Hub "Mid Drive" Build

[docnjoj]

Superglue works good for washer stacks too!
otherDoc

[LI-ghtcycle]

Ahh, yeah I can see how that would! I really lucked out with the shrink, didn't even think about this, but I had enough on each side that it covered the ends too so no worries on painted surfaces.

I got the drop-outs all painted, built some torque plates for a friend yesterday, so put this job on the shelf for a bit, but I am back to work today, more pics coming soon.

[LI-ghtcycle]

More pics, kinda boring, but more just the same:









I have hit a bit of a snag, it looks like the chain line won't clear the exercise bike's drop-outs as easy as I thought, will have to make some adjustments.

[LI-ghtcycle]

Turns out it might not be that hard after all, I just have to run the chain at an angle instead of straight down, and it should clear things fine. The only concern now is if the rack that I am using for the mock-up might interfere with the chain line.







I will try and test fit the mock-up tonight, but my energy level is pretty low, might have over-done it last night, so today I am having to take it easy. (this is particularly hard for me since I am normally high-energy and able to do much more!)

[LI-ghtcycle]

Made some progress this week, had a visit with the neighbor to get some quick welds made for my proto-type:





This eliminated the problem caused by the diagonal drop-out angle and wasn't too hard to do, as a bonus, this made the distance from the mid drive drop-outs to the Nuvinci shorter too.





With these drop-outs, I don't need to worry about chain tension since I can move the drive up and down to get ideal tension.



Had to extend the chain about 8 links beyond a standard length single speed chain, but that will change in the final version where the drive hub will be just beyond the rear wheel going towards the back, freeing up the top of the rack for cargo.





There is some concern around the derailleur, tight clearances, might have to do something different if it becomes an issue.

Just in time for the snow!

Actually here that means cold rain with snow, and soon will be black ice. Hopefully we get some better weather soon, hate to test this in sleet/black ice!

[dogman]

Just nit picking here, maybe you are slim? But it looks to me like the hub rotating is going to rub a hole in your pants.

Mabye a fender of some kind on the finished bike?

[LI-ghtcycle]

Yes, this is just for no-load testing, I have an old fender in mind to prevent that and a coat getting caught.

[docnjoj]

Hey LI-ghtcycle! That is a nice piece of work. I do have a question about having 13 lbs of rotating mass up that high. Have U noticed any problems with handling or gyroscopic reaction?
otherDoc

[LI-ghtcycle]

Hey LI-ghtcycle! That is a nice piece of work. I do have a question about having 13 lbs of rotating mass up that high. Have U noticed any problems with handling or gyroscopic reaction?
otherDoc

Thanks!

Unfortunately I won't have a good answer for probably a few days, today was looking encouraging, but now the sleet/hail is coming down and it's just above freezing, so soon it will be icy.

I got to the garage and replaced the front wheel, and need to solder the phase wire connectors (combining the 9x7 with a different controller, same halls, different bullet connectors for the phase wires) but I'm not very energetic today, so might be a little bit before I get even some no-load testing in.

[katou]

I am curious to see how this will look when it's all done. Do you have a drawing of how it will look after the mechanicals are done and cargo pods are attached?

Was there a reason not to put the hub motor in the triangle? I would have thought that a natural place to put it.

No one has tried exactly this arrangement, I'm very interested to hear your findings after the first ride!

Katou

[LI-ghtcycle]

Yeah, I would like it in the triangle area ideally but this frame is too compact for it to fit. That is also the reason I have looked at the possibility of using a cyclone motor, but it's probably not ideal since it will be running a gear box, and I don't know if it will allow for regen (this has become a major consideration now that I realize that it will save my brake pads!) but who knows? I will also being going a lot slower than people like John in CR, so maybe brake wear won't be as big of a consideration.

I should have it together today for some thorough testing, now that the roads are clear, it's still cold however, and until the garage is at least 50 DEG it's pretty miserable trying to work in there (it's about 45 DEG last I checked, we have a space heater that will do the job, and a propane heater that is great, but I can't really afford the propane, we had a pellet stove, but it's not working anymore and it's a giant garage almost bigger than the house! too bad it's so full of junk! )

I will be taking my father to buy a new electric chain saw, hopefully when I get back it will be warmer, I guess I could try soldering with gloves!

[LI-ghtcycle]

I am curious to see how this will look when it's all done. Do you have a drawing of how it will look after the mechanicals are done and cargo pods are attached?
Katou

I have a general idea how I want to do the cargo pods, mostly I want something more aero than just slapping on some panniers, but my biggest concern is clearance and keeping it as easy as a normal touring bike to maneuver around, it will be hard accomplishing that with the DD hub motor as mid drive, even if I put it to the back since it will make it hard to stand the bike up to turn it around and such, but I guess that will still be difficult just with the weight, just trying to avoid going beyond the normal length of a standard 26" bicycle.

[LI-ghtcycle]

Well, after realizing that the newer controller that I had won't work with the Amped 9 x 7 that I have, I have installed the original controller, and noticed something was wrong, some how the 17T cog had back it's self off the hub while I had been messing with the Forsen controller!!

Unfortunately it and the lock ring backed completely off and took some of the threads with them, in fact the threads that the lock ring will ride on are pretty much gone!!

If I do end up going this route for the final design, I will have my local machinist install a steel insert so I don't have this problem again. For now I have tightened the cog on as tight as I can (I had put anti-seize on the threads, looks like that might have been a mistake ) and then just tightened the lock ring as much as I dared, and tomorrow will be the "maiden voyage".

My no-load speeds have been considerably faster than calculated, top speed with the 'Vinci on max over-drive was 38.7 MPH!!

Max under-drive was 11 MPH which brings into question my RPM calculation for the 9 x 7 at 48V, it will be much lower speed on the road I am sure, and I suspect that the actual loaded hill climbing speed will be closer to 7MPH on our local 10-12% hill with no pedaling, but hey, if it will do that with minimal stress, that is still good news, and although I wonder how fast it will really go (even if I have the gearing, the motor will probably run out of torque and start to bog down around 30MPH) on the flat at 44V (50V hot Turnigy Nano LiPo 25 - 50c) 12S, but in the end I plan on using 14S, so I might need to go to a 13T cog on the motor to slow it down a bit. I have no desire to go 38 MPH on this bike, even with out it's touring load, that is just too fast for how this bike will be ridden.

It is good information for a future built which WILL use that kind of speed, but will be licensed appropriately.

[LI-ghtcycle]

Well, it's sunny and 23 DEG!!!

But as far as I know, the roads are clear, so I am going to test ride the bike!

[LI-ghtcycle]

Well, now I understand probably the main reason that the stoke monkey is a LH drive. Since I am using the motor, or hub to turn the sprocket, instead of a chain turning the hub, the motor's torque is now un-threading the sproket!

So now I need to either use a jack-shaft so I can use the left side, or figure out a better way to secure the sprocket. I might try using some lock-tight, but I have a feeling that won't hold.

If I had a steel insert, I might have been able to really torque down the lock ring so it wouldn't matter. I am thinking that the best would be a hub motor with disc brake spin-on adapter that could be modified to be a permanent attachment on the RH side.

For now I will just clean things up best I can, and lock-tight and see what happens.

[Rassy]

Well, now I understand probably the main reason that the stoke monkey is a LH drive. Since I am using the motor, or hub to turn the sprocket, instead of a chain turning the hub, the motor's torque is now un-threading the sproket!

Yep, That's why my mid-drive hub motor is "backwards" and the chain wraps under the sprocket. When my first freewheel gave out I did exactly what you did, but the lockring only had to keep the sprocket from unscrewing when I was pedaling without using the motor. I had used lock-tite and it worked fine until I got the ENO, but even after using it for a while it was still pretty easy to get it off, so without some sort of a key lock I wouldn't expect it to work against the power of the motor.

[LI-ghtcycle]

Well, now I understand probably the main reason that the stoke monkey is a LH drive. Since I am using the motor, or hub to turn the sprocket, instead of a chain turning the hub, the motor's torque is now un-threading the sproket!

Yep, That's why my mid-drive hub motor is "backwards" and the chain wraps under the sprocket. When my first freewheel gave out I did exactly what you did, but the lockring only had to keep the sprocket from unscrewing when I was pedaling without using the motor. I had used lock-tite and it worked fine until I got the ENO, but even after using it for a while it was still pretty easy to get it off, so without some sort of a key lock I wouldn't expect it to work against the power of the motor.

Yeah, I am thinking the realistic solution will be a splined cog, and having this now damaged thread converted into a splined shaft with a similar pattern as the Nuvinci's freewheel adapter.

I would be still unable to take advantage of regen braking even if I just ran it LH drive, so I'll see what I can work out.

I got some heavy duty lock tight, but I'm not going to trust it unless it holds up to about as much force I can put on it by hand with a chain whip to take it apart, and even then, it's tenuous since I know that the hub motor will put out much more torque!

[LI-ghtcycle]

Well, I am going to try it out, the lock-tight seems to be holding, but I think in the end I will either end up using some form of splined drive, but if the lock-tight will hold, I can live with that.

I also like how the track hubs do it, having a left hand threaded lock ring that contacts the track cog in such a way that if it tries to loosen, it just tightens up the lock ring, however, that would require turning the hub case, and if I can avoid going to that length, I will.

Most likely I will end up using something like either the curry electro drive or other motor that normally turns a bicycle chain, possibly a cyclone motor (the 1000-3000 Watt one with no gear head) but for now, I will test this one out.

[amberwolf]

You could just dremel out a deep rectangular groove in the threads, parallel to the axle, on both the hub threads and the sprocket (or freewheel) threads, and then insert a steel block into that groove. It'll act as a locking pin to keep them from moving.

Make it a bit longer than the groove so you can pull it out with vise-grips or similar, and you can then relatively easily remove it for repairs or gearing ratio changes. Use loctite on it to keep it from just falling out if you like, but mostly the forces across it's short side by the opposing faces of the groove will prevent it from moving along it's length.

[LI-ghtcycle]

You could just dremel out a deep rectangular groove in the threads, parallel to the axle, on both the hub threads and the sprocket (or freewheel) threads, and then insert a steel block into that groove. It'll act as a locking pin to keep them from moving.

Make it a bit longer than the groove so you can pull it out with vise-grips or similar, and you can then relatively easily remove it for repairs or gearing ratio changes. Use loctite on it to keep it from just falling out if you like, but mostly the forces across it's short side by the opposing faces of the groove will prevent it from moving along it's length.

Yeah, I could do that, but if I get that involved, I would want a more "off the shelf" replacement sprocket, otherwise if I broke down on the road, I would need to dremel out the replacement sprocket too.

[LI-ghtcycle]

Got some rides in finally!

Really put it through the paces, took it across the street to an extra long drive-way that is well paved and was a bit disappointed, the flat road speed is the same, the motor just can't give me any more than it's already got at this voltage (12S, 44V of 25-50C Nano Tech LiPo).

I should have realized that, but seeing the 38 MPH no load speed got me excited!

Now, the important part! Climbing the hills. Fortunately for me, I have a 8% grade ravine about 2 blocks from my house, so I could do some hill climbing in the lower gears, and again, the motor is the same, so it didn't really climb any better, maybe gave me 1 MPH more going up the hill, but as JRH's says:

Volt up, Gear down!

I am going to do this very thing, I am thinking that I should have a no-load speed closer to 28-32MPH on the high end, and that will bring my low end gearing down where I need it, 10 MPH on the bottom end isn't really what I was looking for, so I will be buying another cog or 2 (hopefully used this time!) next I will try a 12T or 15T depending on what they have used.

***EDIT**** Well, looks like the smallest track cog nominally I can buy is probably 13T, so that's not going to work especially if I volt up, so I may indeed look at the possibility of going for a slower winding motor like an 8 x 8 if I stick with a hub motor, because I'd rather have more "low gearing" than high.

[John in CR]

Here's my LH 9C mid drive solution. Note that once I test run and I'm happy with the gearing, if I think I need any more rear brake than regen, then I'll use longer bolts and spacers and put my 203mm disk on. That rear 44t is 183mm, so I may be able to squeeze my caliper. If not, I'll just go down a size. My front is a 28t giving me effectively a 13.5" wheel for the hubmotor with the 3"x20 Kendas (21.25" OD) I'm running on the bike. I'm gearing it so low because it's getting a high voltage controller with 148V fresh off the charger, so I should get billy goat climbing with a 9x7 in a 13.5" wheel and plenty of current, and a top speed of about 45mph.

I'm using chainring size sprockets so I can run bike chain without concern, because there's so much less force on the chain than something like a Cyclone that eats chains. Once I dial it all in, I'll probably go to a 218 chain, but this will do for now and gives me regen. Note that a 3rd hand is very helpful for clamping the sprockets to the cover and brake disk for accurately marking the holes and a caliper a must, at least for me.

The holes in the 9C cover are for centrifugal flow ventilation. I expect to push near 10kw into it, so it will need cooling, and turning over 1krpm the motor will be its own fan blowing right at the windings.

9C mid drive sprocket LH.JPG

(36.41 KiB) Downloaded 78 times

[spinningmagnets]

a "keyed shaft" is actually a great solution for a standard RH freewheel on a LH-sprocket mid-drive. An ENO freewheel could have the groove cut in it with a standard hardware store file. The square cross-section pin is called a key, and two of them would hold enough torque that you'd never come close to breaking them at the power levels you have mentioned.

[LI-ghtcycle]

Here's my LH 9C mid drive solution. Note that once I test run and I'm happy with the gearing, if I think I need any more rear brake than regen, then I'll use longer bolts and spacers and put my 203mm disk on. That rear 44t is 183mm, so I may be able to squeeze my caliper. If not, I'll just go down a size. My front is a 28t giving me effectively a 13.5" wheel for the hubmotor with the 3"x20 Kendas (21.25" OD) I'm running on the bike. I'm gearing it so low because it's getting a high voltage controller with 148V fresh off the charger, so I should get billy goat climbing with a 9x7 in a 13.5" wheel and plenty of current, and a top speed of about 45mph.

I'm using chainring size sprockets so I can run bike chain without concern, because there's so much less force on the chain than something like a Cyclone that eats chains. Once I dial it all in, I'll probably go to a 218 chain, but this will do for now and gives me regen. Note that a 3rd hand is very helpful for clamping the sprockets to the cover and brake disk for accurately marking the holes and a caliper a must, at least for me.

9C mid drive sprocket LH.JPG

Nice! Just curious, is the reason for using smaller wheels as much for climbing as it is acceleration? I'm going to have to re-evaluate my gearing if the combo I have now has any trouble climbling hills with a fully loaded bike, but otherwise I might just go with a slower winding to make up for my limited gearing choices. I could always have Staton make a custom 44T Nuvinci FW adapter sprocket too.

You know, on second thought, I think I have the gearing right after all! I forgot to figure in for voltage sag and minimum voltage, i.e. 3.4V per cell, x 12s = 40.8 V, and that puts my gearing right back on target! I do need to try this with a full load however, until then, this is just conjecture.

I'm liking your set up John, I will have to see how it goes for ya, I think that would be a great set-up for my licensed E-Bike!

[LI-ghtcycle]

a "keyed shaft" is actually a great solution for a standard RH freewheel on a LH-sprocket mid-drive. An ENO freewheel could have the groove cut in it with a standard hardware store file. The square cross-section pin is called a key, and two of them would hold enough torque that you'd never come close to breaking them at the power levels you have mentioned.

Yeah, that would be a solution for sure, but I still think having my hub machined to use the same standard FW right-hand thread and LH threaded locking ring (slightly smaller diameter so that it is on the same shaft).



http://www.sheldonbrown.com/gloss_tp-z.html#track

In fact, now that I think about it, it probably wouldn't be that hard to find a trashed track hub that still has good threads on it, and have some form of machining done on my hub motor cover so that it could be press-fit onto it!

That would be a ton easier than having him turn the threads onto my hub cover, and I could get a steel hub (I hope!) so that I would have stronger threads!