search  current discussion  categories  wheels - misc 

building a cone-drive wheel

updated sat 21 sep 02

 

vince pitelka on mon 16 sep 02


David -
I built a Shimpo style wheel of the RK-1 type about 25 years ago, and it
still works perfectly. I used the sliding motor mount, using a
chrome-plated machine shaft and bronze bushings for the sliding surfaces. I
equipped them with zerk (grease) fittings, and I grease it once every few
years. I purchased the drive cone and rubber ring for an RK-1 from Shimpo,
and built the rest. I've replaced the rubber drive ring once in 25 years.
I would not recommend you use any other kind of mechanical movement, because
this is the only one that works really well. You need to have extremely
positive, easily adjustable contact pressure between the drive cone and the
rubber ring, and this system allows that. The cone has to rise up against
the drive ring just after the tip of the cone, and then the pressure has to
remain constant from one end of the throw to the other. I cannot envision
any way you could accomplish that with a "record player tone arm" type of
mechanism. It would have to be extremely precise, and seems a far more
difficult way to accomplish this than the sliding motor mount.

On my wheel, the weight of the motor is carried by the shaft and sliding
bearings, and the pressure is controlled by a sealed bearing running against
a steel rail. The rail is adjustable in order to control the pressure of
the cone against the drive ring. At one end there is a dip in the rail -
when the motor gets to that end, which is the small end of the cone, the
bearing drops into the dip, relieving the pressure against the cone, and the
wheel stops. When you step on the pedal, the linkage pulls the motor along
the shaft, and the bearing climbs out of the dip, raising the cone against
the drive ring.

The RK-8 is a completely different design, with the drive ring mounted on a
large wheel around the inside diameter of the housing. I am not familiar
with the Brent design. On the RK-1 the drive ring is only about 5" in
diameter. It is mounted on a jackshaft, and there is a belt drive from
there to a large pulley on the wheelhead shaft. I just used a stout
two-groove cast iron belt pulley on bottom of the jackshaft, and machined
the bottom surface of the pulley to accept the drive ring, with a steel disk
to hold the ring in place, and used the upper belt groove for the drive
belt.

Saying that the sliding system is prone to jamming if not built precise is
absurd. Don't even bother attempting this project unless you are going to
build it precise. Do it right. It is not at all difficult to build the
sliding system if you just plan it well and use high-quality materials.

But first and foremost, unless you are very experienced in welding and
fabricating and don't mind spending a LOT of your time working on this, it
will be MUCH cheaper to find a used RK-1. I was working as a journeyman
mechanic and welder when I built mine. It was a big job. I built it much
larger than an RK-1, because I am 6'5" tall. If you want one the same size
as a Shimpo RK-1, you are going to have a rough time fitting everything in
that little case. Shimpo does a superb job of engineering and construction.
Remember that Shimpo is a huge corporation specializing in power
transmission devices.

When I built my wheel, the Robert Brent Company was still in Healdsburg, CA,
and they sold wheelhead seconds for ten bucks - that's the wheelhead, the
shaft, and the bearing/flange assembly. I bought two 14" wheelhead
seconds. I picked out the best one, straightened out the slightly-bent
wheelhead, and used it for its original purpose. I cut the wheelhead off
the second one and used the rest of it for the jackshaft mentioned above -
you need both. You might check with Brent-Amaco to see if they still sell
wheelhead seconds. One of the advantages is that you can use a Brent
two-piece splashpan, which is about the best design out there.

The Brent wheelhead uses the standard bat pins on 10" centers. As far as I
know, the only way to buy a Brent wheelhead is to get it with the shaft and
bearing/flange assembly, and as far as I am concerned, that is the ONLY way
to go. I can't imagine it would be worth your while to try to adapt or
machine all this, when the complete 14" wheelhead assembly only costs you
$150.

But perhaps the important lesson here is that in the end you are going to
invest as much money (especially if you value your time at all!) as if you
searched out and bought a used RK-1. In order to take this project on, you
have to love the process and the mechanisms, and you have to be willing to
take the time to build a precision device. Otherwise it isn't worth it.
Good luck -
- Vince

Vince Pitelka
Appalachian Center for Crafts
Tennessee Technological University
1560 Craft Center Drive, Smithville TN 37166
Home - vpitelka@dtccom.net
615/597-5376
Work - wpitelka@tntech.edu
615/597-6801 ext. 111, fax 615/597-6803
http://www.craftcenter.tntech.edu/

Steve Mills on wed 18 sep 02


A very far cry from the ex-automotive Differential wheels many of us
made more than a few years ago! a lovely strong, smooth, simple
mechanism, but rather bulky unless you used a Diff. from a car with
independent rear suspension :-)

Steve
Getting nostalgic in
Bath
UK


In message , vince pitelka writes
>David -
>I built a Shimpo style wheel of the RK-1 type about 25 years ago, and it
>still works perfectly. I used the sliding motor mount, using a
>chrome-plated machine shaft and bronze bushings for the sliding surfaces.=

--
Steve Mills
Bath
UK

David Malicky on thu 19 sep 02


vince,

thanks for your detailed response. your post brings up the main issues i'm
struggling with on this design: how do i maximize the benefit/cost ratio of the
project. seems like there are two known solutions with good b/c ratios:
1) the low-tech "Build A $75 Electric Wheel" by Jolyon Hofsted.
2) the high-tech RK-1 copy, done "downtown," like you did.

my hope was to find a medium-tech solution using a ring-cone design, i.e. a
wood frame w/ RK-1 type parts. But from your post and what i've read in the
archives, ring-cone requires precision that wood cannot deliver, especially
considering moisture-related swelling of wood. i can weld (build 2 racecars,
actually) but don't have the equipment currently and don't really want to
invest the time for an all-metal design. as you noted, it is a lot of work.

aside #1: i agree the motor should run on a linear shaft like shimpo did; any
misalignment of the motor would produce scrub between the roller and the
rubber, and in the top-view that scrub makes for a thrust vector directed right
along the motor-translation-axis ---> did anyone say speed it up!? this also
nixes a wood frame due to lack of precision and swelling.

aside #2: your point on it being a big investment of time and money
(approaching a used RK-1) is also well taken. yes the money adds up quickly.
i'm an engineer and actually like designing and making these kind of projects
about as much as pottery, so i'd like to make it. but i don't want it to
consume my weekends for months on end. if i can find a medium-tech-effort
design i'd do it. if not, a used rk-1 sounds right -- where can one find
those? what do they usually go for?

so why not the hofsted wheel? i just want more function. it's not really a
speed-controlled wheel, but a torque-controlled wheel with some momentum.
i.e., applying different throwing torques will slow the wheel down
considerably. i don't mean to criticize the design overall, because for a
quick and cheap solution it is really beautiful.

some potential medium-tech approaches (brainstorming):
- dc motor. high $, though. plus i have never cared for the speed control of
the pedals - there's something about the shimpo 'feel' that i just like better.
personally, from a "fun project" standpoint it's pointless, since there's no
creativity in it. i'd rather buy a used brent.
- the rk-8 design eliminates the jack-shaft, saving considerable $ and time.
metal frame still required, but perhaps a metal subframe just for the mechanism
(e.g., rivited aluminum square tubing?) and wood for the box.
- an rk-8 design w/ wood frame, but control the cone/motor such that the
resulting scrub/thrust doesn't change speeds. this is similar to shimpo's
table top wheel (?), where instead of a foot pedal there is a hand-screw-
actuated mechanism. the screw wouldn't allow back-driving, so the thrust would
be absorbed instead of causing the cone to climb on the ring. not sure how
well this theory would work in practice, though.
- 4 speed design: abandon ring-cone and have 2 sets of 4-step-pulleys (one
reversed with respect to the other) with 4 belts loosely wound around each.
then 1 of 4 idlers could be selectively engaged (even by a foot pedal) to
tension the desired belt, engaging that set of pulley ratios, giving a choice
of, e.g., 50, 100, 150, 200 rpm. cludgy but it would probably work and not
require much precision.
- any others?

thanks also for the tips on brent wheelheads - i'll check that out.

thanks again,
david





On 16 Sep 2002 at 22:01, vince pitelka wrote:

> David -
> I built a Shimpo style wheel of the RK-1 type about 25 years ago, and it
> still works perfectly. I used the sliding motor mount, using a
> chrome-plated machine shaft and bronze bushings for the sliding surfaces. I
> equipped them with zerk (grease) fittings, and I grease it once every few years.
> I purchased the drive cone and rubber ring for an RK-1 from Shimpo, and built
> the rest. I've replaced the rubber drive ring once in 25 years. I would not
> recommend you use any other kind of mechanical movement, because this is the
> only one that works really well. You need to have extremely positive, easily
> adjustable contact pressure between the drive cone and the rubber ring, and this
> system allows that. The cone has to rise up against the drive ring just after
> the tip of the cone, and then the pressure has to remain constant from one end
> of the throw to the other. I cannot envision any way you could accomplish that
> with a "record player tone arm" type of mechanism. It would have to be
> extremely precise, and seems a far more difficult way to accomplish this than
> the sliding motor mount.
>
> On my wheel, the weight of the motor is carried by the shaft and sliding
> bearings, and the pressure is controlled by a sealed bearing running against a
> steel rail. The rail is adjustable in order to control the pressure of the cone
> against the drive ring. At one end there is a dip in the rail - when the motor
> gets to that end, which is the small end of the cone, the bearing drops into the
> dip, relieving the pressure against the cone, and the wheel stops. When you
> step on the pedal, the linkage pulls the motor along the shaft, and the bearing
> climbs out of the dip, raising the cone against the drive ring.
>
> The RK-8 is a completely different design, with the drive ring mounted on a
> large wheel around the inside diameter of the housing. I am not familiar with
> the Brent design. On the RK-1 the drive ring is only about 5" in diameter. It
> is mounted on a jackshaft, and there is a belt drive from there to a large
> pulley on the wheelhead shaft. I just used a stout two-groove cast iron belt
> pulley on bottom of the jackshaft, and machined the bottom surface of the pulley
> to accept the drive ring, with a steel disk to hold the ring in place, and used
> the upper belt groove for the drive belt.
>
> Saying that the sliding system is prone to jamming if not built precise is
> absurd. Don't even bother attempting this project unless you are going to
> build it precise. Do it right. It is not at all difficult to build the
> sliding system if you just plan it well and use high-quality materials.
>
> But first and foremost, unless you are very experienced in welding and
> fabricating and don't mind spending a LOT of your time working on this, it
> will be MUCH cheaper to find a used RK-1. I was working as a journeyman
> mechanic and welder when I built mine. It was a big job. I built it much
> larger than an RK-1, because I am 6'5" tall. If you want one the same size as a
> Shimpo RK-1, you are going to have a rough time fitting everything in that
> little case. Shimpo does a superb job of engineering and construction. Remember
> that Shimpo is a huge corporation specializing in power transmission devices.
>
> When I built my wheel, the Robert Brent Company was still in Healdsburg, CA, and
> they sold wheelhead seconds for ten bucks - that's the wheelhead, the shaft, and
> the bearing/flange assembly. I bought two 14" wheelhead seconds. I picked out
> the best one, straightened out the slightly-bent wheelhead, and used it for its
> original purpose. I cut the wheelhead off the second one and used the rest of
> it for the jackshaft mentioned above - you need both. You might check with
> Brent-Amaco to see if they still sell wheelhead seconds. One of the advantages
> is that you can use a Brent two-piece splashpan, which is about the best design
> out there.
>
> The Brent wheelhead uses the standard bat pins on 10" centers. As far as I
> know, the only way to buy a Brent wheelhead is to get it with the shaft and
> bearing/flange assembly, and as far as I am concerned, that is the ONLY way to
> go. I can't imagine it would be worth your while to try to adapt or machine all
> this, when the complete 14" wheelhead assembly only costs you $150.
>
> But perhaps the important lesson here is that in the end you are going to
> invest as much money (especially if you value your time at all!) as if you
> searched out and bought a used RK-1. In order to take this project on, you have
> to love the process and the mechanisms, and you have to be willing to take the
> time to build a precision device. Otherwise it isn't worth it. Good luck - -
> Vince
>
> Vince Pitelka
> Appalachian Center for Crafts
> Tennessee Technological University
> 1560 Craft Center Drive, Smithville TN 37166
> Home - vpitelka@dtccom.net
> 615/597-5376
> Work - wpitelka@tntech.edu
> 615/597-6801 ext. 111, fax 615/597-6803
> http://www.craftcenter.tntech.edu/
>
>