friction shifters and internally-geared hubs?



S

Slug

Guest
Is there any really important reason why one shouldn't use a friction shifter with an internally-
geared hub? I've been thinking about building a 7- or 8-speed coaster-brake nexus onto a 700c rim
and throwing it onto a road frame with mustache bars. In the image in my head, it just looks best
with a barcon.

Apparently according to Sheldon Brown, it's possible but not a very good idea. I'm just wondering
the details of why it's not a good idea. I kinda like the occasional (well, okay.. frequent) bad
idea if it's not going to seriously injure me or drastically cut short the working life of pricey
equipment.

Any elaborators?

Thanks,
Y.
 
On Thu, 04 Mar 2004 22:56:33 -0600, slug
<[email protected]> may have said:

>Is there any really important reason why one shouldn't use a friction shifter with an internally-
>geared hub? I've been thinking about building a 7- or 8-speed coaster-brake nexus onto a 700c rim
>and throwing it onto a road frame with mustache bars. In the image in my head, it just looks best
>with a barcon.
>
>Apparently according to Sheldon Brown, it's possible but not a very good idea. I'm just wondering
>the details of why it's not a good idea. I kinda like the occasional (well, okay.. frequent) bad
>idea if it's not going to seriously injure me or drastically cut short the working life of pricey
>equipment.
>
>Any elaborators?

The shifter's accuracy is essential. You're not dealing with a mechanism that is somewhat self-
centering like a der in this case; the cable's motion is what actually moves the gearsets into and
out of engagement, and if the correct throw isn't achieved, you'll have gears that are only partly
engaged. This is a Bad Thing. In the case of the old 3-speed SA hubs, there was even a spot where
you'd get no gear at all.

No, the correct shifter isn't the most cosmetically esthetic device for a road bike. Use it anyway.

--
My email address is antispammed; pull WEEDS if replying via e-mail.
Typoes are not a bug, they're a feature.
Words processed in a facility that contains nuts.
 
On 3/4/04 11:38 PM, in article [email protected],
"Werehatrack" <[email protected]> wrote:

> On Thu, 04 Mar 2004 22:56:33 -0600, slug <[email protected]> may have said:
>
>> Is there any really important reason why one shouldn't use a friction shifter with an internally-
>> geared hub? I've been thinking about building a 7- or 8-speed coaster-brake nexus onto a 700c rim
>> and throwing it onto a road frame with mustache bars. In the image in my head, it just looks best
>> with a barcon.
>>
>> Apparently according to Sheldon Brown, it's possible but not a very good idea. I'm just wondering
>> the details of why it's not a good idea. I kinda like the occasional (well, okay.. frequent) bad
>> idea if it's not going to seriously injure me or drastically cut short the working life of pricey
>> equipment.
>>
>> Any elaborators?
>
> The shifter's accuracy is essential. You're not dealing with a mechanism that is somewhat self-
> centering like a der in this case; the cable's motion is what actually moves the gearsets into and
> out of engagement, and if the correct throw isn't achieved, you'll have gears that are only partly
> engaged. This is a Bad Thing. In the case of the old 3-speed SA hubs, there was even a spot where
> you'd get no gear at all.
>
> No, the correct shifter isn't the most cosmetically esthetic device for a road bike. Use
> it anyway.

Yeah, sounds like it falls into the drastically-cuts-short-the-working-life category. I was worried
that'd be the case.

I guess I'd just need to find an indexed 7- or 8- speed barcon and figure out if there's any way to
file the detents so that the cable travel exactly matches that in the hub's "real" shifter. But I
don't know if I have faith enough in my hand-eye coordination for that feat.

Thanks for the advice..
y.
 
slug wrote:

> Is there any really important reason why one shouldn't use a friction shifter with an internally-
> geared hub? I've been thinking about building a 7- or 8-speed coaster-brake nexus onto a 700c rim
> and throwing it onto a road frame with mustache bars. In the image in my head, it just looks best
> with a barcon.
>
> Apparently according to Sheldon Brown, it's possible but not a very good idea. I'm just wondering
> the details of why it's not a good idea. I kinda like the occasional (well, okay.. frequent) bad
> idea if it's not going to seriously injure me or drastically cut short the working life of pricey
> equipment.

Sheldon's right. On a good day you might get the shifter close enough to ride but standing through a
yellow light wouldn't be prudent. If you ever saw an internal hub apart you wouldn't be so enamored
of this. Mating edges of drive components are small and precise. Loading drive forces across four
small contacts 1/2mm wide versus 3mm wide can result in chipped edges besides just simple
disengagement. Perhaps if there were no sideplay in mating parts you could manage but in the real
world, pieces with high loads on tiny edges fail.

--
Andrew Muzi www.yellowjersey.org Open every day since 1 April, 1971
 
slug <[email protected]> wrote in message news:<BC6D897A.31A8%[email protected]>...
>
> I guess I'd just need to find an indexed 7- or 8- speed barcon and figure out if there's any way
> to file the detents so that the cable travel exactly matches that in the hub's "real" shifter. But
> I don't know if I have faith enough in my hand-eye coordination for that feat.
>

File the detents?

Wouldn't a bellcrank be better?
 
Probably. What's a bellcrank?

On 3/5/04 8:59 AM, "Brian Huntley" <[email protected]> wrote:

> File the detents?
>
> Wouldn't a bellcrank be better?
 
slug <[email protected]> wrote in message news:<BC6DFA0F.32DA%[email protected]>...
> Probably. What's a bellcrank?
>
It's a lever, used to change a cable or rod movement in direction or amount or both. In this
case, I'd imagine using a short piece of stainless steel with a central pivot point somewhere
convient (under the seat?) You'd attach the shifter cable to a calibrated spot on one side of the
lever, and attach a second cable from another calibrated spot on the other side of the pivot and
down to the hub.

Fixed font drawing:

||-------------- cable to shifter
||
____||
|____*| * = pivot
|
|

cable to hub

The ratio of the distances from the pivot to the attachements would be carefully chose to give the
correct pull. For example, an 8 speed Shimano shifter pulls 3.07 mm per click. If you needed 4 mm,
you would set the distances from the cables to the pivots to something like 30mm and 39mm. The trick
is, the level arms have to be long enough to accomodate the full range of movement, and the pivot
has to be very stable.
 
On 5 Mar 2004 14:03:41 -0800, [email protected] (Brian
Huntley) may have said:

> The trick is, the level arms have to be long enough to accomodate the full range of movement, and
> the pivot has to be very stable.

And the relative step spacing must be the same for both the lever and the device. If there are
irregular steps on the gearhub in this case, the Shimano lever's steps will still produce the wrong
result for some gears even though the system may have the correct overall amount of total travel in
the bellcrank-to-hub section.

--
My email address is antispammed; pull WEEDS if replying via e-mail.
Typoes are not a bug, they're a feature.
Words processed in a facility that contains nuts.
 
On 3/5/04 6:42 PM, in article [email protected],
"Werehatrack" <[email protected]> wrote:

> On 5 Mar 2004 14:03:41 -0800, [email protected]
> (Brian Huntley) may have said:
>
>> The trick is, the level arms have to be long enough to
>> accomodate the full range of movement, and the pivot has
>> to be very stable.
>
> And the relative step spacing must be the same for both
> the lever and the device. If there are irregular steps on
> the gearhub in this case, the Shimano lever's steps will
> still produce the wrong result for some gears even though
> the system may have the correct overall amount of total
> travel in the bellcrank-to-hub section.

Right, right. So, can some additional sources corroborate
the assertion that an 8-speed Shimano indexed shifter pulls
3.07 mm per click? And does anybody have the specs for the
cable pull of the Nexus shifters or of a 7-speed shimano
indexed derailleur shifter?

It seems like maybe getting a machine shop to drill a new
retainer plate for a barcon would be a little easier than
coming up with a reliable bellcrank. But I'm having a hard
time visualizing the inside of the barcon anyway.. Hopefully
I'll come across one I can take apart. Barcons don't show up
too often in the piles of parts I regularly sift through,
let alone indexed ones..

Anyway, thanks for all the input. I'll get in touch if I
ever actually get started in earnest on this..

Y.
 
In article <[email protected]>,
[email protected] ( Doug Goncz ) wrote:

> Try, what is it, a Nexus-D automatic hub? No barcons at
> all.
>
> I've had the same desire.
>
> Try to get a barcon with drilled retaining plates holding
> ball detents. Then manufacture new plates precisely and
> it'll work. I can drill to 50 divisions and index to
> arbitrary divisions. So can your local machine shop. They
> use an optical comparator to determine the spacing of your
> detents.
>
>
>
>
>
> My physics project at NVCC: Google Groups, then "dgoncz"
> and some of: ultracapacitor bicycle fluorescent flywheel
> inverter

Just in case you haven't thought of this, you could do what
I did (following a picture I found somewhere on Sheldon's
site): mounted the twist shifter on a cut-down mountain bike
barend, then mounted the whole thing on the base of my stem
(1" steerer). Much simpler and safer, if not as elegant as
what you're trying to do.

art strum
 
So, here's what I've gathered so far:

- You need precise cable pull amounts to shift internally-
geared hubs.
- STI/Ergo shifters are precisely indexed for the cable
pull necessary to shift their corresponding derailers.
- If you can find the initial pull-per-click of an indexed
shifter, there are 2 ways you can probably change it:
- 1: replace the notched detent plate with a freshly-
machined one with new notches in it (assuming the
initial total cable pull of the device covers a wide
enough range for the intended new amount)
- 2: leave the shifter unmodified and come up with some
kind of lever (bellcrank) to modify the amount of pull
between the shifter and the cable anchor (this assumes
the target setting needs the same amount of pull for
each click)
- I can fairly easily find out the cable-pull specs of
various indexed shifters designed for derailers.

So this whole conversion thing may be do-able if I can find
out how much Nexus shifters pull, both per click, and over
their range. I haven't had any luck with this.

Do any of y'all out there have the details?

thanks.
Y.

On 3/4/04 10:56 PM, in article
BC6D64A1.3077%[email protected], "slug"
<[email protected]> wrote:

> Is there any really important reason why one shouldn't use
> a friction shifter with an internally-geared hub? I've
> been thinking about building a 7- or 8-speed coaster-brake
> nexus onto a 700c rim and throwing it onto a road frame
> with mustache bars. In the image in my head, it just looks
> best with a barcon.
...
 
Okay, I realize that last post was really long-winded, considering I really
wanted to ask just this:

Does anyone have the specs for the exact cable pull per
click (with details if it's not the same for each click) of
the 7- and/or 8-speed Nexus shifters?

(said info is conspicuously absent from bike.shimano.com).

Anybody? Please?

On 5/28/04 6:44 PM, "slug"
<[email protected]> wrote:

> So, here's what I've gathered so far:
>
> [blah blah blah blah blah]
>
> So this whole conversion thing may be do-able if I can
> find out how much Nexus shifters pull, both per click, and
> over their range. I haven't had any luck with this.
>
> Do any of y'all out there have the details?
>
> thanks.
> Y.
>
> [blah blah.]
 
slug wrote:
> So, here's what I've gathered so far:
> - You need precise cable pull amounts to shift internally-
> geared hubs.
> - STI/Ergo shifters are precisely indexed for the cable
> pull necessary to shift their corresponding derailers.
> - If you can find the initial pull-per-click of an
> indexed shifter, there are 2 ways you can probably
> change it:
> - 1: replace the notched detent plate with a freshly-
> machined one with new notches in it (assuming the
> initial total cable pull of the device covers a wide
> enough range for the intended new amount)
> - 2: leave the shifter unmodified and come up with
> some kind of lever (bellcrank) to modify the amount
> of pull between the shifter and the cable anchor
> (this assumes the target setting needs the same
> amount of pull for each click)
> - I can fairly easily find out the cable-pull specs of
> various indexed shifters designed for derailers. So
> this whole conversion thing may be do-able if I can
> find out how much Nexus shifters pull, both per click,
> and over their range. I haven't had any luck with this.
-snip- You might set up a shift system with a bare wire
someplace, paint the wire and a mark behind it on your frame
and just measure the travel per click. (I don't know whether
it is even or progressive travel or what the distances are).

Round Ergos and STi have a mechanism to pull slightly past a
gear and then settle back on the cog center. In an Ergo it's
the peg on the spring carrier - the one that breaks off. You
should obviate that feature for an internal gearbox, i.e.,
use a broken spring carrier.

--
Andrew Muzi www.yellowjersey.org Open every day since 1
April, 1971
 
On 5/30/04 8:43 PM, in article [email protected], "A Muzi"
<[email protected]> wrote:
> -snip- You might set up a shift system with a bare wire
> someplace, paint the wire and a mark behind it on your
> frame and just measure the travel per click. (I don't know
> whether it is even or progressive travel or what the
> distances are).
Yeah, I was figuring I could do that if I have to.. But I
think I want to do this with the 8-speed Nexus, which as far
as I can tell isn't available on its own yet. But it seems
like it couldn't hurt to try and dig up the tech info before
I can get my hands on the part. I thought maybe it was in
one of those newsletters Shimano sends to dealers..

I s'pose if I haven't scrounged up the specs by the time I
can buy the hub, I'll just do it myself..

> Round Ergos and STi have a mechanism to pull slightly past
> a gear and then settle back on the cog center. In an Ergo
> it's the peg on the spring carrier - the one that breaks
> off. You should obviate that feature for an internal
> gearbox, i.e., use a broken spring carrier.
I'm a teeny bit unclear on what you're saying here, but I
think I get it. by "round Ergos," I take it you mean the
last few years' models, in comparison to the older ones I've
heard described as pointy? And then as far as the spring
goes: I think you mean that there's no good reason for the
shifter to overpull the cable at first, right? I just want
the gear to click right into place? So if I use a round
Ergo, then I should find "the peg on the spring carrier" (I
think I could figure this out from an exploded drawing) and
break it off?

Indexed downtube or bar-end shifters wouldn't present this
issue at all, correct?

Thanks for all the patient explaining, Andrew (and the other
patient explainers I've seen in so many other r.b.t.
threads..)

later. Yoni.