Epicyclic hub gear

Unicycles and Equipment
1

Hello all-

I am developing an epicyclic hub gear for unicycles. This is a topic that has been beaten into the ground in the past along with wheel size for an ideal big wheel so grit your teeth (good engineering pun, that) and I will explain my motivation, describe the plan and schedule, and ask for input. Keep in mind that a geared unicycle can only compete with a high end, pneumatic tired big wheel in the catagories of speed and weight. A big wheel has an esthetic appeal unattainable by a “standard” sized unicycle (I don’t know about you but that is why I bought a Coker) as well as enhanced ride comfort due to the wheel diameter and associated angular momentum. There is also the belief that big wheels are more easily controlled than geared unicycles, but that may be a problem due to backlash in geared systems more than anything else.

Several threads have already dealt with geared unicycles. It seems that geared giraffes, chain driven standard unicycles, modified Sturmey-Archer (S-A) hubs, and external gearing have been tried or seen by others. I don’t think anyone has designed and built an epicyclic hub gear strictly for unicycles. This would necessarily differ from the S-A hub in that the crank axle, or applied drive shaft, must be the inner most rotating shaft. In the S-A, the drive is applied to a chain sprocket which is inside the fork and is the middle rotating shaft. The difficulties encountered with modified S-A hubs, as I understand it, are hokiness, backlash, and lack of strength and reliability. After all, they were designed to be freewheeling (infinite backlash) and chain driven.

Chain driven attempts all have to deal with the problem of chain stretch and slackness and its associated backlash. Uniformly driving both sides of the hub seems to be a problem also. These are difficulties that will always be encountered in giraffe designs as well as chain driven geared unicycles.

Externally geared unicycles all seem to require split crank axles. This has to be a serious mechanical weakness and I would think that the fork is the part that has to be reinforced to accommodate the design.

A properly designed epicyclic hub gear for a unicycle should have a continuous (unsplit) crank axle 17mm or so in diameter. It should have very small backlash. The gear ratio limits for simple sun/planet/ring systems vary from 1 to 2 with realistic (attainable) ratios from 1.2 to about 1.8 or so. Compound gear systems can produce much higher ratios with a cost of increased weight, volume, and complexity. It should use off-the-shelf, standard gears, retainers, and bearings.

My prototype design will have (or has) a simple gear ratio of either 1.5:1 or 1.625:1, an aluminum hub with a flange diameter of 3.75" (big), a continuous 17mm crank axle, standard steel gears, standard, sealed ball bearings, and is servicable by removable caps on each end of the hub. The backlash should be 4 mrad or 0.23 degrees which is about 0.024" or 0.6mm measured at the pedal on a 6" crank.

The ratio of 1.5:1 is highly desirable because it makes a 24" wheel into a 36" wheel equivalent. There are many people who have experience riding on a 36" wheel with 6" cranks and could therefore make an accurate comparison. I live in Seattle and my goal is to build at least one prototype to have available for people to try at the 2002 NUC and UNICON 11 in North Bend, WA next year. Hopefully, lots of people who ride Cokers will show up and test it so I can determine the control weaknesses. Hopefully there will be riders there strong enough to break it so I can determine the mechanical weaknesses too.

A gear ratio of 1.625:1 will make a 700c wheel (which is a standard mountain bike size about 27" in diameter) into a 44" wheel equivalent behemoth. There are lots of different 700c wheels and tires. This design can probably be easily altered to make a hub that can be shifted to direct drive (1:1) so, if it is rugged enough, it could be shifted to high gear to cruise to an offroad trail and 1:1 to go trail riding.

Any feedback would be appreciated. If anyone has (supportable) arguments against an initial gear ratio of 1.5:1, let me know. If anyone wants to donate time or parts to this project (lace a wheel, provide a fork, provide a variety of seatpost lengths with seats, etc.), please let me know. I have a rough AutoCAD drawing of the stackup of my third design pass. I will probably make two more drafts before I start machining.

Thanks.

2

Hello all I’ve just signed on and saw this. I built a unicycle with a
modified pedal-back brake 3 speed hub - that worked pretty well but as you
say not reliable - it broke soon after i learnt to ride it - breaks are
tricky things on a one wheeled bike!

3

> Chain driven attempts all have to deal with the problem of chain stretch
> and slackness and its associated backlash. Uniformly driving both sides
> of the hub seems to be a problem also. These are difficulties that will
> always be encountered in giraffe designs as well as chain driven geared
> unicycles.
>
> Externally geared unicycles all seem to require split crank axles. This
> has to be a serious mechanical weakness and I would think that the fork
> is the part that has to be reinforced to accommodate the design.

I’ve actually got an idea about this, which is basically to use an axle
which is like a unicycle axle on the outside where the cranks fit on and
like a bike axle in the middle bit. The hub could be on bearings in a
similar way to a back hub on a bike. A big cog is fitted to the cranks, on
a bike like spider. This cog drives a smaller cog attached to one side of
the fork, getting the gearing up required. The fork cog then drives a cog
attached to a standard bicycle rear hub.

This has one very very major advantage, which is that it would allow a
standard bicycle rear hub and cranks to be used. This would all work
because bike hubs nowadays tend to have separate freewheel cassettes, so
the actual hub doesn’t have any freewheel mechanism. What would be
needed as far as componenents are concerned would be a new axle, a cog
to fit on a normal bike spider, two cogs to fit on the fork and a cog
with a hole in that fitted onto a rear hub instead of a cassette. It
might even be possible to persuade someone to make splined axles other
than the profile splines for this system, allowing a greater variety of
splined cranks to be used, although the spline patterns are patented so
you can’t copy them yourself.

The major disadvantage of this system would be the extra external
complexity, requiring three extra bearings over a normal unicycle and also
an extra bearing holder on the fork on one side of the wheel. It would
also require a bit more fork strength, but not too much as the weight
bearing axle would still be un-broken. There would be some slackness, but
nothing like a chain driven unicycle, it would probably be fine. It might
lose some strength as I’m not sure if a bike rear axle is narrower than a
unicycle axle.

It would also be not too hard to make this have two gearings by shifting
the smaller cog and having two cogs on the hub and on the inside small
cog. All in all this would be wicked, a two geared unicycle with access to
standard bicycle parts. Also, as the only engineering required by a bike
person to support unicycles using this system would be the axle, we might
be able to persuade a few more bike companies to support it.

> A properly designed epicyclic hub gear for a unicycle should have a
> continuous (unsplit) crank axle 17mm or so in diameter. It should have
> very small backlash. The gear ratio limits for simple sun/planet/ring
> systems vary from 1 to 2 with realistic (attainable) ratios from 1.2 to
> about 1.8 or so. Compound gear systems can produce much higher ratios
> with a cost of increased weight, volume, and complexity. It should use
> off-the-shelf, standard gears, retainers, and bearings.
>
> My prototype design will have (or has) a simple gear ratio of either
> 1.5:1 or 1.625:1, an aluminum hub with a flange diameter of 3.75" (big),
> a continuous 17mm crank axle, standard steel gears, standard, sealed
> ball bearings, and is servicable by removable caps on each end of
> the hub. The backlash should be 4 mrad or 0.23 degrees which is
> about 0.024" or 0.6mm measured at the pedal on a 6" crank.
>
> The ratio of 1.5:1 is highly desirable because it makes a 24" wheel into
> a 36" wheel equivalent. There are many people who have experience riding
> on a 36" wheel with 6" cranks and could therefore make an accurate
> comparison. I live in Seattle and my goal is to build at least one
> prototype to have available for people to try at the 2002 NUC and UNICON
> 11 in North Bend, WA next year. Hopefully, lots of people who ride
> Cokers will show up and test it so I can determine the control
> weaknesses. Hopefully there will be riders there strong enough to break
> it so I can determine the mechanical weaknesses too.

Aha, this is what I was trying to describe in an earlier message, except I
couldn’t quite work out how to make it all fit together. I could work out
how to gear down, but not how to gear up (I’m not an engineer at all). I
also didn’t manage to find anyone capable of building a custom cycle hub.
I’m glad someone’s turned up who knows enough about the engineering to
actually try to design and make an epicyclic hub gear, I could work out it
was possible, but not work out how it could be done.

> A gear ratio of 1.625:1 will make a 700c wheel (which is a standard
> mountain bike size about 27" in diameter) into a 44" wheel equivalent
> behemoth. There are lots of different 700c wheels and tires. This design
> can probably be easily altered to make a hub that can be shifted to
> direct drive (1:1) so, if it is rugged enough, it could be shifted to
> high gear to cruise to an offroad trail and 1:1 to go trail riding.

Now that would be wicked, kind of like a singlespeed bike with a
speed-drive (two gear bottom bracket). This would be the perfect unicycle
for doing mountain bike XC type competitions.

> Any feedback would be appreciated. If anyone has (supportable) arguments
> against an initial gear ratio of 1.5:1, let me know. If anyone wants to
> donate time or parts to this project (lace a wheel, provide a fork,
> provide a variety of seatpost lengths with seats, etc.), please let me
> know. I have a rough AutoCAD drawing of the stackup of my third design
> pass. I will probably make two more drafts before I start machining.

Any chance of a gif / pdf or something of the designs?

Joe

4

What are the differences/pros/cons of using a planetary geared
bottombracket system as described at:
http://www.schlumpf.ch/speed/speeddrivee.html ? I was given to understand
that it didn’t have the same magnitude of backlash problems as S/A-based
systems. Good or bad info? Jim

5

> What are the differences/pros/cons of using a planetary geared
bottombracket
> system as described at: http://www.schlumpf.ch/speed/speeddrivee.html ?
> I was given to understand that it didn’t have the same magnitude of
backlash
> problems as S/A-based systems. Good or bad info? Jim

The speeddrives say not to use them on fixed wheels, they can’t take back
pressure, so you can’t just use them out of the box on a giraffe. Also I
think the main problem with a bottom-bracket system is that the bottom
bracket has to be above the wheel, putting you at least 20inches off the
ground and that you have a chain involved, hence all the problems of
slack chains you get on a normal giraffe. A hub-gear or some other kind
of gear that can work without a chain should remove a lot of the slack
problems and hopefully make something requiring a lot less maintenance
than a giraffe.

Joe

6

Joe-

As I said, I am on my 3rd draft so my drawing has no cross hatching, dimensioning, section views, or part details yet. All I have is the stackup. If you have access to AutoCAD release 12 or later, you can see the entire .dwg file and peel away the layers (one for each part type). I can export *.dxf, *.iges, *wmf, and *.eps files (drawing exchange file, international graphics exchange file, windows metafile, and encapsulated PostScript file, respectively). Someone around my lab will know how to produce *.pdf files. Let me know what you can use.

15:25 P.D.T.: O.K., I have a sort-of-all-right pdf file if you’re interested.

7

harper wrote:
>
> Hello all-
>
> I am developing an epicyclic hub gear for unicycles.

<big snip>

What the heck is an epileptic hub gear anyway!? :wink:

Has anyone ever thought of using an enclosed hub with electronics and
electric servos etc. to change gears in the hub all controled remotely
with radio such as ‘Bluetooth’?

Christopher Grove

–
“I never MET a man I didn’t like.” -Will Rogers

Read my short story at: http://www.rosedog.com/manuscript.asp?m=9389256&m-
anuscript=6862016&t=The+Determination+of+Jeffery&fn=Christopher&ln=Grove&-
srdt=9%2F27%2F01+5%3A32%3A05+AM&fs=2&ff=Arial%2C+Helvetica%2C+Verdana&ps=-
4000&s=735143750&folder=Search+Results

8

Yes; they call IT Ginger- an evil device that Kamen is working on to make unicyclists obsolete.

Christopher

9

Send me the drawing and I will do the exchange/upload it for people to
look at. Keep it pure and send it as ACAD V12 I can deal with that.

Roger

“harper” <forum.member@unicyclist.com> wrote in message
news:9r4m27$37p$1@laurel.tc.umn.edu…
> Joe-
>
> As I said, I am on my 3rd draft so my drawing has no cross hatching,
> dimensioning, section views, or part details yet. All I have is the
> stackup. If you have access to AutoCAD release 12 or later, you can see
> the entire .dwg file and peel away the layers (one for each part type).
> I can export *.dxf, *.iges, *wmf, and *.eps files (drawing exchange
> file, international graphics exchange file, windows metafile, and
> encapsulated PostScript file, respectively). Someone around my lab will
> know how to produce *.pdf files. Let me know what you can use.
>
>
>
>
> –
> harper Posted via the Unicyclist Community -
> http://unicyclist.com/forums

10

On Tue, 23 Oct 2001 21:05:11 +0000 (UTC), harper
<forum.member@unicyclist.com> wrote:

>As I said, I am on my 3rd draft so my drawing has no cross hatching,
>dimensioning, section views, or part details yet. All I have is the
>stackup. If you have access to AutoCAD release 12 or later, you can see
>the entire .dwg file and peel away the layers (one for each part type).

Hi!

As I am also thinking about building a geared up hub, I would like to get
the .dwg file. Wouldn´t it be great to have a few geared up´s at NUC???

My e-mail is einrad@epost.de

CU at NUC!

Frank

11

The thought of wireless control has occurred to me, but that’s all. I am
not sure that Harper’s epicyclic hub could switch gears on the fly, which
IMHO would be the juice of it. But maybe it could??

Klaas Bil

On Wed, 24 Oct 2001 15:00:48 GMT, Christopher Grove
<c_r_grove@yahoo.com> wrote:

>harper wrote:
>>
>> Hello all-
>>
>> I am developing an epicyclic hub gear for unicycles.
>
><big snip>
>
>What the heck is an epileptic hub gear anyway!?
>
>Has anyone ever thought of using an enclosed hub with electronics and
>electric servos etc. to change gears in the hub all controled remotely
>with radio such as ‘Bluetooth’?
>
>Christopher Grove
>
>–
>“I never MET a man I didn’t like.” -Will Rogers
>
>Read my short story at: http://www.rosedog.com/manuscript.asp?m=9389256&-
>manuscript=6862016&t=The+Determination+of+Jeffery&fn=Christopher&ln=Grov-
>e&srdt=9%2F27%2F01+5%3A32%3A05+AM&fs=2&ff=Arial%2C+Helvetica%2C+Verdana&-
>ps=4000&s=735143750&folder=Search+Results

–
“To trigger/fool/saturate/overload Echelon, the following has been picked
automagically from a database:” “Mossad, EP-3 Aries, PGP 7.0”

12

The prototype will be a fixed gear unit. If I ever design a two speed unit I am fairly sure that it will not be shift on the fly. I do try to incorporate exploding bolts in all of my designs, though, because they are so cool. That would ensure that the gear ratio changes rapidly on the fly.

I have sent the .dwg files to Roger Davies and to Frank as requested. Hope they are received. Let me know if they are not.

13

Roger Davies has kindly posted the drawing on his web site at
http://homepage.ntlworld.com/rdavies21/greg/epicyc03.jpg

Frank has a similar design that appears to me to be much more rugged. Hopefully his drawing will be posted somewhere also.