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

[Chalo]

They're 19mm spindles. Here is the Gusset store link:

http://www.gussetbikes.com/products-bmx-list.php?c=D

Here are the 19mm spindles in various lengths in chrome moly steel:

http://www.profileracing.com/estore/product.php?productId=343

19mm is a very common size, but the Profile spindles are 3/4" (19.05mm)

The bore in the freewheel adapter above looks larger than 19mm. The OD of that thread is about 35mm.

I'm working on a sprag bearing adaptor for these axles.

Cool. Torrington roller clutch, CSK bearing, or something else? No one-way clutch I have ever seen for a 3/4" shaft can withstand bicycle crank torques, so I'm guessing you'll have to mount a sleeve to fit the bore of the clutch.

 

[spinningmagnets]

Do you have a link to that item?

http://endless-sphere.com/forums/viewtopic.php?f=28&t=47618&start=50#p708963

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

also: 19mm is 0.748-inch

Staton-inc FW adapter bore with keyway and two set-screws, $14. (I wouldn't use the bolts in the pic, but at least the threaded holes are there already)

1.000
0.748___19mm
0.252 = sleeve adapter with 0.126-inch thick walls...very doable!

http://www.staton-inc.com/store/pro..._sprockets_1_10_wide_2_set_screws-984-27.html

 

[Chalo]

Do you have a link to that item?

http://endless-sphere.com/forums/viewtopic.php?f=28&t=47618&start=50#p708963

That looks the same as the Sick Bike Parts long bottom bracket. It's for a standard square taper crank, and it's 17mm outside diameter. (On a square taper crank spindle, the flats are tapered 1 degree per side.) It is of no use with the freewheel adapter in question, because it fits a completely different kind of crank.

Edit:

Also, the goobers at Staton Inc need to study some basic mechanical engineering, because they make a lot of boneheaded mistakes like cutting keyways all the way up to the bottom of a thread, or machining a 10mm axle with a 12.7mm dropout clamping face and a snap ring groove cut at the point of highest stress:

This stuff is just retarded. I know they're from Oklahoma so they probably can't help it, but c'mon. Really.

 

[LightningRods]

Cool. Torrington roller clutch, CSK bearing, or something else? No one-way clutch I have ever seen for a 3/4" shaft can withstand bicycle crank torques, so I'm guessing you'll have to mount a sleeve to fit the bore of the clutch.

CSK bearing fitted with an steel adaptor sleeve that has a 48 spline bore and a square keyed OD.

 

[gwhy!1]

Oh hey, this acetone vapor polishing works great on the surface of the gears

After fudging with the parameters a bit, I got a 14 to 18t sleeve to print without any gap in the wall. This one will probably need a bit of filing to fit since I had to lie about the thickness of the print head a bit to get the g-code generator to squeeze plastic into the tight areas.

I'll send a few of these out to anyone that wants to try them. You will need the 95t or larger sprocket to use these sleeves, they are useless on their own since they change the gear reduction too much.

The offer is still there if you want to send me one I will cast a few in the resins that I have here and send them out to anyone who wants one to test for the cost of the postage.

 

[bee]

noticed a wobble on the crank gears today, did I bend my crank?
[youtube]X57ZEak1xFM[/youtube]

PM me your address and which gears you want, gwhy!

I'll send you some gears today LightingRods.

 

[skyungjae]

noticed a wobble on the crank gears today, did I bend my crank?
[youtube]X57ZEak1xFM[/youtube]

PM me your address and which gears you want, gwhy!

I'll send you some gears today LightingRods.

Completely normal. Even my Sick Bike Parts set up does that too. Just make sure it spins true under pedal and motor power.

 

[Samd]

I used to do 2 part resin jewellery. The process of silicon mould making was cool, but the stock was maybe a bit brittle. I am amazed at the toughness (in the true engineering sense) of ABS and to me it'd seem a better material on the surface of it. Trials should tell I guess!

If the print time is a concern, as much as they can take an hour to print each there's nothing to stop you queueing up a dozen on your printer and waking up to them the next morrning.

People seem to be trialing the PLA equivalent of lost wax sand casting elsewhere online, by sand casting the 3D print with a metal replacement/filler. Now that'd be strong...

 

[LightningRods]

I used to do 2 part resin jewellery. The process of silicon mould making was cool, but the stock was maybe a bit brittle. I am amazed at the toughness (in the true engineering sense) of ABS and to me it'd seem a better material on the surface of it. Trials should tell I guess!

If the print time is a concern, as much as they can take an hour to print each there's nothing to stop you queueing up a dozen on your printer and waking up to them the next morrning.

People seem to be trialing the PLA equivalent of lost wax sand casting elsewhere online, by sand casting the 3D print with a metal replacement/filler. Now that'd be strong...

Materials are interesting in their application. Many hard materials are also brittle (glass, high carbon steel). In the case of pulleys it's not so much impact resistance or flexibility that we want as it is resistance to de-lamination, shredding and erosion from wear. ABS might be a superior material to 2 part resin but if it has to be 3D printed or injection molded in an expensive mold every time it's not good for low volume, low cost production.

Sand casting is good for low cost, low volume production because it uses inexpensive green sand molds for every part cast. The problem with sand casting is that the sand imparts it's rough surface to the part. A sand finish is not good for a pulley surface.

Silicone molds are beautifully smooth but even high temperature silicones are limited to low melting temperature metals like pewter. Vulcanized rubber molds used in spin casting can handle temperatures up to metals like zinc. Ideal industrial metals like aluminum and iron have to be die cast (smooth finish and expensive for low production) or sand cast (rough finish and affordable for low production).

It seems to me that the best materials for good finish low volume pulleys are 2 part resins or zinc.

 

[Samd]

The word tough gets thrown about, but in a material science sense it has deliberate meaning. From wikipedia: "In materials science and metallurgy, toughness is the ability of a material to absorb energy and plastically deform without fracturing" and "Toughness requires a balance of strength and ductility."

It's the ability of ABS to take up energy that sees it used in so many car bumpers over the past decade or so.

I plugged my wattmeter into my 3D printer last week. To print the three GNG cogs for full-throttle cost me 12 cents in energy and about 19 cents in material. Ten cents each. Now if I was trying to make my wages it might be a different thing to amortise say fifty bucks an hour into three hours of production. But the beauty of it at home is I just set up three gears, head off to work or bed, and the machine does the rest. Ten cents a pop. I suppose I could amortise in the $700 cost of the machine over say a three year life, but I call it fully sunk as if I'd bought a playstation or something else of the ilk. I reckon 3D is cheap.

 

[maramusa]

I have been trawling this thread for hours deciding if its the right setup for me...finger on the buy button
GNG BB Drive 450W brushless belt motor, PING 15AH 45V battery pack
I wont be doing any trails, just urban commuting on my large hybrid 700c bike.
I live in a hilly area nothing too mad.
I like the idea of using the bike gears for efficiency, I want range and some hill climbing ...and power, 35-40km/hr would be nice,
but now have my doubts as per parajared post below

I wouldn't recommend this kit to anyone except those who want an off-road system. Hubs just plain can't handle the trails; 9c, magic pie, and e-bikekit.com gear hub all quickly reach the failure point for the kind of riding I use them for. GNG tackles the steepest, gnarliest, terrain with ease.

However hubs are great for going fast: 9c and magic pie both taken to 45mph, hubs are great for going up steep hills: 9c 2810 and magic pie both tackle urban trails and steep city streets like champs. (9c 2805 and e-bikekit.com fail!). They are easy to install, reliable, less complex, use less power for commuting, and require almost no maintenance. I think almost everyone should go with a hub.

Should I just focus on a good HS Crystalyte or MagicPie 3.
Please any opinions appreciated

 

[spinningmagnets]

A cell_man 10T geared hub will do about 25-MPH on 48V, and I'm told it has very good hill climbing. Same price, but no hassles.

http://em3ev.com/store/index.php?route=product/product&path=36&product_id=55

 

[speedmd]

I am waiting a few more weeks before I grind off the old pulley teeth on the motor shaft. I still have decided on a battery so I have some time. Interested in the pulley sprocket cap but have some reservations with the potential delam of the printed material while in use. Certain it will work, but not certain for how long and what type of failure mode it may have. Investment cast steel, Now that would be the ticket.

Are the 120 tooth pulleys still available. Last I checked, they were no longer available. 95 teeth is most likely still not enough reduction. It is looking like it will be a 3 chain bike.

Has anyone tried eccentric type bearing cups for the jack shaft bearings. Thinking I can turn some up, custom, and have potentially (if I am lucky with chain lengths) enough adjustment to handle both sides with one adjustment and possibly eliminate the need for any idlers with the stock mounting brackets. At least one can be eliminated for certain this way. Now, I have to think eccentrically for a few hours. Will post if I come up with anything useful.

 

[LightningRods]

This is an excerpt from the Net Gain Motors web site. Net Gain does EV conversions of all kinds.

"Our motors like to spin 2000-4000 RPMs. Running the motors at very low RPMs will generally draw
significant amperage and not allow the fan to cool the motor. Direct drive works well in racing
applications, however it is not the best choice for a daily street driven vehicle. The generally
accepted rule of thumb is this: Direct drive will require twice the motor and twice the controller of
vehicle with a transmission. This means you would have to use a WarP 9 coupled to a TransWarP
9 in an application where a single WarP 9 would normally suffice if a transmission was used.
Additionally, if a single Zilla 1K controller could have been used, you will need a Zilla 2K for a
direct drive application. Additionally, you must force cool air into direct drive motors if the normal
RPMs of the driven vehicle are below 2000 RPMs."

Hub motors are simple and can take a lot of voltage due to their copper mass. I don't see how they
can possibly be more efficient than a smaller motor with a transmission.

 

[justlooking]

Re: GNG CRANK ARM REPLACEMENT ,, THREADS KNACKARD

Postby justlooking » Wed Mar 06, 2013 7:33 am
i was looking on the GNG site and found this
http://www.gngebike.com/withcrankcrankarm.htm

it looks like the arms are made from steel ??
they fit the freewheel with the thread so might be a better option,

cost ?? USD $68.9 or $116.70 with airmail to the uk

i have also found a place in the uk that sells the alloy ones
i have found a place in the uk where postage is just over £6

they have the arms and crank freeweels

http://eclipsebikes.com/freewheel-chainwheels-c-32.html

 

[Chalo]

Hub motors are simple and can take a lot of voltage due to their copper mass. I don't see how they
can possibly be more efficient than a smaller motor with a transmission.

It's because you disregard the efficiency effects of a transmission, which in the real world eats up between 10% (well-maintained conventional bike gears operated at a couple hundred watts of mechanical power) and a majority of the motor's power.

The GNG drive starts as an enhancement to a conventional bicycle drivetrain that can be close to 90% efficient when the components are in good shape and everything is aligned, lubricated, and adjusted correctly. Then it imposes poor chainline, which increases friction and wear. Then it is run at higher than sustained human power levels, which also increases wear and promotes overuse of the least efficient and most wear-prone small sprockets. And it does this after first passing its motor power through two reduction stages, each with their own efficiency losses that get compounded at successive stages.

The only categorical efficiency benefit of a crank drive is in matching optimum motor speed to optimum pedal speed, so that by maintaining ratios that are effective for pedaling, the motor is consequently kept in its most efficient speed range. But unlike Stokemonkey, GNG's kit uses an overrunning chainring, so it doesn't provide that benefit. Folks who overvolt it so that the crank runs in the 100s of RPMs under power are clearly incapable of realizing this benefit.

The GNG kit may constitute efficient use of resources that are already present on the bicycle (the multi-speed transmission), but in terms of mechanical efficiency it's just poor; there is no getting around it. However efficient or inefficient a direct drive hub motor is, its power is not being shaved away by stacked ratio reductions (or in the GNG's case, two reductions followed by a step-up stage). The hub motor's power flows directly to the tire without incurring any loss along the way.

For someone who does not usually pedal, a crank drive doesn't make much sense at all. For someone who wants to use an amount of power and/or gear ratio well in excess of what bicycle components are designed for, crank drive is a plain bad idea. It seems like many of the participants in this discussion fall into both categories, and would be better served by some kind of hub motor, even if power efficiency is not important.

 

[LightningRods]

All transmissions have parasitic losses as do all reduction systems. And yet they proliferate everywhere including the most efficient hub motors which have planetary gear sets in them. The reason we see so many transmissions and reduction drives is that the benefits of the reduction, or the variable reduction in the case of a transmission, outweigh the parasitic losses.

Here is an article in electricbike.com comparing a 15 pound direct drive Crystalyte to an 8 pound BMC. http://www.electricbike.com/hubmotor/ At half the motor mass and at the same power consumption, "The BMC proved to be a better hill climber using the same power as the Crystalite when climbing hills without pedaling. This is because of the planetary gears." Additionally, the bike handled better due to the reduced weight.

Unless you ride your bike at only one speed, and never climb a hill too steep for the motor to stay at full rpm and high voltage/low amperage, it's better to have a transmission. It will be interesting to compare the mid mounted MACs and BMCs to hub mounted units once the mid drive kits are available.

 

[full-throttle]

It's because you disregard the efficiency effects of a transmission, which in the real world eats up between 10% (well-maintained conventional bike gears operated at a couple hundred watts of mechanical power) and a majority of the motor's power.

The GNG drive starts as an enhancement to a conventional bicycle drivetrain that can be close to 90% efficient when the components are in good shape and everything is aligned, lubricated, and adjusted correctly. Then it imposes poor chainline, which increases friction and wear.

Got any evidence to support your claim, Penny Farthing lover?

Single stage chain transmission is 99% eff at average human 200W input
Multispeed bicycle transmission is 97% eff
Planetary transmission with ball bearings on the planets is 95-97% eff
Look it up

 

[bee]

Any power loss in the gear reduction is made totally null and void by the ability to apply the power of the motor to 9 different gears and thus 9 different output loads. By using gearing to adjust the torque and speed required to pull our load, we can effectively overdrive our little motors to hit much higher RPMs in order to achieve a better top speed.

Yes, theoretically under ideal circumstances the non-geared hub motors can reach a higher top speed with the same power/weight ratio, but it will take the motor much longer to get to that speed if we can't use gears to reduce the load on the motor while it's spinning up to more efficient RPMs. I don't know anyone that only rides their bike under ideal circumstances though, there are plenty of hills and obstacles where I ride my bike that allow me to alternate gearing so that the motor is always maxing out RPMs and making peak torque, at any speed.

There's no arguing here, you're just going to have to try it yourself. I'm surprised this argument is even happening, but looking back to my 2-stroke tuning days I realize that it also took me a while to wrap my head around gearing efficiency too, and I setup a few transmissions that broke records on flat ground but couldn't climb some of the hills around here

Unfortunately on the cheaper end of two-wheeled sports we don't have the same access to testing equipment like a dyno which could end this argument with some hard scientific facts. Instead, we have a bunch of internet experts that have read TONS of blogs on physics and think that means they know how exactly how something will work in the real world.

I got the larger 3d printer working, here it is printing an 80 to 119t adapter:


Still needs more tuning of the retraction to clean up the little plastic boogers and I gotta sort out why the extruder jams after half an hour. Then I can start to tweak the print speeds to hopefully reduce the 7.5 hour print job so I can finish printing one of these.

 

[LightningRods]

Still needs more tuning of the retraction to clean up the little plastic boogers and I gotta sort out why the extruder jams after half an hour. Then I can start to tweak the print speeds to hopefully reduce the 7.5 hour print job so I can finish printing one of these.

Man! That's a family size pizza! Does it come as close to the front wheel as it appears in this photo???

 

[bee]

Here's a better angle



With how close the gears are to each other, I'm tempted by a herringbone gear setup

 

[gknowes]

this is my first post but i have been here along time trying to learn all that i can from you guys. i have a GNG 36v BB-drive. I too were missing hardware in my bottom sheets .BUT that is no reason for mike to pay with his own money trying to HELP US with these custom parts. so i started a BEER FUND for him of $ 10.00 to his paypal account please help me help him and THANK YOU all for the great work you guys do here. it has been a big help, gary

 

[KMB]

this is my first post but i have been here along time trying to learn all that i can from you guys. i have a GNG 36v BB-drive. I too were missing hardware in my bottom sheets .BUT that is no reason for mike to pay with his own money trying to HELP US with these custom parts. so i started a BEER FUND for him of $ 10.00 to his paypal account please help me help him and THANK YOU all for the great work you guys do here. it has been a big help, gary

+++1! The beer fund!

 

[Chalo]

The GNG drive starts as an enhancement to a conventional bicycle drivetrain that can be close to 90% efficient when the components are in good shape and everything is aligned, lubricated, and adjusted correctly. Then it imposes poor chainline, which increases friction and wear.

Got any evidence to support your claim, Penny Farthing lover?

Single stage chain transmission is 99% eff at average human 200W input

Nope.

Consult Wilson's Bicycling Science for well vetted and authoritative test data on chain drives. Even he admits that his cited reports' controlled test conditions (lower than your figures) may be more favorable than real world results.

Multispeed bicycle transmission is 97% eff

Nope. Not even close. Maybe on a test stand, with an unvarying load and brand new premium level components, in the best gear ratios. But that's not how it works in The Big Room.

Planetary transmission with ball bearings on the planets is 95-97% eff
Look it up

Nope, not in the real world. Besides, bicycle planetary transmissions other than the $2000 Rohloff hub use plain pinion bearings and have averaged efficiencies in the 80-something percent range-- plus the losses from the chain primaries, of course.

 

[Chalo]

I'm surprised this argument is even happening, but looking back to my 2-stroke tuning days I realize that it also took me a while to wrap my head around gearing efficiency too, and I setup a few transmissions that broke records on flat ground but couldn't climb some of the hills around here

Really?

Are you honestly likening the power curves of two-cycle stinkers to those of permanent magnet electric motors? Or are you just taking the piss?