3 speed hubs

Following the thread on geared wheels, I thought I’d toss in my $0.02. Standard
3 speed bicycle hubs used the same type of planetary gear system as a 3 speed
automatic transmission in an automobile. There are a set of planetary gears
which run around a central sun gear, and are held in place inside a race. The
different gear ratios are obtained by holding various parts of the system still
(no movement). This gives gear ratios from 1/3 to 1 (as I recall). The problem
is that these are all geared_down from the input, so you’d make your effective
wheel size smaller, not larger… There should be no problem with freewheeling,
as the gearset changes gear ratios through use of various clutches, which can be
applies while the gears are moving or while stationary. However, you might have
to gear up the input somehow, so that you could make use of the gears to give
you greater speed. In my mind, that would be the hard part.

All this information is what I remember from automobile transmissions, not
bicycle hubs, but to the best of my memory, they both work on the same
basic idea.

Keep us updated as to the progress… It should be an interesting challenge!

David S. Bowden dbowden@ford.com “I’m not a unicyclist in real life (yet), but I
play one on the internet”

RE: 3 speed hubs

From: SMTP%“schecter@TFS.COM” 31-AUG-1995 13:58:37.30 X-To: DBOWDEN
CC: Subj: Re: 3 speed hubs

Message-Id: <m0soDqo-0003wTC@TFS.COM> From: schecter@TFS.COM (Mark Schecter)
Subject: Re: 3 speed hubs To: dbowden@etcv01.eld.ford.com Date: Thu, 31 Aug 1995
10:56:29 -0700 (PDT) In-Reply-To: <422akb$mil@eccdb1.pms.ford.com> from
dbowden@etcv01.eld.ford.com” at Aug 30, 95 06:28:27 pm X-Mailer: ELM [version
2.4 PL23] Content-Type: text Content-Length: 3512

David,

To sum up my response, the automotive analogy is not an exact fit. I’ll be
specific below.

> Following the thread on geared wheels, I thought I’d toss in my $0.02.
> Standard 3 speed bicycle hubs used the same type of planetary gear system as a
> 3 speed automatic transmission in an automobile. There are a set of planetary
> gears which run around a central sun gear, and are held in place inside a
> race. The different gear ratios are obtained by holding various parts of the
> system still (no movement).

So far, so good.

> This gives gear ratios from 1/3 to 1 (as I recall). The problem is that these
> are all geared_down from the input, so you’d make your effective wheel size
> smaller, not larger…

This is where the systems are not alike. In the bicycle, the second of three
gears is direct drive, i.e. 1:1. In second gear, the input driver is locked to
the hub shell, which is locked to the output sprocket. First (low) gear is
approximately 1.2:1, that is, 20% reduction. Third (high) gear is
approximately .8:1, or 20% overdrive. (The first numeral is the input revs,
the second the output.)

> There should be no problem with freewheeling, as the gearset changes gear
> ratios through use of various clutches, which can be applied while the gears
> are moving or while stationary.

Sorry, but your theoretical “should” is not borne out in practice in the bicycle
hub. Thrust is applied to the hub by ratcheting pawls, which, when not engaged,
create the clicking you hear as you pedal, and the whirring you hear when you
stop pedaling. Gear changes are permitted by slop in the system, and can only
take place when torque is not being applied, a situation that is quite unlike
that inside an automatic transmission. The design of the hub and its ability to
revolve at rates that are slower than, the same as, and faster than the
mechanism inside it, i.e. to freewheel, are so intertwined that it would take
serious design changes and machine work to get around it, if indeed it’s even
possible. Both of these would cost substantial bucks and effort, and require
more knowledge, ability and bucks than I have. I don’t think it would be cost
effective.

> However, you might have to gear up the input somehow, so that you could make
> use of the gears to give you greater speed. In my mind, that would be the
> hard part.

See above. I think the speed of the input shaft in an auto trans is on the order
of 100 times the speed of the input shaft on a bicycle or uni.

> All this information is what I remember from automobile transmissions, not
> bicycle hubs, but to the best of my memory, they both work on the same
> basic idea.

Yes, the same basic idea, but only in the most general sense.

As has been said before, Sturmey-Archer created a non-freewheeling hub quite a
few years ago for its track racing bikes. A few of these may still exist
somewhere in the physical (as opposed to theoretical) world. Finding one would I
think be easier than designing and building one.

Also, as has been pointed out by a couple riders who have experience on geared
up unis and giraffes, (see recent posts by Ken Fuchs, Dennis Kathrens and John
Foss) the problems associated with gearing up may make the effort misguided. Not
to rain on anyone’s parade, but it has been done, and the results were less
than thrilling, fondest hopes notwithstanding.

Nevertheless, best of luck to anyone continuing this effort - often the reward
is in the pursuit.

-Mark Schecter

Re: 3 speed hubs

David,

To sum up my response, the automotive analogy is not an exact fit. I’ll be
specific below.

> Following the thread on geared wheels, I thought I’d toss in my $0.02.
> Standard 3 speed bicycle hubs used the same type of planetary gear system as a
> 3 speed automatic transmission in an automobile. There are a set of planetary
> gears which run around a central sun gear, and are held in place inside a
> race. The different gear ratios are obtained by holding various parts of the
> system still (no movement).

So far, so good.

> This gives gear ratios from 1/3 to 1 (as I recall). The problem is that these
> are all geared_down from the input, so you’d make your effective wheel size
> smaller, not larger…

This is where the systems are not alike. In the bicycle, the second of three
gears is direct drive, i.e. 1:1. In second gear, the input driver is locked to
the hub shell, which is locked to the output sprocket. First (low) gear is
approximately 1.2:1, that is, 20% reduction. Third (high) gear is
approximately .8:1, or 20% overdrive. (The first numeral is the input revs,
the second the output.)

> There should be no problem with freewheeling, as the gearset changes gear
> ratios through use of various clutches, which can be applied while the gears
> are moving or while stationary.

Sorry, but your theoretical “should” is not borne out in practice in the bicycle
hub. Thrust is applied to the hub by ratcheting pawls, which, when not engaged,
create the clicking you hear as you pedal, and the whirring you hear when you
stop pedaling. Gear changes are permitted by slop in the system, and can only
take place when torque is not being applied, a situation that is quite unlike
that inside an automatic transmission. The design of the hub and its ability to
revolve at rates that are slower than, the same as, and faster than the
mechanism inside it, i.e. to freewheel, are so intertwined that it would take
serious design changes and machine work to get around it, if indeed it’s even
possible. Both of these would cost substantial bucks and effort, and require
more knowledge, ability and bucks than I have. I don’t think it would be cost
effective.

> However, you might have to gear up the input somehow, so that you could make
> use of the gears to give you greater speed. In my mind, that would be the
> hard part.

See above. I think the speed of the input shaft in an auto trans is on the order
of 100 times the speed of the input shaft on a bicycle or uni.

> All this information is what I remember from automobile transmissions, not
> bicycle hubs, but to the best of my memory, they both work on the same
> basic idea.

Yes, the same basic idea, but only in the most general sense.

As has been said before, Sturmey-Archer created a non-freewheeling hub quite a
few years ago for its track racing bikes. A few of these may still exist
somewhere in the physical (as opposed to theoretical) world. Finding one would I
think be easier than designing and building one.

Also, as has been pointed out by a couple riders who have experience on geared
up unis and giraffes, (see recent posts by Ken Fuchs, Dennis Kathrens and John
Foss) the problems associated with gearing up may make the effort misguided. Not
to rain on anyone’s parade, but it has been done, and the results were less
than thrilling, fondest hopes notwithstanding.

Nevertheless, best of luck to anyone continuing this effort - often the reward
is in the pursuit.

-Mark Schecter

---------------------------------------------------------------*
| | |
| Mark Schecter | “Please identify yourself.” |
| schecter@tfs.com | (Self checks pocket mirror) |
| | “Yes, that’s me.” |
*------------------------- -------------------------------------