Sturmy Archer 8 speed



Frank Krygowski writes:

>> This is a dead design. Let it rest in peace.


> Except that I've got a live one or two, with applications in mind.
> I'm interested in the real extent of the problem, and if justified,
> possible solutions.


As I suggested, take the planet pins (axles) and grind a one degree
funnel shaped taper on their stepped ends being careful to not take
off any more of the case hardening than needed. Then do the same for
the four leading faces of the "clutch, only so far in to where the
pins bear on it. That place should be obvious from polish markings
from use. 1 degree is about 0.0175inch per inch. Since this is a
matter of less than 1/10 inch it isn't much depth but it is a greater
angle than the elastic disengagement angle.

If you haven't had a failure and you don't intend to pedal hard while
standing at significant speed, it may never be a problem, especially
if your pedaling doesn't produce continuous torque. In that case the
clutch can home between each pedal stroke and will never pop out, as
it doesn't with most users (and I suppose the justification for not
changing the design) although the S5 hub did. However, that was
necessary because it uses only three planets but uses the same driver
parts with four quadrants.

Jobst Brandt
 
On Sat, 10 Dec 2005 00:48:07 GMT, Jasper Janssen <[email protected]>
wrote:

>On Thu, 08 Dec 2005 22:22:43 -0600, Tim McNamara <[email protected]>
>wrote:
>
>>IIUC they replaced the steel hub body
>>with an aluminum alloy body, also a nice touch.

>
>Actually, they make most of their hubs in both. The real S-A did that as
>well, somewhat: all their hubs were available in steel and some were
>*also* available in alu. The new one is different in that much more of the
>line is available in alu and that some hubs, the high end ones, are alu
>only.


It remains to be seen if they're good, though. I'll give Sun the
benefit of the doubt more readily than I would the old SA simply
because Sun's owners seem to understand that their products must be
good enough that the buyer feels like there's good value for the money
even several years down the road. That one factor alone bodes better
for Sun. The reported improvements that have already been
incorporated in the old SA designs are a step in the right direction.
If the Sun-designed hubs can match the overall durability of the best
of the SA units while providing more gears and fewer annoyances, they
shouldn't have any problem capturing a good slice of the market. If
they can do it while offering a choice of grip or trigger shifting,
they should sell well.
--
Typoes are a feature, not a bug.
Some gardening required to reply via email.
Words processed in a facility that contains nuts.
 
> A Muzi wrote:
>>-The key, I now believe, is that the AW pins are spinning and
>>at a low angle to the axis. They behave like a screw and
>>force the clutch upward until it travels over the end of the
>>pin. S-5 pins have a flatted side and do not spin. My
>>all-muscle employee and a young Jobst were able to exert
>>greater consistent pressure to drive that system to failure.


[email protected] wrote:
> Do the pins spin? I'd assumed the planet gears spun on the pins, and
> that the pins were stationary w.r.t. the planet carrier (or whatever
> it's called). And if the 5 speed has a similar mechanism with flats on
> one side of the pins, that seems to preclude spinning pins.
>
> Back to the Dremel idea: If the pins don't spin, putting that flat on
> the drive side of the pins wouldn't be extremely difficult. I'd try it
> with a very slight angle in the direction to fight disengagement.
> IIRC, motorcycle transmissions use just 1 degree or so on the
> components that are vaguely similar - the dogs that engage the sides of
> the constant-mesh gears. Admittedly, it would take some patience to
> take equal amounts of metal off each pin, and each arm of the clutch.
> But ISTM a slight angle would make the result pretty forgiving.
>
> The clutch doesn't look very precise to me when brand new. I'd think
> some skill with a Dremel and a little patience could make it work.
>
> Andrew, did your employee get it to jump out of gear immediately? If I
> go through the trouble of doing this (it would involve building a
> wheel, as I've got a naked hub) I'm wondering how long the "test" would
> have to last.


The pins do spin as can be easily seen on a worn set. Five
speed pins are flatted on one end so they can't.

I assumed you were takling about making a slightly conical
shape out of the cylindrical end. That would do it and it
needn't be much of an angle either. As Jobst mentioned, a
lathe with a grinder attachment would be ideal as they are
cyanide hardened steel.

I've never experienced the phenomenon myself. As I
understand it, try a powerful smooth cadence in high gear.
The conditions would be where the gears spin the large
diameter of the pin and once spinning the pin's angle to the
axis screws the clucth up and over the top.

If you do try that, I wonder if pins cut from drill rod,
without the necked-down end at the clutch side, would have
enough extra material to make the job easier.

--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971
 
> On 09 Dec 2005 05:40:21 GMT, [email protected] wrote:
>>First, the flexing of the axle AND the loose fit in the planet housing
>>generate a skew between "clutch" and the pins that it drives. That is
>>a disengaging mechanism.


Jasper Janssen wrote:
> Ah, of course, the pins themselves aren't made non-straight, the drive
> force just kicks them into a slanting position. That'd actually be fixable
> simply by moving the two holes (left/right) they fit in slightly. Make 'em
> so they're crooked when non-engaged and straight while engaged.


There is a slight skew. As the clutch touches the pin just
above the base it stays there under constant pressure. It is
free to walk up the pin a small amount with each rotation,
the skew of the driver/clutch/axle being slight but constant
and the pedal pressure constant.

Yes if the pins were made a conical shape, large end up , it
would prevent that but may introduce other problems,
probably faster wear at the edge. At any rate it would be a
major undertaking to make special pins.

--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971
 
> On 9 Dec 2005 10:19:19 -0800, [email protected] wrote:
>>The clutch doesn't look very precise to me when brand new. I'd think
>>some skill with a Dremel and a little patience could make it work.


Jasper Janssen wrote:
> Old clutches were very precise when brand new. The new ones (Taiwanese,
> presumably) are ****.
> At least when you judge by how sharp all the edges are.


Old machined clutches are indeed beautiful.
New sintered clutches are ugly but wear well, cost less,
work fine.
The 'new' model has been around 20 years without any problems.
--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971
 
> On 9 Dec 2005 10:19:19 -0800, [email protected] wrote:
>>Back to the Dremel idea: If the pins don't spin, putting that flat on
>>the drive side of the pins wouldn't be extremely difficult.


Jasper Janssen wrote:
> They might not spin, especially when they're being driven, but they're not
> keyed either, so you don't know which position they'll be in at any one
> time. They *can* spin, or at least rotate little bits on occasion.
>
> The advantage to that is that the highest-load and thus highest-wearing
> bits are swapped out regularly with the other half that doesn't wear at
> all.


Pins frequently come out shiny with striations showing they
do spin.

But the phenomenon is that the clucth, not being perfectly
aligned, creeps up with each rotation. It is free to move up
when both the pressure and skew angle are constant.

You need a constant forceful pressure to complete the
process all the way over the top of the pin.

--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971
 
>>Frank Krygowski writes:
>>>The clutch doesn't look very precise to me when brand new. I'd think
>>>some skill with a Dremel and a little patience could make it work.


> On 09 Dec 2005 19:50:42 GMT, [email protected] wrote:
>>It may have an unburnished shape but it is precisely flat and 90
>>degrees from face to face.

-snip-
>>This is a dead design. Let it rest in peace.


Jasper Janssen wrote:
> Not the new ones. The clutch is, what, 2-3 mm thick? The
> not-particularly-exact radius of the edges is now well over a mm. The
> clutch quite literally has a driving face that's more cylindrical than
> straight.
> It's less than academic for me in particular as there are quite literally
> millions of bicycles around here with them (ABs, that is, not AWs --
> difference is a drum brake on the other side plus the planet carrier is
> shorter left/right, no room between the planets and the pawls, unlike the
> AW carrier. The rest of the internals are identical). At a guess, about
> 30-50% or 6-10 million of the 21 million bikes we've got in this country
> of 16-17 million inhabitants.
>
> Also, my shopping/cinema bike will not have a derailer for vandalism,
> theft, and general functional reasons, I don't really want to use a single
> (though I might try it next time it gets stolen/breaks), and >3 speed hubs
> are way too expensive.
>
> On the other hand, it *is* pretty much academic since, at least in that
> usage, I won't encounter the problem, ever.


My own is an AB!
That few riders have experienced this says more about the
typical AW/AB rider than about the design. Still they are
cheap and reliable for the bulk of users. Mine will likely
outlast me!

--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971
 
> On 09 Dec 2005 18:59:12 GMT, [email protected] wrote:
>>No! It has nothing to do with rotation. Imagine two castellated cups
>>with their loosely but precise square teeth engaged, cups
>>face-to-face, one driving the other. If their intersecting axes are
>>at a slight angle, their square teeth will pump in and out of
>>engagement as they rotate, gradually pushing each other apart. As I


Jasper Janssen wrote:
> But if the teeth are designed for it, that doesn't happen (as on all
> direction-changing bevel gears). Is that just tooth shape or is there
> another factor? Wouldn't the effect be (at the very least) much greater is
> the (line the) axles (describe) didn't actually intersect, but passed
> slightly beside eachother?


Exactly!
A CV joint has shaped fingers which mesh properly.
The flat face of the clutch (yes even the new thin ones) can
walk up a flat face of the pin - in fact it is forced up as
it meets at a slight angle. If the pressure is constant it
will continue to climb with each rotation.

--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971