93% Peak Efficiency 2 speed Hubmotor $285

[Bluefang]

Hey john and everyone else playing with the motors, has anyone got anything up and running yet?

I am just waiting for my motor to arrive but i have the frame mostly modified, the chain line hopefully figured out and the battery pack testing and about to be built for 20s4p Lipo out of 5s5ah packs.

Motor will be mounted on the chain stays with about 15mm of clearance to the metal around the magnets, controller above that. The batteries will be mounted in a saddle type arrangement on either side of the frame top tube. With the chains there is just enough room for double free wheels on the Nuvinci drive with the driven been closest to the hub and 38 tooth sprocket and something around a 18t sprocket on the hub. The seat will be layed back as i need abit more length to the handle bars to be comfortable

[johnrobholmes]

I'm shooting to mount mine this weekend. I just have to finish up wiring on a battery I built for a buddy, then it is time to get my scooter running again. I have the ebike withdrawals!

[John in CR]

Looking great Bluefang. I can't wait to see it rolling.

I got some sprockets done up on good centers. I'll need to do a post and pics of the process, because others may want to go the same route, ie just home tools. My issue now is rain . I really want to get some test riding in before deciding on which wheel size (I've got sprockets ready for both), because it could impact battery location. If I like the ride, then go all out waterproofing, get a quality shock mounted on the back, and whatever else is needed for a beach and trail capable ride.


HELPFUL HINT:
Any time you twist the throttle and get a null result, whether it's during initial spin-up or later after a stop, the first thing to check is that the 2 speed switch didn't get bumped and is in between modes. I don't have any linkage on mind and it's still never jumped out of position while riding, but I have bumped it a couple times while out and get back on and nothing leading to minutes of frustration before remembering "ah, check the 2 speed switch". One time I was on the phone calling for a rescue, when I remembered to check.

Once I do linkage, I'm going to figure out a way to have a spring that works both ways in the same manner as a kickstand spring to provide force to keep it in high or in low. In between means all 3 phase wires are disconnected.

John

[Bluefang]

For the 2 speed you could clamp a bar parallel to the shifter onto the axle of the motor so its above the 1/4 turn motion of the shifter and just a spring linking it to the end of the shifter linkage should work

Spring.png (15.4 KiB) Viewed 382 times

[John in CR]

For the 2 speed you could clamp a bar parallel to the shifter onto the axle of the motor so its above the 1/4 turn motion of the shifter and just a spring linking it to the end of the shifter linkage should work

Spring.png

Exactly. That's the way kickstands work. A couple of things: 1. The shift range isn't anywhere near 90° and 2. Going that route puts something out even wider, and width at the ends of the axle is one of our compromises using this motor.

Thinking about it some more, just an easy rig with a bicycle cable would work fine. A spring load in the extended position just like on a front or rear derailleur and friction or clicks in the mechanism to hold the other position should work fine.

For those like me who shift to high or low rarely, the easy way is no linkage at all. That probably leaves the shifter the most open to getting bumped when you're off the bike and essentially knock it out of gear. It happened to me twice in 9 months, so I bring it up just in case you get on your bike, turn the key on, turn the throttle and nothing...check the shift first, not after 5 minutes of trouble shooting like happened to me both times.

[Kin]

I'm not sure if this has been covered, but I was rereading a bit of the thread and wanted to asked about the part where there's no change in torque because <<the controller responds to the greater resistance and higher inductances.>>

I thought the basic controllers simply limits current based on the drop across the shunt? So although the driving characteristics of the motor would change, I don't understand why the current allowed would decrease.

[Bluefang]

Lucky me with the shifting i will hopefully be able to lock it into high speed drive and use the nuvinci to do the rest. Kinda hoping i can get a 3-1 ratio otherwise i may have to stick to the low drive and use 1.5 ratio which is a lot easier to get a cog mounted to the motor then.

[John in CR]

I'm not sure if this has been covered, but I was rereading a bit of the thread and wanted to asked about the part where there's no change in torque because <<the controller responds to the greater resistance and higher inductances.>>

I thought the basic controllers simply limits current based on the drop across the shunt? So although the driving characteristics of the motor would change, I don't understand why the current allowed would decrease.

I've never seen the explanation of why on the forum. We only have the real world results. eg DoctorBass went through a lot of trouble and expense giving an X5 series/parallel switching to turn an X5303 into an X5306 at the flip of a switch. He thought he would get a big torque difference, but it was virtually unnoticeable. Too bad, despite my suggestion to try it, the good Doctor didn't check to see how much better the low setting was at hills. There have been other examples too. I've definitely seen it first hand.

Keep in mind that it's not just one continuous flow of electricity. It's short pulses, and when you switch the windings like that, not only do you change the resistance greatly, but you also change the inductance too. I suspect the difference in inductance is the bigger effect, because it slows the electricity so much.

Bluefang will see changes in torque, but that's because his Nuvinci changes the gearing, not the windings.

John

[John in CR]

Lucky me with the shifting i will hopefully be able to lock it into high speed drive and use the nuvinci to do the rest. Kinda hoping i can get a 3-1 ratio otherwise i may have to stick to the low drive and use 1.5 ratio which is a lot easier to get a cog mounted to the motor then.

Yes, high is definitely what you will want unless it's geared too high, but with right at 2:1, you should be golden. The only time I could see low coming into play would be some off road rock crawling at very low speed, or for lower speed cruising, say below 40kph or so. In that case it may be better running the motor in low speed mode and allow the NuVinci to operate at a higher gearing where it's more efficient in it's own operation.

You're going to have all kinds of fun experimenting with all those variables.

John

[Bluefang]

I think if i gear it for a 2-1 ratio my top speed is something like 140km/h....way over what i will ever do on a bike i have welded on

Nuvinci has a 1.75 overdrive hence the 3.3-1 ratio is needed or i end up with a high speed monster or having to use the low speed settings, really hoping the motor gets here this week which is about my normal shipping time from China. If so i should be good for getting the bike going soon as were all be asked to use anual leave due to lack of work atm lol.

[John in CR]

I kinda screwed the pooch on that one and forgot about NuVinci's overdrive. Now if you did some heavy power modding (good ventilation and gut the 2 speed switching), and put on a streamliner, then it could do the 140.

Actually after thinking about it, all you'd need is the streamliner, since that's just 87mph and Sam broke 82mph on pedal power alone.

John

ps- I've already got a note in to the shipping dept, because you should have had it already. I'll send you a pm after they get in Monday morning china time.

[Kin]

Thanks John, your comment led me to
viewtopic.php?f=2&t=9215&hilit=nine+continent

It's a 42 page thread though, so it will take some time to read a bit more. I don't have that time right now, but hopefully will check out soon. Initial reactions seemed quite positive...so I'll have to see what the eventually results are about torque not actually improving. (I am assuming the mechanical switch is a delta/wye(star) change, not a turn count change)


Edit: found the thread about torque potentials not changing: viewtopic.php?f=28&t=20484. Still trying to tease apart the implications here. I still believe it to be true that for starting up a wye setup will allow you more torque at your current limited controller. But it sounds like LFP is saying only the current you put in will matter, not whether it's in wye or delta. .... sleeping for now.

[Bluefang]

Isnt the no noticable change in torque more related to how the power is distributed? Were all measuring amps on the intake side of the controller so say you have 100V 100A controller, so 10kw of power. When you take off in low gear you have essentialy 500 phase amps(controller limited again) and really low voltage due to the controller, then you reach top speed in the low gear and your motor is seeing 100V100A that is on the intake on the phases. Your in low gear so it produced better torque down low but now that your limiting the amps into the controller the torque is suffering at top speed, you change parallel to serial making the motor double its KV.

From there your phase volts drops to ~40V and your amps perk up abit and hit 300a or something on the phases, so even with a faster wind happening you still feel the same torque on the motor because the controller had been limited by the amount it could supply in low gear vrs the start of high gear.

I am probably completly wrong but that would be how i picture it happening

PS, thanks for checking John, i always expected it this week.

[Kin]

I think you're right in some sense, but I am concerning myself more on the initial torque of the start, not the transitional situation that you're describing. But it sounds like a reasonable explanation for why Doc and everyone were saying it feels like "turbo" because there's actually a clear torque increase and secondary acceleration moment when going from wye to delta as the phase amps go up / controller-bemf difference goes up.

But I still find myself confused at the starting torque. It sounds like
http://www.rcgroups.com/forums/showthread.php?t=1096719
"What I am trying to say is that teh MAXIMUM torque a motor can do is a fixed thing. It cannot be magically increased by rewinding it."
"Sure. But what most people mean is that a low KV motor is capable of producing more torque, It simply is false. It isn't. All its doing is producing the same torque at a lower RPM, current and power than its high KV equivalent (given the same pack voltage). [Kin's note: I feel like this is saying starting torque will be improved, but the gist of the entire thread is NO, it won't.
http://www.astroflight.com/pdfs/3220.pdf
Gets at some of the problem. Though the whole thread is about turns, I think the rationale applies to delta/wye(star)

The situation with having a switch might have to do with the fact that we must limit our pack voltage for the delta setting. But because we have a current limited controller (and wiring, etc) I still think we should have better starting torque with wye, but I hate to be that stubborn clueless poster. I am just getting grumpy at my cluelessness . Ought to learn more. At this point though there's no sense in posting more about it, I'm taking the thread entirely off topic and I apologize about that.

Back to business!

[Bluefang]

Kin, i kinda think your on the wrong wave length. Pretty sure this motor is parallel to series shifting. Ie changing the turn count rather then the delta/wye change your talking about. In this case you will produce more torque per amp in parallel then you would in series.

John, i have recieved my motor, well its at my place atleast i have not been home to have a look at it Any helpfull hints on how to strip the excess metal from it? Did you just cut the bulk of the flange off with a cut off disk on a angle grinder with alot of breaks for cooling, then set it up in a mount to run the motor and use a file to get it finished off with a constant stream of water cooling it down.

With the brake side did you cut it off with it all on the motor or did you remove that cover rather then have to worry about heat sink to the magnets?

[John in CR]

Bluefang,

Great, you got it. They sent me a note saying they were waiting for out-of-stock shifter arms, and your box was still in China. F'ing crazy.

I took the the drum flange off in one clean piece. I had a helper work the throttle and pour some water after cutting for a short time. We gauged the time by temp early on. I used a hacksaw with the spinning motor for most of the cut. The cast metal chews blades, so have a few and use a sawing action. Then I had to open the cut up with a grinder anyway once it got most of the way through, so just the really thin stainless cutting discs is fine. Just don't heat it up too much. An air driven grinder would be great, then I think you could use water at the same time and make quick work of it.

The reason I did it with the motor spinning nicely was to ensure a nice even straight cut, so the balance of the motor wouldn't be thrown off.

John

[John in CR]

Kin,

Actually, I don't know definitively if it's Delta/WYE switching or Series/Parallel. Until I spun one up with a speedo attached, I thought it was delta/WYE because that's what LFP thought it was just looking at it. However, the no-load rpm difference is right at 2:1 . The reason I've backed off in this thread, is because looking at the wiring of the switching mechanism, it has connection one way that is a triangle loop. Maybe the other mode just has a hidden triangle, so they're both delta and it is series parallel after all. Series/Parallel or Delta/WYE means a change in Kv only, so they're only different in their magnitude. One really odd coincidence is the change in no load being 2:1, but my change in top speed is 1.7:1, which is the Delta/WYE difference, though that .3 is easily explained by the effect of wind resistance.


In the big thread, what is talked about being equal is max torque in either mode, which is theoretically true. We have current limiting though, so you'd think there would be a torque difference due to same amps, which is also theoretically true. What I found though with the same controller, just switch the switch, that there's no big difference in thrust. The only explanation for that is that the controller must behave differently looking at essentially 2 different loads of the motor. One has much higher resistance and inductance than the other.

Also keep in mind that the controller limits primarily the battery current, but the motor only sees phase current, and it's in the phase current that things are so different with big changes in inductance.

The bottom line and I can't definitively explain it, is that the lower Kv mode doesn't give you markedly more takeoff thrust in real world use. The same is true for those high turn count motors...they aren't thrust monsters compared to their speedier wind counterparts using the same controller. They do better on hills because they get into their more efficient band of operation at proportionately lower speeds. The exact same thing is true with this motor, except it has both in one. People have found same result with Delta/WYE as well as Series/Parallel switching. They were looking for more torque, but didn't get it. What they generally missed was that they did have better hill climbing, but never tried.

John

[_GonZo_]

Here is a chart with same bike, same battery, same controller, same motor but with different windings, almost to a series parallel difference.
There is two ways to look at it, one is from the point of view of getting the most power and torque out of the motor/system (Sporty), and the other witch I find more interesting is to get the most efficiency from the motor/system (Economy).

From what this chart says you get a sightly higher torque with the series wining at the start,and more from the parallel winding at higher rev. But what is more important are the differences between the power curves and efficiency curves.

On my point of view parallel/series switching acts very similar to a mechanical gearing on a IC motor, but we have to change the way we think about it because IC motors and EM have a big difference, IC motors gives you the max torque at high speed (Usually around 3000rpm) while EM gives you max torque at low speed (It can be said 0rpm)
So the points in order to get a more sporty or economy results from the motor are different and actually quite opposite.

We are all used to drive IC motors and when we want to drive more sporty we change gears at high revs, and when we want to drive economy we change gear at low rev (Up to a certain point) on EM it is completely different then, looking at this particular case:
If you want to get out the best performance out of the motor (Sporty) you should change "gear" at the point where torque curves meet, that is around 14Km/h
If you want to get the best efficiency out of the motor (Economy) then you should change gear where the efficiency curves meet, that is around 25Km/h

Hope this helps.
Sorry about my English, it is not my mother tong.

[John in CR]

Nice post Gonzo. There we have it, a slight increase in startup torque/thrust, and I'm not sure the simulation accurately factors in current limiting or the difference in the controller behavior looking at the 2 different loads. I would disagree about looking at it like mechanical gearing, because just like "torque wind" motors don't have more torque it is misleading since changes in mechanical gearing changes torque directly. More accurately it's High Speed/Higher Power in high, and Low Speed/Lower Power in low, and low is better on hills.

I agree that a shifting approach is theoretically better, but I would stay away from that in the interest of long term durability, because the gain is small. I shifted my test unit back and forth many many hundreds of times as part of my testing, and it's still fine, but wires do have to flex slightly to accomplish the change from one set of contacts to the other. Then again, the wires on my bikes with suspension do flex some with every bump in the road, so it's possible that I'm being overly cautious. My tech contact at the factory says switching while on the throttle is ok too. Maybe that is true since the phase currents are AC, and yes I did do it during testing, but I definitely recommend against it out of concerns of scarring the contacts with tiny arcs, and concerns about the controller seeing phase connections briefly brake under load with a huge change in what the motor looks like to the controller.

I'm sure you want to avoid shifting while regen braking, something I didn't even test because it seems like such a bad idea, especially going to low while regen braking at higher speeds. Switching to low while at high speeds, whether braking or not, also seems like a bad idea. DD hubbies always seemed "odd" to me while coasting down hills above no load speed. Shifting to low while at high speed can make that like coasting down a hill at almost double the no load speed, something I've never done, since that would be way over 100mph on my bikes.

The other thing I want to caution again against is trying to push this motor in stock form to extreme currents. The factory uses 50A controllers. Those I supply are 60-70A, and I did run 80A for months without issue. We have to keep in mind that the phase currents go through those contacts in the motor, and there's always going to be some added resistance there compared to just continuous wire.

The big question is who is going to be the first to gut that switching, add ventilation to dissipate the extra heat that comes with higher power, and see what this baby will really do. Maybe we can talk JohnRobHolmes into it. He's got an extra motor in hand anyway, since his wiring harness got damaged in transit, and I sent him a replacement motor. I'll contribute that motor for finding if he'll contribute the elbow grease.

That brings up another important point. Protect those axle ends. Redoing the wiring harness on these motors is not a job I envy. Below is what I came up with, my usual quick and crude but effective to get the test bike on the road. They just slide onto the axle flats. I didn't reinstall them when I had the motor off, because my bike is protected from falling over, and if I take a spill on the road with that bike I'll have much bigger worries than some broken wires. I have clamping dropouts, so the axle nut's purpose was simply to hold these on. The wrap around the ends probably wasn't necessary, just something solid extending past the end of the axle to take the force if the bike goes on it's side would have more visual appeal. I'm sure you guys will come up with something better.

John

Axle end shields for MiniMonster.jpg (35.38 KiB) Viewed 1418 times

[_GonZo_]

I run another simulation with same motors but limiting the Amps, the results are more less the same but with obviously less torque and less speed.

[Bluefang]

yeah John, your right it was not my motor, my father indicated a large package arrived while he was working on his car at my place, turns out it was my 37tooth sprocket.....much smaller then the 1foot square box he indicated I am fine waiting i still have a few bits i want to build before the motor is needed

[johnrobholmes]

I didn't get a chance to install this past weekend, but I did get my buddies bike running! So now I just need to button up a few wheel builds before I can work on my own parts guilt free

[John in CR]

I didn't get a chance to install this past weekend, but I did get my buddies bike running! So now I just need to button up a few wheel builds before I can work on my own parts guilt free

If you end up having to switch a pair of halls to get the direction of spin the way you need with whatever controller you're connecting, I'd cut and swap 2 on the controller side. The motor side wires are glued into the hall connector, and they use an odd wire color scheme to try to force people to use their controllers. You don't want to lose track of +5V and gnd for the motor side hall wires.

That's the kind of stuff you end up dealing with when dealing directly with a factory that makes it's motors for it's own scooters and not for sale, so they try to keep things proprietary. 4 almost identical wires of the 5 halls is an example. These little hassles are worth the effort though, because the motor really is significantly better and the introductory price a steal. I think I should add $100 to the price and eat $100 of the shipping though. People have sticker shock over the shipping cost, because it's not buried in the price like with other motors that often have multiple layers of shipping built in.

John

[_GonZo_]

25% more torque and 20% more efficiency, no comment ))

[John in CR]

25% more torque and 20% more efficiency, no comment ))

Simulations are useful tools, and I use them a lot, but I only use them to look at the the range from near peak power up to peak efficiency. They're not particularly useful to me below peak power, because how the throttle is used and controller sets play a huge roll that the simulator can't take into account. My experience, as we as those who have done series/parallel or Delta/WYE switching, is that it has negligible effect on starting thrust in real world use. There's no way I've seen a 25% increase in starting torque with any of the controllers I've tried. An increase in current of 25% definitely makes a notable increase in thrust, so I'd disagree with the 25% figure from the sim.

I don't doubt the efficiency difference, but the only time I can see it making a useful difference in overall efficiency is in stop and go traffic. You don't need high speed in heavy traffic anyway, so sure, leave it in low. I'll pass on trying to save a fraction of a penny per ride, but ensuring a cool motor is always worth a bit of extra effort.

John