E.S. bottom bracket drive.

[Suspect]

I have a bike that's close to this design. I've been thinking about doing a Kickstarter project to see if there was interest in the mid-drive approach.
I have a freewheel on the bottom bracket and a MAC like motor with a planetary reduction.
Here are some rough specs:
300 W
24 V, 12 Ah battery
Geared low for hill climbing so top speed is ~18 mph. Low gear top speed is ~5 mph. Personally I prefer the mountain gearing but if flats and speed were preferred I could adjust accordingly, however, I would hit a motor power limit fairly quickly which would require some beefier gears (to hit speeds in the 30 mph range).
~20 mile range
On trek 4200 weight is 42 lbs with battery
The gears are tuned to pedal cadence.

I've made a few revisions of this basic setup. The latest adapts to the bottom bracket and I'm in the process of assembling this version. The version I currently ride required me to weld some tabs to the down tube. I've put lots of miles on this version.

It's a tough setup. I take it on trails and bash it hard. I love it, though it's not without shortcomings.
I would like to switch to an IGH hub, and I have had some cassettes lose teeth due to late shifts while under full power.

Also pictured are my very original version with a chain drive. It was loud but still surprisingly reliable. I put ~4000 miles on that setup. The CAD is for the current build I'm working on.

 

[full-throttle]

Thought you might find the size/power specs of these motors interesting? They do look like big versions of the GNG motor - but I haven't checked speed specs etc. Anyways here it is:
http://www.motenergy.com/brdcmo2.html

Those ME020101xx look interesting: 149.2mm dia, 107.4mm wide (+ shaft, terminals)
What do they weight?
What ever happened to the Mars Electric LLC??

 

[Samd]

Actually check out the me1016 alternator on the homepage. 1000/12 is about a 83 Kv inrunner. Take off the fan and it looks narrower.

The lower rpm per volt and narrow design could help.


No hall sensors tho using an alternator as a motor of course.....

 

[Miles]

Actually check out the me1016 alternator on the homepage. 1000/12 is about a 83 Kv inrunner.

If you look at the drawing you can see that it's an outrunner. Perhaps you just meant radial flux?

 

[neptronix]

Thought you might find the size/power specs of these motors interesting? They do look like big versions of the GNG motor - but I haven't checked speed specs etc. Anyways here it is:
http://www.motenergy.com/brdcmo2.html

Those ME020101xx look interesting: 149.2mm dia, 107.4mm wide (+ shaft, terminals)
What do they weight?
What ever happened to the Mars Electric LLC??

They don't list the weight, but it is of similar dimensions of the 4201 series, so i would assume 20-22lbs ( ~10kg )

MARS changed their name to Motenergy due to a trademark dispute over the name.

 

[crossbreak]

A 420 chain derailleur would be awesome. sadly I' found nothing in the net.

Why should the ES BB drive have more than 1000W continuous? Isn't that enough to empty the battery? I'm more keen to see a drive that is lightweight and well cooled than an overpowered device that fails within a few hundred miles. A 12Fet controller@ 45amps max can suck dry my battery so fast, I doubt i'd ever build something even better sucking for my bike.

My dream motor would be:
enclosed outrunner, with cooling fins
80mm rotor
20p 18n
0.3mm laminations
70mm over all length
75KV
hall sensors

the 600W mac scooter motors come close, but the laminations are thick, nothing that can be run at high rpm The 50cc turnigy rotomax seems to be perfect. There is just much rework to be done

It's so sad, I feel like I searched the whole net, but did not find a motor that is like the GNG one. Maybe we should just order from china all together. Btw. does anyone know yet the specs of the motor (pole count, inrunner/outrunner or axial flux and so on?)

 

[Suspect]

The current gear housing and planetary sets I'm working with require a 8 mm OD D-shaft from the motor input. Torque is limited to ~8 N*m.
Does anybody know of a source for motors with the same 8 mm D-shaft output but with higher power capability?
Does anybody know of a motor that comes with an 11 tooth spline that would insert directly into a matex planetary set? The 11 tooth spline input gear sets have a much higher capacity.

Currently it costs me ~$400 to build this gearbox plus I spend another $100 for the motor. If I could build ten at a time it would cost $300.

 

[wildharemtbkr]

Are the torque figures on those matex gears the input torque or the output torque? I looked into those a couple years back, and the guy said it was the output toque. That is fairly low once you factor the reduction into the motor torque figure. Example 9.55/72.72 kv of the motor pictured? = 0.1322714 x 30amp (stock controller) =3.968142 which is under all the reductions, but if it is the output then we need to times the reduction into it. Mine is 4.8:1 4.8x3.968142=19.047081 which is over my particular unit, and all of the D shaft input ones. Maybe my math is wrong somewhere.

 

[Suspect]

If they are not input values then I've been lucky. My current build runs two stages. The second stage has the male spline input with a rating of 22.5 N*m.
I've got spares at home, maybe I'll do a destructive test with one.
I've ridden 100's of hard miles with no troubles. The gears and lubrication don't show signs of wear.

 

[Samd]

Actually check out the me1016 alternator on the homepage. 1000/12 is about a 83 Kv inrunner.

If you look at the drawing you can see that it's an outrunner. Perhaps you just meant radial flux?

Indeed - my bad. It would be radial flux. But with that can radius, gearing would be impossible. The concept of the me1016 is no good as an outrunner. Too bad.

 

[Suspect]

Are the torque figures on those matex gears the input torque or the output torque? I looked into those a couple years back, and the guy said it was the output toque. That is fairly low once you factor the reduction into the motor torque figure. Example 9.55/72.72 kv of the motor pictured? = 0.1322714 x 30amp (stock controller) =3.968142 which is under all the reductions, but if it is the output then we need to times the reduction into it. Mine is 4.8:1 4.8x3.968142=19.047081 which is over my particular unit, and all of the D shaft input ones. Maybe my math is wrong somewhere.

I decided against doing a destructive test because I doubt the failure mode is the same for a static load test. I did however check ES for references to Matex and found a lot of evidence that the planetary units should be adequate. There were a few references of using them for years with decent power.
http://endless-sphere.com/forums/viewtopic.php?f=28&t=28761&p=659492&hilit=matex#p659492

 

[Miles]

They are output torque values.

If you look in the Matex eCatalogue, it gives a nominal maximum and usually a graph showing the recommended de-rating for speed.

Example: http://www.matex-japan.com/e/catalog/lgu75-mlg8.php

 

[Suspect]

They are output torque values.

If you look in the Matex eCatalogue, it gives a nominal maximum and usually a graph showing the recommended de-rating for speed.

Example: http://www.matex-japan.com/e/catalog/lgu75-mlg8.php

Thanks for the reference, it clears all doubt.

 

[Geoff V]

Sorry for being a bit later arriving at this party but am fairly new to ebikes, I have a 250w rear hub bike and a 1000w front hub MTB (Ugh).

I do have quite a lot of experience in design and machining and am clear in my own mind that mid drive is the way forward and have therefore decided to make this my next project, being retired affords me the time.

It seems to me that the main problem is trying to achieve the necessary speed reduction by employing chains or belts and I am wondering if a Worm Gear set has been considered. The available reduction ratios easily meet the needs of an ebike and with a 20:1 reduction set, 600w input is rated at 5000hrs without a hardened worm, with an output rating of nearly 38Nm and a reported efficiency of 84%.

Have I missed something along the way? if so please advise before I start manufacturing swarf.

GeoffV

 

[crossbreak]

efficiency of 84%.

No one would use such a reduction. Gears and chains get 98%.

 

[Geoff V]

Sir

I would be very suprised if the GNG primary drive belt with that horrible idler wheel achieves anything like 98% and to make matters worse it only survives a fracton of the time of a worm gear set.

 

[Miles]

Let's make a fair comparison.

The GNG drive, as it stands, is probably less than 90% efficient.

To get get 20:1 you'd need 2 stages of planetary gears or chains and sprockets.

 

[crossbreak]

I would only use belt if we go faster than 6000rpm. we should stick to chains. Worm gears are only suitable if the the auto-stop function is used like in servo mechanisms .

 

[Cyclebutt]

Let's make a fair comparison.

The GNG drive, as it stands, is probably less than 90% efficient.

To get get 20:1 you'd need 2 stages of planetary gears or chains and sprockets.

I may be getting a bit off track here but the efficiency of this kit is surprising. I'm getting lower watt usage per mile out of this thing than I am my rear hub DD. Granted, the DD pulls around 1500w under load and the GNG about 1100w, I'm still getting around the same average speed for road use. OK, I'll have to admit I am better able to offer pedal assist because of gear selection ability. DD pulls around 36w/mile where GNG pulls 27w/mi. These are early figures but I'm keeping data logs on both, we'll see.

 

[Miles]

I may be getting a bit off track here but the efficiency of this kit is surprising. I'm getting lower watt usage per mile out of this thing than I am my rear hub DD. Granted, the DD pulls around 1500w under load and the GNG about 1100w, I'm still getting around the same average speed for road use. OK, I'll have to admit I am better able to offer pedal assist because of gear selection ability. DD pulls around 36w/mile where GNG pulls 27w/mi. These are early figures but I'm keeping data logs on both, we'll see.

Well, this isn't one of the many GNG threads but...... What are your journeys like? Hills? Lots of stop-start? Or, mostly constant speed on the flat?

 

[neptronix]

I've yet to see wh/mi figures, or even watts per continuous speed on the flats for the GNG kit despite asking for it :/
If you could provide some data on that, that would be wonderful.
Maybe post it up on the GNG kit thread too.

 

[Miles]

What would be an interesting control test, would be to do a typical journey without pedal assistance using each drive but using a single gear for the GNG, equivalent to that of your DD hub.

Then do the same journey on the GNG, again without pedal assistance, but using the gears as you would normally.

 

[deecanio]

Hi All,

In answer to the early posts in the thread the rc setup on my kona had several problems, worst offender was the freewheel at the end of the jackshaft meant when pedaling i was circling the 219 chain which was taught (no tensioner) and two bearings before i got to the freewheel, the belt sprocket Matt s made for me was also under tight belt tension resulting in severe drag, unusable without power.
Secondly the motor was too noisy, in a silent forest it was a banshee in comparison to a hub, the high rpm it ran at didn't help but you couldn't pootle around, any slow riding meant a very hot motor, it liked to be thrashed.
Also i didn't like running 2 chains, meant i had to have skinny ass tyres and always worried that the 2 might meet one day
To be fair it ran ok, but this was 3 years ago, i didn't have a current CA but i did have the short run of fecther current limiter which worked superbly on the hv110 (motor was the 3210).

I was hoping that this thread was going to take the doubtless skillset we have communally and apply it to the existing gng, swap out the parts we know are tried and tested and strengthen the brackets for 3kw usage, if the existing users testify to very little drag when pedaling normally i dont see too much in the way of a straight forward es mod kit?

Anyway fingers crossed something comes up, no doubt its mid drive for the the win, especially for those that have dual suss bikes, the best thing about the rc was that i could bounce around without issue.

i've never really understood the need to use reduction to get 100rpm at the crank? in my small mind the best system is one that allows you to pedal normally with no or very little drag, has the ability to haul the rider by itself, or any combination of the two, although i understand how great it is to do both to achieve synergy i dont understand why you would need to limit the rpm at the crank to be equal to the riders potential input, surely you can control this using the throttle?


D

 

[rodgah]

surely you can control this using the throttle

my thoughts exactly mate, it actually makes more sense to have it this way if you want to run more power as you can use the higher teeth count sprockets on the rear cluster which in other words handles the power better anyway.

I was thinking along the lines of the bracketry getting changed so the belt length is a big longer and overall the structure stronger, an add on (extension) to the teeth on the motor sprocket to allow for a 25mm wide belt (if the outer flange is just a press fit one) and a wider recumpence/ffr sprocket to match.

 

[Miles]

If you are happy with the compromise of only being able to contribute meaningful pedal input at low duty cycle.