GNG, 1000W 48V BB-drive, $400

[Ebikeman]

While I am waiting the arrival of some of the LighteningRods side sheets I ordered. I have been working on the drive pulleys. So far, I have managed to mould and resin cast the 14 to 18T slip on adapter, seems ok, and fits snugly on the motor. I tapped a 5mm thread in the motor shaft and fastened an end plate on to retain the adapter.
Now I am making a mould to cast a slip on adapter for the big pulley which will be 107T. This will give a slightly improved ratio compared to the original 80/14 pulleys. Until I can test the casting out under load I am not sure if they will stand up to long term use.
I used a slow cure casting resin with aluminium filler. I hope this will give a hard none brittle cast that can resist wear. The tooth profile has come out well with a nice smooth surface which should help to reduce belt wear, especially when compared to the original GNG belt/pulley setup.

My suspicion is that I will finish up having to go for a primary chain drive mod at some stage. I have all the bits and bobs for the chain drive but until I know the belt & pulley is not going to work I don’t want to file off the motor shaft.

 

[rp3]

Nice work Ebikeman, let us know if you end up wanting to sell a set.

...

 

[speedmd]

easy fix. Times 18.

 

[speedmd]

Someone who ordered my primary adjustable sheets just sent me this picture of the "New and Improved" Gen 1 GNG. They didn't make the sheets adjustable but they did change the design and eliminate the alloy bearing blocks that held the old design together. I'm asking the member to send the upper sheets and jackshaft to me so that I can see what modifications I have to make.

Couldn't they leave bad enough alone?

Yes, still no adjustment but much improved with the bearing barrel spacer / frame support. I added a center alloy spacer to my rebuild and was also able to eliminated the other two cross supports. Looks like all steel so it should be good and solid.

Also I noticed today that the stock spacer behind the little 12t freewheel is much too thin walled to give the shaft proper support.

 

[redstone02]

Hello,

I need help with selecting a bottom bracket tool. What type is needed for the 1st gen GNG?
How about this one:
http://amzn.com/B00811WQJS
?

This for the square tamper crank:
http://amzn.com/B001VS1M20

or

http://amzn.com/B00BUL4K54
?

When first fitting up the GNG, the bike shop did the work. It would be nice to have the tools on hand.

Thanks,
Tom

 

[Rassy]

Those tools you referenced look like the right ones to me. My crank puller doesn't have the handle, so I have to use a wrench, but it works great and was a little bit cheaper.

Edit: I missed your third link. That's the crank puller I have.

 

[spinningmagnets]

Not the first one, but either one of the last two would work. edit: this pic is to show two methods. I wanted to swap out a bad crank-arm, so I ordered the proper tool for future fixes (the black bolt-looking device) and I immediately used my cut-off wheel in a cordless drill. I cut through 90% and then pried the last 10% into snapping apart.

http://endless-sphere.com/forums/viewtopic.php?f=3&t=18157#p267415

You can see that the multiple splines on the first one pulls out the BB bearing cup from this style of cartridge:

 

[christerljung]

This is my first 3d-printout of a Herringbone reduction.
Unfortunately the big sprocket needs about ten more teeht. Gonna walkthrough a bunsh of Solidworks tutorials and find out how to make them perfect.

 

[speedmd]

Possibly a good strong quite design for some modeling.

 

[LightningRods]

Here's another shot of GNG Gen 1.1. The main point of the mods seem to be protecting the jackshaft and drive from castoff from the front wheel.

Forum member hillyterrain is sending this GNG to me to design a revised set of adjustable upper sheets. I'll take detailed photos and then disassemble the top end to see how it all goes together. I'll also post dimensional drawings. Feel free to ask for specific views and information and I will post it here.

 

[redstone02]

Here's another shot of GNG Gen 1.1. The main point of the mods seem to be protecting the jackshaft and drive from castoff from the front wheel.


Forum member hillyterrain is sending this GNG to me to design a revised set of adjustable upper sheets. I'll take detailed photos and then disassemble the top end to see how it all goes together. I'll also post dimensional drawings. Feel free to ask for specific views and information and I will post it here.

Hello LightningRods,

With these new sheets, are you working toward the 120T pulley?

Thanks,
Tom

 

[redstone02]

Here's another shot of GNG Gen 1.1. The main point of the mods seem to be protecting the jackshaft and drive from castoff from the front wheel.


Forum member hillyterrain is sending this GNG to me to design a revised set of adjustable upper sheets. I'll take detailed photos and then disassemble the top end to see how it all goes together. I'll also post dimensional drawings. Feel free to ask for specific views and information and I will post it here.

Hello LightningRods,

With these new sheets, are you working toward the 120T pulley?

Thanks,
Tom

.... sorry not enough information....
Enough room for the 120T and 18T pulleys?

 

[LightningRods]

With these new sheets, are you working toward.... Enough room for the 120T and 18T pulleys?

Yessir. If I have to completely redesign the upper sheets they will have enough extension to comfortably fit a 120 tooth driven pulley.

 

[speedmd]

Way more solid and stubby looking than the orig. Do like the crash pan. I have decided to mod the bb mount on mine. I am working on making a mount now that will mount to outside of bb shell and not need the gnc BB or rely on the bb threads to hold motor. May not be a universal solution, but should be solid and adjustable.

 

[spinningmagnets]

If using my recommended 13T freewheel and the 60T chainring (4.6:1 secondary reduction):

Several builders plan on trying the 120T:18T pulley set, = 6.7:1 reduction = total 30.6:1 reduction.

pedal-cadence unloaded/loaded
48V =______________105/84
36V =_______________79/63

 

[bee]

While I am waiting the arrival of some of the LighteningRods side sheets I ordered. I have been working on the drive pulleys. So far, I have managed to mould and resin cast the 14 to 18T slip on adapter, seems ok, and fits snugly on the motor. I tapped a 5mm thread in the motor shaft and fastened an end plate on to retain the adapter.
Now I am making a mould to cast a slip on adapter for the big pulley which will be 107T. This will give a slightly improved ratio compared to the original 80/14 pulleys. Until I can test the casting out under load I am not sure if they will stand up to long term use.
I used a slow cure casting resin with aluminium filler. I hope this will give a hard none brittle cast that can resist wear. The tooth profile has come out well with a nice smooth surface which should help to reduce belt wear, especially when compared to the original GNG belt/pulley setup.

My suspicion is that I will finish up having to go for a primary chain drive mod at some stage. I have all the bits and bobs for the chain drive but until I know the belt & pulley is not going to work I don’t want to file off the motor shaft.

That's a clever way of mounting the slip on pulley! Wish I had thought of that

This is my first 3d-printout of a Herringbone reduction.
Unfortunately the big sprocket needs about ten more teeht. Gonna walkthrough a bunsh of Solidworks tutorials and find out how to make them perfect.

You won't be able to modify the tooth count without wasting a lot of time if you're working with the STLs and solidworks. If you load up the openscad source code you just need to add 10 to the tooth count parameter and it will automatically generate a new model for you.

That's exactly what I'm doing in the first few seconds of this video:

[youtube]btz9jiiJcJs[/youtube]

Just gotta change the number and hit render

Working with STL files is generally a huge pain, it is similar to trying to change how a program works using only binaries. You want the source code to the model, whatever file format it was in before someone exported the STL.

Also I have a feeling the tiny little set-screw on the small gear won't last very long. Those things are fragile enough when used with a little nema17 motor on the 3d printer, I can't imagine them holding up to the torque of the GNG motor. With the stock 14t HTD motor shaft, we have a fairly convenient spline for 3d printed objects to attach to.

If you're feeling crazy, speedmd has a great idea:

Possibly a good strong quite design for some modeling.

There are a ton of easily adaptable planetary gear designs like this on thingiverse: http://www.thingiverse.com/search?q=planetary

I love the ones that use herringbones like http://www.thingiverse.com/thing:34778 since herringbone gears are apparently patented, and patents can become meaningless if enough infringement happens and the patent owner doesn't do anything to enforce their patent. Forced open source hardware!

 

[rp3]

Here's another shot of GNG Gen 1.1.

Does the belt setup look the same?

 

[gknowes]

spinningmagnets, using your recommended setup. will your pedal-cadence be good/ok or right on the money. also is there a link to learn all about pedal-cadence. THANK

 

[mwkeefer]

If using my recommended 13T freewheel and the 60T chainring (4.6:1 secondary reduction):

Several builders plan on trying the 120T:18T pulley set, = 6.7:1 reduction = total 30.6:1 reduction.

unloaded pedal-cadence/loaded
48V =______________105/84
36V =_______________79/63

My math or perspective might be wrong here but does this equate to:
80% Efficiency at 48v under load vs 79.74% Efficiency under load making 36v the more efficient voltage for the motor but also letting us know to expect an overall 80% ish range from the motor regardless the condition?

Just curious

-Mike

 

[spinningmagnets]

Mike, the numbers I posted are a very rough estimate, using a 20% reduction in RPMs between unloaded (wheel in the air) and loaded by pulling the bike and rider along. Actual numbers may vary quite a bit. I only meant to help prospective buyers understand the approximate results of the various mods.

I am anxious for actual numbers to be posted, as many of the most desirable mods have not yet been fully implemented. The real world results will allow everyone to evaluate the relative worth of the various mods.

gknowes, I am not personally worried about pedal cadence. If you raise the voltage of the system (the GNG has been used at 72V), then the motor-RPMs go up. With the stock 44T chainring at 48V, the system will spin faster than just about anybody will pedal. If you are off-road, it would be like riding an electric dirt bike, and trying to pedal will add nothing, because your pedaling would feel like the rear wheel is up in the air. At that point, the pedals are pretty much foot-rests.

If your pedaling can add some work to the bike, the battery will get more range (good for commuters). You can do this by various methods. One way is to lower the motor RPMs by using 36V. If you want to keep 48V, you could use a larger pulley at the jackshaft (this is why there's discussion about the 95T, and 120T pulley), and you could also change the Bottom-Bracket (BB) chainring from 44T to something bigger (like 52T or 60T).

Many bike riders seem to like 80-RPM at the pedals. Your personal tastes may prefer a little lower or higher, but 80-RPM is a good starting point (if you desire to pedal with the motor)

 

[gknowes]

thank's spinningmagnets i have a 36v. i added lightingrods upper and lower sheets. also just installed a 95t with 535l belt ,i have no slipping on the 14t drive so far. after your post i starter thinking about my next step . working on the secondary drive system.

 

[spinningmagnets]

I highly recommend you sand off the sharp edges of the motor-pulley, if the teeth are still pretty sharp, the belts will wear out very fast. At 36V, it would be better for them to be just a little too rounded than too sharp.

Perhaps sand them just until they seem to fit the curve of the belt teeth as well as possible.

 

[gknowes]

thinks spinningmagnets ,i filed down the teeth but i think your wright i will need to sand a little more

 

[Miles]

since herringbone gears are apparently patented................

What?????

As they've been used since the 19th century..................

 

[bee]

since herringbone gears are apparently patented................

What?????

As they've been used since the 19th century..................

Unfortunately it's totally ok to file invalid patents, and companies will do it in the hope that it'll scare an engineer from giving a second thought to their idea. A quick google search shows that there are dozens of patents for various herringbone gear implementations.

The best way to erode the validity of these patents and open up the knowledge for the greater good is to release and use open source derivatives