 Is it able to shift under load, such as accelerating, or do I need to back off motor power to shift?

2 What really is the efficiency of it. Just spinning by hand takes a good amount of effort compared to a regular hub?

3 Does the efficiency change depending on the load?

4 Does the efficiency change depending on the the gear ratio selected?

 

[dumbass]

I got my hub this week, and will have my rim by Monday, and hopefully laced up by mid week so that I can start testing. I went over to the Fallbrook Nuvince forum to hopefully get some answers, but so far it is awful quiet over there. See if anybody here can answer.

1 Is it able to shift under load, such as accelerating, or do I need to back off motor power to shift?

2 What really is the efficiency of it. Just spinning by hand takes a good amount of effort compared to a regular hub?

3 Does the efficiency change depending on the load?

4 Does the efficiency change depending on the the gear ratio selected?

I can answer question one for you based on my 171. I can down shift under a fair load but can not up shift under a load. This is one of the problems that I see with the power shifter because it won't know if your under a oad or not. They may have corrected this problem in newer models but don't count on it. I was really screwed a few times on really steep but short hills a few times. I started up the hill from a stop and because the hill was so steep I thought I would need to gear all the way down. But I quickly relized I was wrong but couldn't up shift because I didn't want to get off the power. Not the worse thing to happen but I crawled up the hill.

I think I read a Nuvinci claim that the eff. was around 95% in general but it's something they don't like to talk about much. Keep in mind that you should have an overall eff gain by being in exactly the right ratio rather then excepting what a gear can give you. I know there were always time for me when I felt I needed a gear between my gears. But I have been really happy with the Nunvinci because I can settle in on th perfect ratio for my exact situation. I wish I had a crank gear set so as I gain speed I could keep up with the motor speed. Bob

 

[amberwolf]

2 What really is the efficiency of it. Just spinning by hand takes a good amount of effort compared to a regular hub?

3 Does the efficiency change depending on the load?

4 Does the efficiency change depending on the the gear ratio selected?

What we need is someone with a dyno to test their bike with a regular wheel and cassette drivetrain at a few of the ratios equal to the NV (highest, lowest, and a scattering in the middle).

Then swap the wheel out for the NV wheel, otherwise identical, and test the three points above.

That should give us a good comparison set of data, right?


I know that I have read of varying eff reports, from 95% down to 66%, and it seems based on my probably poor memory that the higher the power input and/or load, the lower the efficiency.

 

[GGoodrum]

Gary, What motor setup you are planning to use and how you are going to reduce the RPMs. I'm asking because as mentioned I have a very simular setup already. I am planning to replace my drive with a new motor this winter that is rated for (I think) 100rpm/volt and I am planning on a 48v or 60v system. The motor/controller is rated to 72v. Like you I was planning on using a belt for the first reduction and go with chain for the second (final) reduction. But I really would like to try a belt for the second reduction as well because of the possible noise of chain. Do you think the smaller chain (#25) is going to be strong enough? Bob

I'm using an AstroFlight 3220-7t which has a kV of 97, terminated in wye mode. These are sensorless, higher end, RC-type motors that are extremely efficient and pack a ton of power for their size (3" in diameter and about 3" long...). The only downside is that most of the RC-type ESCs, like the popular Castle Creations HV-160, have an absolute voltage limit of about 53V, or 12s if LiPos are used. For me, this setup is going on a 20" folding bike, and the max top speed I'm looking for is about 35 mph, so having a 48-50V limit is not an issue. I will use a 12s2p LiPo pack. On my existing setup on this bike, I have a Sturmey Archer 3-speed hub and this same 3220-7t setup, but with a direct drive configuration between the motor and the 3-speed hub. I did a special freewheel adapter that allows the pedal chain to be separate, with its own freewheel. The other side of this adapter sandwiches the 16t cog for the hub between the adapter and a 94T #35 sprocket. This size chain is smaller pitch than the normal 1/2" bicycle chain, but it is much beefier. Anyway, I had an 11T drive sprocket on the motor. As for #25 chain, I would say no way. I tried using this on a 1st stage setup that was only about 6" between the motor and the jackshaft, but the chain got tweaked. It stretches/bends pretty easy.



This setup was extremely powerful and I could not come close to giving full power in "low" gear. It just flips you on your butt, no matter what. I never even had to use low gear unless I was stopped at the bottom of a hill. 2nd gear, is 1:1, so most of the time was spent there, unless I needed to go a bit faster. in that case, 3rd would get me up to well over 40 mPH, but I gotta tell you, that's way too fast for me on a 20" bike.

I let my 23-year old son thrash this about for awhile, and now the shifting is sloppy and it makes even more racket. Even before we trashed the hub, this setup makes way too much noise, for my tastes. This is mainly due to the 11T motor sprocket. I've since learned that due to the chordal effect, the chain noise increases significantly as you go below about 13-14T.

Since it doesn't look like I can do a dual freewheel setup with the Nuvinci, I'm going to go back to adding an intermediate reduction, and use the Staton splined sprocket on the Nuvinci with the 13T Oddessey on the jackshaft. I'll then use Matt's high-power belt drive setup on the first stage, between the motor and the input side of the jackshaft. The total ratio will be similar to the roughly 8.5:1 I have now, but this should be a whole lot quieter.

-- Gary

 

[dumbass]

As I've mentioned I already have a Nuvinci 171 on 1 of my ebikes and now I'm trying to decide if I want to add one of these units to me second bike or not but I have a few question if anyone can help me.

Now that so many have received there test units can anyone confirm if it can be shifted manually in some way? I'm wondering if there's an electric shift switch (up/down) of some kind that is mounted to the handle bar so you can override the processor if/when needed. OR maybe id youjust don't want it to be programed for automatic shifting and only want the the value of pressing a buttom for it to shift. Or do you thing there's a way of adding a manual switch? On a standard unit with manual shifing there is an indicator that tells you where you are in it's shift range. Is there any type of indicator on these units that tells you where it is in it's shift range? Thanks, Bob

 

[GGoodrum]

As I've mentioned I already have a Nuvinci 171 on 1 of my ebikes and now I'm trying to decide if I want to add one of these units to me second bike or not but I have a few question if anyone can help me.

Now that so many have received there test units can anyone confirm if it can be shifted manually in some way? I'm wondering if there's an electric shift switch (up/down) of some kind that is mounted to the handle bar so you can override the processor if/when needed. OR maybe id youjust don't want it to be programed for automatic shifting and only want the the value of pressing a buttom for it to shift. Or do you thing there's a way of adding a manual switch? On a standard unit with manual shifing there is an indicator that tells you where you are in it's shift range. Is there any type of indicator on these units that tells you where it is in it's shift range? Thanks, Bob

Yes, there's a couple of ways to do this. You can use a switch, to at least cycle between a couple of different shift profiles, but these are still going to be speed-related shifts. If you only wanted two speeds, you could program one two shift tables with just one setting, I guess. Frankly, I'll have to look at the manual again, to figure this out.

The other way, which I'm really leaning towards now is to use the 0-5V analog input to vary the ratio. All you would need is something like a Magura throttle and a +5V source from somewhere. This will basically "mimic" the way the manual control works, but it won't take any effort to adjust the setting. Although the shift table approach is quite clever, I, like many here, worry that this approach just isn't going to work very well with a motor. It doesn't have any load sensing input, other than a drop in speed. Anyway, the analog version is much easier for me to get my arms around. What is really needed, however, is a throttle that is spring-loaded to return to a mid-point. this could then represent 1:1 and then you twist it one way to go into "low" gear, and twist it the other way to go faster. I'm going to noodle this some more, and see if there isn't some way to do this. Since I am hopelessly mechanically challanged, I might just try to come up with a electronic way to do this. The easiest would be a 3-position rocker switch and a resistor network that would output 0V, 2.5V and 5V, for the three positions.

-- Gary

 

[dumbass]

Yes, there's a couple of ways to do this. You can use a switch, to at least cycle between a couple of different shift profiles, but these are still going to be speed-related shifts. If you only wanted two speeds, you could program one two shift tables with just one setting, I guess. Frankly, I'll have to look at the manual again, to figure this out.

The other way, which I'm really leaning towards now is to use the 0-5V analog input to vary the ratio. All you would need is something like a Magura throttle and a +5V source from somewhere. This will basically "mimic" the way the manual control works, but it won't take any effort to adjust the setting. Although the shift table approach is quite clever, I, like many here, worry that this approach just isn't going to work very well with a motor. It doesn't have any load sensing input, other than a drop in speed. Anyway, the analog version is much easier for me to get my arms around. What is really needed, however, is a throttle that is spring-loaded to return to a mid-point. this could then represent 1:1 and then you twist it one way to go into "low" gear, and twist it the other way to go faster. I'm going to noodle this some more, and see if there isn't some way to do this. Since I am hopelessly mechanically challanged, I might just try to come up with a electronic way to do this. The easiest would be a 3-position rocker switch and a resistor network that would output 0V, 2.5V and 5V, for the three positions.

-- Gary

Sounds like you are the opposite of me. I am challenged electronically but being a mechanical engineer I manage mechanically.

I honestly don't think your going to be happy with just shifting based on speed. To start with there's no way to program for different starting conditions. On my setup I start out about 30% off the lowest end. But if I was starting out on a hill I might want a lower range. On a manual setup as I have now I can do this. But with a programed setup this is going to be a problem. Personally I would like to just scrap the program all to gether ahd just be able to have an up/down switch jumpered into the drive.

Personally I think Nuvinci relized some of these problems which is why they are now offering these systems so cheap. Being an elecrical person do you think the processor could be setup to also read amp load? Then it might work well. Bob

 

[GGoodrum]

Sounds like you are the opposite of me. I am challenged electronically but being a mechanical engineer I manage mechanically.

I honestly don't think your going to be happy with just shifting based on speed. To start with there's no way to program for different starting conditions. On my setup I start out about 30% off the lowest end. But if I was starting out on a hill I might want a lower range. On a manual setup as I have now I can do this. But with a programed setup this is going to be a problem. Personally I would like to just scrap the program all to gether ahd just be able to have an up/down switch jumpered into the drive.

Personally I think Nuvinci relized some of these problems which is why they are now offering these systems so cheap. Being an elecrical person do you think the processor could be setup to also read amp load? Then it might work well. Bob

Actually, I started out life as an aerospace sowtware engineer, back in the '70s. Back then, there wasn't a Computer Science degree type until my 2nd junior year ( ), and it ended up being pretty close to an EE curiculm. Whatever electronics knowledge I received I quickly forgot but in the last several years, Richard has been tutoring me.

The anaolg input bypasses all the shift table stuff, which is based on speed. It simply translates the 0-5V analog input into the desired ratio. 0V is the maximum reduction and +5V is the max overdrive ratio. This is why I said it can mimic the way the manual shift works, with the twist grip.

-- Gary

 

[recumpence]

Well, Gary, you understand WAY more than I do, I can tell you that. I understand just enough to properly burn the tips of my fingers with a well placed plasma arc created by too much lipo in a way too overpowered bike equipped with far too few safety features.

I guess you could call it a highly refined skill.

Matt

 

[GGoodrum]

Well, Gary, you understand WAY more than I do, I can tell you that. I understand just enough to properly burn the tips of my fingers with a well placed plasma arc created by too much lipo in a way too overpowered bike equipped with far too few safety features.

I guess you could call it a highly refined skill.

Matt

Oh, I've had my share of burnt appendages. Remember this one?:



This was my first inadvertant attempt at a123 arc welding. :lol:

 

[amberwolf]

For a shifting setup that does work on load and speed, take a gander at that block diagram I put up on a previous page of thsi thread (and also here: http://www.endless-sphere.com/forums/viewtopic.php?p=449421#p449421 ). I think it can be done relatively simply, in pure analog, without having to build and program an MCU setup for the purpose. It would just use the analog input to the NV shifter to do all the grunt work of shifting.

 

[amberwolf]

Reading the latest in the Beta CA build for R/C controllers,
http://www.endless-sphere.com/forums/viewtopic.php?p=451753#p451753
I *think* the CA itself could be used for most of what we want.

Use the CA's "throttle" output as the NV shifter control input, or process it a bit with some external electronics (like a manual override input shifter, etc).

 

[johnrobholmes]

Haaaay, what a grand idea! I will mull this over a bit, I bet there is a way we could interface them.



I have been pricing frames for my build lately. I don't think I will have the time to build the frame I want, so I may just get a standard fat bike and bolt on components.

 

[justin_le]

Reading the latest in the Beta CA build for R/C controllers,
http://www.endless-sphere.com/forums/viewtopic.php?p=451753#p451753
I *think* the CA itself could be used for most of what we want.

Use the CA's "throttle" output as the NV shifter control input, or process it a bit with some external electronics (like a manual override input shifter, etc).

I have a nuvinci developers kit on my way for an entirely different application, but that could give me a chance to play around with this first hand.

Thinking about this, it should work fine to have a CA control the Nuvinci with the CA in speed control mode if the goal was to have constant motor RPM. Basically, you would set it up with with the motor shaft going into the CA speedo signal and configure the display to show RPM (so set the wheel size to 1666mm). Set up the CA Speed Limit to the target motor RPM, with ITermMax = 5V, ITermMin = 0V. Then map the nuvinci shift table such that 5V is the lowest gear and 0V is the highest gear (assuming it lets you do this).

You would then quite independently setup your throttle / ESC into a current throttle mode (either by programming the ESC accordingly if it allows it, or using a 2nd CA device to control the controller rather than to control the Nuvinci). This is important, since it is rather pointless to aim for a constant motor RPM if the rider has a voltage throttle.

It would then work like this. You would twist the throttle, allowing a certain power to the motor. The CA controlling the Nuvinci would then attempt to gear the Nuvinci drive ratio such that the motor was spinning at the target RPM while still drawing the commanded power. If the bike hit a hill and started to slow down, then the CA would see a lower speed on the motor, and increase the Throttle Override output voltage accordingly, which would switch the Nuvinci into a lower gear, allowing the motor to spin at a higher RPM etc. closing the loop.

If at any point the Nuvinci was maxed out at one extreme or the other, then you simply have a current throttle at one or the other ends of the gearing.

Justin

 

[kevo]

Reading the latest in the Beta CA build for R/C controllers,
http://www.endless-sphere.com/forums/viewtopic.php?p=451753#p451753
I *think* the CA itself could be used for most of what we want.

I have a nuvinci developers kit on my way for an entirely different application, but that could give me a chance to play around with this first hand.

Thinking about this, it should work fine to have a CA control the Nuvinci ...
Justin

Amberwolf and Justin, we are very fortunate to have your talents working on this. I have been considering the Nuvinci developer kit for a while. To have it controlled my CA-LRC is very tempting indeed. Thank you for your efforts!

 

[GGoodrum]

Interesting, Justin. One potential issue, I think, is that on the Nuvinci, I believe 0V is the lowest ratio and 5V is the highest, which I think is the opposite of what you said you needed.

-- Gary

 

[justin_le]

Interesting, Justin. One potential issue, I think, is that on the Nuvinci, I believe 0V is the lowest ratio and 5V is the highest, which I think is the opposite of what you said you needed.
-- Gary

Yup, that would be a potential issue! One potential (non electric) solution if the Nuvinci is going on a jackshaft is just to change the input/output direction of the Nuvinci hub. You can take the flange and spin it in reverse and have it drive the sprocket backwards as well, so your motor turns the Nuvinci casing and then the sprocket on the nuvinci is what drives the bike wheel.

Not exactly how they were intended to be used but the internal mechanism functions exactly the same this way, and then the direction of up vs. down gearing is effectively inverted.

I think this also lends itself better to the sprocket sizes on hand to achieve the high gear reductions needed for RC setups. Small sprocket on motor goes to large sprocket on Nuvinci disk mount or whatever, and then the small freewheel sprocket of the Nuvinci goes to large chainring on the bike wheel. But I am guessing that most people were planning to have the hub installed in the wheel, in which case this doesn't really work.

Justin

 

[amberwolf]

Will it really work that way? Because if it will, that'd make my jackshaft setup a lot easier.

But even if it can't be used that way (cuz it's in a wheel for someone's build) then it's simple to use an op-amp or transistor circuit to invert the analog signal going out to the NV shifter box, right?



@Justin: I may have missed it (I'm kinda worn out) but I didn't see the current limiting addressed in there, on the CA. In additon to the speed shifting, I'd like the CA to also shift based on power usage (just like it can cut throttle back if current limit is reached).

Is that going to require something different in the CA setup or is that just going to be like setting up your regular current limit in the CA and letting it do it's magic?

(I have not yet used current limiting or speed limiting on the CA yet, so perhaps it won't do both at once in the way I want it to?)


I also wish there was another speed sensor on the CA, because I'd like to sense wheel speed not just motor speed, to verify that I have or haven't reached the speed I'm trying to (I guess it'd also need to read the commanded throttle voltage for commanded speed, and need a setting in there for what exactly that range equals).

But some of this stuff can still be done externally via a tiny analog board, and mixed with the CA's control output, to do what we want. It'd just need pots to adjust, and take some tweaking to get the desired results, rather than just plugging in the numbers into the CA and riding off.

 

[drewjet]

Has anyone laced there wheel yet?

 

[GGoodrum]

Has anyone laced there wheel yet?

Not me. I'm still trying to sort out the drive/reduction part. I'm planning on using my 3220-7t (97 kV...) with one of Matt's v4-type drives, mounted to the seat tube. The 1st stage reduction will be his 20mm belt drive, with a built-in torque limiter. I'm not sure what the reduction is, but I'm guessing around 3.5-4:1. The output shaft on the v4 unit will drive an Oddessey 13T freewheel that I got recently from Tommy at FFR Trikes. On the Nuvinci, there will be one of Staton's splined sprockets with a second 17T freewheel attached, for the pedal chain. Normally, this is a 27T sprocket, but it turns out they had a 36T version laying around, so I snagged it. This will make the final motor-to-Nuvinci ratio between 9.7-11:1. This is perfect.

-- Gary

 

[drewjet]

I am debating between a 1 cross and a 2 cross. From my research it would seam in a 20" wheel a 1 cross would be better.

The problem is that this is going (maybe, depends on effeicency) in my electrathon and it is subjected to high side loads as it does not lean.

 

[dumbass]

I am debating between a 1 cross and a 2 cross. From my research it would seam in a 20" wheel a 1 cross would be better.

The problem is that this is going (maybe, depends on effeicency) in my electrathon and it is subjected to high side loads as it does not lean.

If I remember correctly the manual that came with my 171 said to use a 1 cross on 26" and up and a zero cross on anything smaller then 26". The manual also says not to alternate the spokes (front to back). The guy that did my lacing didn't read the manual I gave him and her used a double cross on my 26" wheel and he bent the crap out of the spokes to fit them. I refused the wheel and he redid it as a single cross. Personally, I think even a single cross was boarder line but it's been like that for 2 seasons now with no problem. If your kit didn't come with a manual indicating the lacing I would recomend finding one online. Bob

EDIT: This is the manual for the 171 hub see page #6. http://www.fallbrooktech.com/docs/NuVinci_Technical_Manual.pdf

 

[dumbass]

I just came across this. It looks like these units (if they are the same) can be setup for automatic and/or manual shifting with a switch. If that's the case I may look in to it now to replave my 171. . Bob

http://finance.renewableenergyworld.com/pennwell.renewableenergy/news/read?GUID=19186613

 

[johnrobholmes]

I am debating between a 1 cross and a 2 cross. From my research it would seam in a 20" wheel a 1 cross would be better.

The problem is that this is going (maybe, depends on effeicency) in my electrathon and it is subjected to high side loads as it does not lean.

Radial is fine with high side loads. It is actually the stiffest. 1 cross would be really bending the spokes hard on a small rim. If you are worried about the wheel wrapping up under power, lace the drive side with 1 cross and see how it holds up.