Nine Continent Disk Motors

[justin_le]

So we received samples of the new front and rear motors from Nine Continent that have an integrated 44mm ISO standard disk bolt pattern just over a week ago. Most of the mechanical details check out pretty good, and it means there is a viable alternative to Crystalyte for direct drive hubs that have proper disk compatibility.

In the original NC hub design, there is a silly amount of wasted space due to the depth of each side cover. The rotor is only 35mm wide, yet the outer width of the two side plates is a full 68mm. In the front disk model, the right side cover is the same as we're used to, but the left one is substantially flat, leading to an overall width of 61mm:



This, combined with the fact that the hub is not centered but is offset about 2mm to the right, means that there is an extra 9mm between the left side cover and the fork that didn't used to be there.

As for the alignment of the disk rotor, on most of the standard front bicycle hubs that I measure there is 10.5mm from the axle shoulder to the surface on which the disk bolts, and that matches the spacing on this motor too:



That leaves just 14.5mm of space for the caliper between the disk rotor and the flat side cover. Lots of standard bike calipers, especially the hydraulic ones, need more like 16-17mm of clearance here, so people will need to watch out about this detail and possibly downgrade their calipers to a narrow one that fits.



The only other substantial difference is that the rotor and spoke flanges on these hubs are solid steel. I'm not sure why they did this, because normally the front hub rotors have a steel inner ring but are otherwise cast aluminum. As a result of this plus the extra metal for the disk mount, the motors are about a pound heavier than the standard NC front which weigh just 5.3kg:

 

[justin_le]

The specs on the rear hubs are similarly pretty good. In this case, both the left and right side plates are redesigned and square rather than rounded. The left side plate is still the same depth for some reason. I'm not sure why both sides weren't made skinny which would enable up to an 8 speed freewheel and more clearance for the caliper, but oh well.



On the rear hubs, more like 15.5 mm is needed for the proper rotor alignment to the frame, and this has that:



The Caliper clearance is 16mm, better than on the front but still not wide enough for all calipers.



Over on the freewheel side, the axle sticks out 36 mm, which is just a tad shy of what you need for a 7 speed freewheel, though it fits a 6 no problem. So for any 7, 8, or 9 speed freewheels a spacer washer is still going to be required.



What is nice on this side is that the freewheel threads are now machined on a steel insert rather than on the die cast aluminum, so this bit is much stronger. We haven't seen as many threads shear off on the NC rear side covers as on Crystalyte, for instance, but it has happened and so a steel thread is some reassurance:



Overall weight is heavier than the front, due both to the longer axle, and the steel inserts (the disc side also has a steel insert too, not quite sure why since the tapped holes for the disk itself are in the aluminum):

 

[Drunkskunk]

Wow, nice!
I'd seen this on 9C's website and was impressed then, without the details. With your review, that looks like a very good motor.
I got a 6X10 of the old design a few weeks back, and was really impressed with it's torque curve and ability to climb without overheating. It was Fundamentaly better than my Crystalyte 4012, except for the problems that this motor seems to adress.

So, got any bare motors in 6X10 you need tested?

 

[torker]

Justin, I was wondering about 6x10 winds as well. Have you ever discussed them with 9c? Any idea if they will make that winding avail. again anytime soon. Awesome pics and info btw.

 

[justin_le]

Justin, I was wondering about 6x10 winds as well. Have you ever discussed them with 9c? Any idea if they will make that winding avail. again anytime soon.

For sure, the winding is just something you specify when you order a big batch of hubs. Slow windings like a 10 turn just aren't my thing though. Super underwhelming with normal 36V and 48V batteries, so it forces users to go to non-standard voltages and series connected packs just to achieve the same performance that a faster wind gets off a standard pack.

Justin

 

[liveforphysics]

The 9*7 has 63 pieces of copper in each slot.

The 6*10 has 60.

Therefore, the 9*7 is always capable of more torque and efficiency, and has a greater speed range as well. The trade off is in additional FET heating during any time the controller is in battery current limiting mode due to running 30% lower PWM cycle, and hence 30% higher phase current loads on the FETs.

In other words, 9*7 is a good choice. 8*8 would be most ideal for hot-rodding type applications.

 

[amberwolf]

Do you think the covers from this new model will fit the old ones (like the 2807 from E-BikeKit)? If so, what's the chance of getting hold of a few of the new covers separately?

 

[dogman dan]

My CA verified what LFP says about the 6x10 pulling less current because it has less copper. It's part of the reason it runs so cool, it's 48v max watts is 200 watts less than a 9x7.

But it is a nice motor for slow riding at 48v, due to it's ability to lug up hills seemingly endlessly at 5 mph without getting very hot, even in hot summer weather. I found that 6x10 motor just what I needed for riding singletrack, and I just bought another in a front hub to put on my commuter. It wil slow me down to a legal 20 mph, and extend range. I ride too much on a bike path to be going 27 mph all the way to work. It should also be perfect for putting in 60-80 mile days on weekend tours in the mountains.

In a rear disk, the 6x10 would be nice for a few, but it will never be that popular to be slow. I don't know if it makes economic sense to order 10 at a time for Justin. Mabye he could add a few to future containers on a pay for it first basis?

 

[El_Steak]

My CA verified what LFP says about the 6x10 pulling less current because it has less copper. It's part of the reason it runs so cool, it's 48v max watts is 200 watts less than a 9x7.

I don't think LFP says less copper = less current. The 6x10 is pulling less current mainly because of the winding wire number of strands and number of turns. My 9x7 draws less current than my 10x6 (faster rpm/v) eventhough it has more copper. OTOH It allows me to run in a more efficient zone than the 10x6 at high voltages, so it runs cooler.

I guess more copper means it can more efficiently handle more current, not that it will necessarily pull more current.

 

[grindz145]

Thanks for the review Justin!

I'm thinking about picking up the 10x6s on a bargain and Wye-Delta converting it. Should end up with more like 6X10 performance.

 

[dogman dan]

As usuall I misunderstand when it gets tech. All I know is my 9x7 pulls 1200 watts and the 6x10 only pulls 1000. Isn't 1000 watts of 48v less current than 1200 watts of 48v? Sorry, about higjacking the thread.

 

[masserin]

This is excellent news indeed. Like amberwolf I'd like to know if just the plate will be made available to convert a front 9C I have on my ride. I'd love to have more breaking power on my chopper and without this new plate there is not another way short of machining a converter for the existing cover.

 

[justin_le]

Do you think the covers from this new model will fit the old ones (like the 2807 from E-BikeKit)? If so, what's the chance of getting hold of a few of the new covers separately?

So, I did order a handful of spare covers just in case for those who might want to try this out. However, it would require an axle swap as well, or a little bit of machining. As mentioned, the disk side plate is flatter than the normal one, which means that the location of the ball bearing is not exactly the same on the axle.

Here are the two side covers side by side:


If you measure across the flats of the original side cover, then there is a 10mm gap from the lip that fits inside the rotor to the ball bearing face:



On the disk side plate, there is just a 3mm gap here:



Soo, if you were to go ahead and try to swap the new plate on, you'd find that there would be a 7mm space between your the rotor and the side plate. One way to fix this if you have a lathe is just to turn down the axle so that the bearing can slide farther down the shaft.



Another approach that I might look into is that you can get a thin section 15mm ID ball bearings that are only 5mm wide (6802), as opposed to the the 11mm width of the stock Nine Continent bearing (6202). We could have an adapter made up with one of these much narrower bearings that properly fits into the cavity on the disk side cover, but allows it to slide a full 6mm further down the axle, and that might be just enough. Like this:



-Justin

 

[dnmun]

justin,

in that picture where you point out where one could machine 7mm off the boss on the axle, what does the new axle look like there and is it already machined back that far to match the new side cover?

will you attempt to source these covers and would you like for us to make a commitment to order some from you in advance to support it?

i bet we could get commitments for 50-100 side covers from those of us interested in this. and maybe the new axle too.

thanks so much for everything, dm

 

[amberwolf]

Any downside to turning the axle down? I have a lathe, and thus could do this, assuming normal lathe toolbits will cut into this axle material.

What would be involved in pressing out the old axle and pressing in a new one? Probably significantly more work than turning the axle, since I'd have to pull all the phase/hall wires back thru the axle, etc.

Optionally, as you say, a larger ID bearing could be used, if one exists that has the same OD. Or a different bearing with a ring adapter/spacer to fill the space the new one currently fills.


These narrower covers could allow me to use this motor where I might not have before, and use a disk or sprocket on either side without drilling into the cover itself.

Do you think the new rear sprocket side cover will fit the front 2807? I'd like to be able to try out a disc cover (front or rear) and a sprocket cover on my front 2807, if they'd fit.

 

[hydro-one]

Great news on the improvments !! I just last night sheared off a third disk brake boss off of last years motor, Justin. Hard stop, Dry pavment. I hope these new covers will solve my problems I have been having with braking bosses!! Im also surprised they did not improve the caliper clearance issue for us. Seems like NineC is asleep at the wheel here......I will be ordering a couple of these right away!!

mike

 

[Hyena]

In the original NC hub design, there is a silly amount of wasted space due to the depth of each side cover.

Justin how do you think this will affect cooling for those of us pumping alot of current into these motors ? You'd think the older design with more air gaps would be better for cooling ?

 

[masserin]

Do you by chance know when these motors or covers will be available for purchase?
Thanks

 

[liveforphysics]

Less air inbetween will definitely make for better cooling. Less delta-T between the stator to case with
A lower volume of air = better stator cooling.

In the original NC hub design, there is a silly amount of wasted space due to the depth of each side cover.

Justin how do you think this will affect cooling for those of us pumping alot of current into these motors ? You'd think the older design with more air gaps would be better for cooling ?

 

[doc007]

Less air inbetween will definitely make for better cooling. Less delta-T between the stator to case with
A lower volume of air = better stator cooling.

In the original NC hub design, there is a silly amount of wasted space due to the depth of each side cover.

Justin how do you think this will affect cooling for those of us pumping alot of current into these motors ? You'd think the older design with more air gaps would be better for cooling ?

LFP- That's good to hear that it would run cooler. I would have thought the opposite. Any idea on how much cooler it would run?

Justin- When will these motors be available to us?

 

[liveforphysics]

Nope. Not a magician, and don't have the specs, and really wouldn't want to have to input all the shapes into the CFDTA software either. lol

You close down the air gaps on things, and decrease the air space, you increase the rate of air contact and air turn-over between the outside and the stator, you decrease the temp difference between them, making the outside of the case hotter for a given stator temp, which means better ability to transfer that heat into the outside air passing by it. Less air volume, higher air motion and turblence = lower thermal resistance = higher continous power levels. In this case, it's going to be maybe a 5-15% type improvement(total WAG), nothing to get wild about, but a step in the right direction.

 

[doc007]

Nope. Not a magician, and don't have the specs, and really wouldn't want to have to input all the shapes into the CFDTA software either. lol

You close down the air gaps on things, and decrease the air space, you increase the rate of air contact and air turn-over between the outside and the stator, you decrease the temp difference between them, making the outside of the case hotter for a given stator temp, which means better ability to transfer that heat into the outside air passing by it. Less air volume, higher air motion and turblence = lower thermal resistance = higher continous power levels. In this case, it's going to be maybe a 5-15% type improvement(total WAG), nothing to get wild about, but a step in the right direction.

haha thanks for the lesson, LFP. you're explanation makes sense as does your s/n. would that mean if these things were run under vacuum they would produce even less heat?


Now the question is, when can we start buying these puppies? ... I was going to say Justin must be busy testing out his electric winch. The hat threw me off, its actually bzhwindtalker's.
Anyway, Justin if you are out there what do you think the timeframe will be for these motors?

 

[amberwolf]

would that mean if these things were run under vacuum they would produce even less heat?

Well, in that radiation in a vacuum would let less heat out compared to radiation+convection in air or other media, then yes. :lol:

But they would get hotter inside, as less heat could escape. The reason that a lower volume of air inside should make it run cooler is because the outer plates would be that much closer to the stator/etc, and radiation and convection would have less distance to transport the heat over, making the trip faster and cooling more efficient.

Ideally, having direct conduction to the outside would be better, but since the outside is rotating and the inside is not, we can't do that.

Next best would be convection of a liquid coolant with a high heat-carrying-capacity, but that will add extra drag between the rotor and stator, and decrease mechanical efficiency (maybe not much, but it would thus add heat--hopefully less than it helped take away, still increasing heat-removal efficiency). It also requires anticorrosive additives or a liquid that doesn't corrode in the first place, and also one that doesn't dissolve the insulation on wires, components, or windings. It also requires very good seals on all bearings and cover edges, as well as the axle-hole for wires.

Next best would be *pressurized* air (or a neutral gas of some type; perhaps nitrogen) to increase convection efficiency, but maintaining pressure would be even more difficult than with keeping the liquid from escaping.

So we're left with normal air convection at whatever pressure your local air is at, with no seals to keep it from changing. So the best we can do with that is to keep it stirred up inside, convecting between the stator out to the rotor, and keeping the space between those as small as possible.

 

[doc007]

would that mean if these things were run under vacuum they would produce even less heat?

Well, in that radiation in a vacuum would let less heat out compared to radiation+convection in air or other media, then yes. :lol:

But they would get hotter inside, as less heat could escape. The reason that a lower volume of air inside should make it run cooler is because the outer plates would be that much closer to the stator/etc, and radiation and convection would have less distance to transport the heat over, making the trip faster and cooling more efficient.

Ideally, having direct conduction to the outside would be better, but since the outside is rotating and the inside is not, we can't do that.

Next best would be convection of a liquid coolant with a high heat-carrying-capacity, but that will add extra drag between the rotor and stator, and decrease mechanical efficiency (maybe not much, but it would thus add heat--hopefully less than it helped take away, still increasing heat-removal efficiency). It also requires anticorrosive additives or a liquid that doesn't corrode in the first place, and also one that doesn't dissolve the insulation on wires, components, or windings. It also requires very good seals on all bearings and cover edges, as well as the axle-hole for wires.

Next best would be *pressurized* air (or a neutral gas of some type; perhaps nitrogen) to increase convection efficiency, but maintaining pressure would be even more difficult than with keeping the liquid from escaping.

So we're left with normal air convection at whatever pressure your local air is at, with no seals to keep it from changing. So the best we can do with that is to keep it stirred up inside, convecting between the stator out to the rotor, and keeping the space between those as small as possible.

I should have never taken physics. The endlessphere forum would have sufficed. :wink:

 

[amberwolf]

Fizzix? Yeah, I guess I picked up a little here and there. I might not have it all correct, though. Odd, but lots of the things I've learned came from older sci fi books where the sci was more important than it is today. Rest of it came from highschool science classes, PBS, or other books skimmed thru for various reasons, but I don't think I could actually pass a physics class. Too much math; I would fail.