Sturmy Archer 8 speed

Dan Burkhart said:

I was just checking out the specs on this thing on the Sturmy Archer web
site. Shimano and Sram will be playing catch up now

You were, I take it, unaware that Sun borged Sturmey-Archer some time
ago, then? Note the contact email address, info at sunrace dot com.
--
Typoes are a feature, not a bug.
Some gardening required to reply via email.
Words processed in a facility that contains nuts.

Borged? A misleading representation of how Sun recycled Sturmey Archer
from a dust bin after SA was neglected for many years and robbed by con
artists. Sun has done for SA what SA needed for 75 years. Put money
in to engineering new and improved designs.

An all overdrive design may be more efficient or cheaper to produce.
Anyhow I must rejoice that SA is born again and not another great
company gone for good. I have 1 new AW 3 speed hub on a bike. It has
a beautiful aluminum shell, sealed bearings and the dreaded neutral
between 2 & 3rd is no more. Works perfectly.

Rick Paulos said:

Borged? A misleading representation of how Sun recycled Sturmey
Archer from a dust bin after SA was neglected for many years and
robbed by con artists. Sun has done for SA what SA needed for 75
years. Put money in to engineering new and improved designs.

Quoted message said:

An all overdrive design may be more efficient or cheaper to produce.
Anyhow I must rejoice that SA is born again and not another great
company gone for good. I have 1 new AW 3 speed hub on a bike. It
has a beautiful aluminum shell, sealed bearings and the dreaded
neutral between 2 & 3rd is no more. Works perfectly.

Well, the sealed bearings aren't an improvement, judging from history.
SA hubs had excellent non-contact labyrinth seals that worked so well
that I have never found a hub from the 1950's that had any dirt or
water intrusion. So that's a step backward. As to the, jumping out
of top gear, I'm curious about what deign change was made to prevent
this. I suspect it is only so because your unit is new.

Have you taken it apart, and is it the same design as before, all held
together using no screws or fasteners. If so, I would like to know
what the "clutch" (driver cross) shape is that might prevent popping
out of top gear under continuous high torque. This syndrome plagued
early auto transmission until someone realized that through elasticity
and shafts flex, gears tend to disengage in the absence of a positive
retention design.

Jobst Brandt

Quoted message said:

Well, the sealed bearings aren't an improvement, judging from history.
SA hubs had excellent non-contact labyrinth seals that worked so well
that I have never found a hub from the 1950's that had any dirt or
water intrusion. So that's a step backward. As to the, jumping out
of top gear, I'm curious about what deign change was made to prevent
this. I suspect it is only so because your unit is new.

Sturmey's invented numerous ways to do without the blind spot over the
years. Actually, the first units in the 18th century didn't have it, by
using three sets of pawls, and it was introduced as a cost-cutting
measure. One of the ingenious ways that would be virtually free I've seen
a patent for involved slightly longer teeth on the ring wheel thingy with
edges beveled, so that if the star clutch traveled right far enough to
disengage from the planet wheel axles it would be forced back onto the
planets, or similarly if it traveled left far enough to get out of second
it'd be forced onto three.

Also, better triggers & cables mean more positive engagement and fewer
opportunities for the star clutch to wear in the first place.

If you mean the diagonally worn star clutch climbing up from the planet
wheel axle ends under force, I can think of a few ways to remedy that
problem: 1) service the damn thing and replace the star clutch when it's
done enough miles to be iffy (I always replace it and the pawls when I
have one open anyway for other reasons -- usually the thread between the
indicator rod and the little beam being gone) 2) make sure the shifter's
appropriately adjusted at all times, and don't shift under power 3) make
the ends of those axles about twice as long and/or the star clutch half as
thick and edit cable pull appropriately. The problem only exists because
the top of the star clutch teeth is even with or above (especially when
adjusted too-tight) the ends of the axle ends. The most diagonally worn
star clutch in the world won't climb up over axle ends that are a bit
higher 4) give the planet wheel axle mounts a slot as key, so that they
don't rotate (or put in the axles from the other side and just have tabs
on the planet wheel housing), and file a reverse-diagonal slope onto them
that acts toward positive engagement rather than against it.

Anyway, as long as you're prepared to throw backwards compatibility for at
least *some* of the bits out of the window it's entirely possible to do
it.

Jasper

Jasper Janssen said:

Quoted message said:

Well, the sealed bearings aren't an improvement, judging from
history. SA hubs had excellent non-contact labyrinth seals that
worked so well that I have never found a hub from the 1950's that
had any dirt or water intrusion. So that's a step backward. As to
the, jumping out of top gear, I'm curious about what deign change
was made to prevent this. I suspect it is only so because your
unit is new.

Quoted message said:

Sturmey's invented numerous ways to do without the blind spot over
the years.

I'm sorry, I don't understand what you mean by "blind spot" and what
that has to do with the basic flaw in torque transmission in AW hubs.
The design flaw is a basic misunderstanding of gear engagement in
general.

Quoted message said:

Actually, the first units in the 18th century didn't have it, by
using three sets of pawls, and it was introduced as a cost-cutting
measure.

I'm sorry, I think we are not on the same subject. The failure mode
is one in which driving elements on the same but elastically bent
shaft, push each other apart. This has nothing to do with ratchet
pawls. The bending load comes from the chain pulling on a sprocket
that lies between support bearings for that shaft.

Quoted message said:

One of the ingenious ways that would be virtually free I've seen a
patent for involved slightly longer teeth on the ring wheel thingy
with edges beveled, so that if the star clutch traveled right far
enough to disengage from the planet wheel axles it would be forced
back onto the planets, or similarly if it traveled left far enough
to get out of second it'd be forced onto three.

That does not prevent the type of failure that causes the problem.
Just making it so there is no undefined drive location cannot occur
because otherwise it could be in two gears at once.

Quoted message said:

Also, better triggers & cables mean more positive engagement and
fewer opportunities for the star clutch to wear in the first place.

That falls into the response that SA gave over the many years of these
failures... "you mis-adjusted the shift linkage." That claim that was
repeated endlessly over the years is false. Not only is it a
misunderstanding of the failure mode but it defies the definition of a
safe design, one that cannot throw you onto the street with a slight
misadjustment.

You'll notice that most of today's hub gears do not take the
expedient means of entering the hub by using a chain through the axle.
That in itself was asking for failure because that place is often
banged against hard objects. Just think how many smashed shift chains
bicycle shops replaced over the years. That was one of the most
common failures and it was usually noticed by failure to shift
correctly and to jump out of gear.

Quoted message said:

If you mean the diagonally worn star clutch climbing up from the planet
wheel axle ends under force, I can think of a few ways to remedy that
problem:

Quoted message said:

1) service the damn thing and replace the star clutch when it's done
enough miles to be iffy (I always replace it and the pawls when I
have one open anyway for other reasons -- usually the thread between
the indicator rod and the little beam being gone)

Quoted message said:

2) make sure the shifter is appropriately adjusted at all times, and
don't shift under power

Quoted message said:

3) make the ends of those axles about twice as long and/or the star
clutch half as thick and edit cable pull appropriately. The problem
only exists because the top of the star clutch teeth is even with or
above (especially when adjusted too-tight) the ends of the axle
ends. The most diagonally worn star clutch in the world won't climb
up over axle ends that are a bit higher

Quoted message said:

4) give the planet wheel axle mounts a slot as key, so that they
don't rotate (or put in the axles from the other side and just have
tabs on the planet wheel housing), and file a reverse-diagonal slope
onto them that acts toward positive engagement rather than against
it.

You are not aware of the mechanism at work here or you wouldn't say
all that... in true SA style. I rode these hubs to school for years
and got all that good advice that you repeat practically from the SA
bible. I used new driver crosses, new planet pins and went over the
bars anyway in sprints on two occasions besides failures while seated.

Quoted message said:

Anyway, as long as you're prepared to throw backwards compatibility
for at least *some* of the bits out of the window it's entirely
possible to do it.

It took me a while to be convinced that there are technical matters
which manufacturers and their minions do not understand. I came
across the same disengagement problem on some older motorcycle gear
boxes. What occurred there was that the engine revved damn high when
it jumped out under full throttle. That hurt less than falling from
the bicycle.

If the SA planet shafts had a slightly funnel shaped tip and the
driver a matching slant, torque would force a more secure engagement
rather than the driver weaseling its way out of engagement because its
rotation axes was skewed a small fraction of a degree from the parts
it was driving. This skew is elastic and is no longer present when
the hub is disassembled to prove to the user that he is the culprit.

Jobst Brandt

Quoted message said:

Borged? A misleading representation of how Sun recycled Sturmey Archer
from a dust bin after SA was neglected for many years and robbed by con
artists. Sun has done for SA what SA needed for 75 years. Put money
in to engineering new and improved designs.

An all overdrive design may be more efficient or cheaper to produce.
Anyhow I must rejoice that SA is born again and not another great
company gone for good. I have 1 new AW 3 speed hub on a bike. It has
a beautiful aluminum shell, sealed bearings and the dreaded neutral
between 2 & 3rd is no more. Works perfectly.

Funny, I was just re-reading Tony Hadland's page about S-A, then came
over here to read this thread. The page shows how the 3-speed AW works
(older version, anyway), and how the dreaded "no-drive" event happens.

http://www.users.globalnet.co.uk/~hadland/elegy.htm

Mark

Quoted message said:

You are not aware of the mechanism at work here

Yes that's entirely possible so goddamn explain what the hell you're
talking about instead of sniping ten times a post that I don't understand
what you're saying. *breathe*

What we have here is a failure to communicate, and in this particular
case, it's you that's failing to communicate what you're talking about,
not me that's failing to understand it.

Jasper

Quoted message said:

Quoted message said:

4) give the planet wheel axle mounts a slot as key, so that they
don't rotate (or put in the axles from the other side and just have
tabs on the planet wheel housing), and file a reverse-diagonal slope
onto them that acts toward positive engagement rather than against
it.

[...]

Quoted message said:

If the SA planet shafts had a slightly funnel shaped tip and the
driver a matching slant, torque would force a more secure engagement
rather than the driver weaseling its way out of engagement because its
rotation axes was skewed a small fraction of a degree from the parts

That's exactly what I just suggested! How can I be misunderstanding what's
happening so terribly if I come up with the same solutions as you?

Jasper

Jasper Janssen said:

Quoted message said:

Quoted message said:

4) give the planet wheel axle mounts a slot as key, so that they
don't rotate (or put in the axles from the other side and just have
tabs on the planet wheel housing), and file a reverse-diagonal slope
onto them that acts toward positive engagement rather than against
it.

Quoted message said:

Quoted message said:

If the SA planet shafts had a slightly funnel shaped tip and the
driver a matching slant, torque would force a more secure
engagement rather than the driver weaseling its way out of
engagement because its rotation axes was skewed a small fraction of
a degree from the parts

Quoted message said:

That's exactly what I just suggested! How can I be misunderstanding
what's happening so terribly if I come up with the same solutions as
you?

I didn't find that you suggested this in your numbered points that
explain how to avoid forward freewheeling. Point 4.) above does not
address that issue nor does the web site:

http://www.users.globalnet.co.uk/~hadland/elegy.htm

From that I have heard and read, the cause for disengagement under
load is not understood. I've spent much time discussing this with
people claiming to be SA experts and found that the concept of a
skewed driver causing disengagement is not understood. I have not
seen it mentioned in anyone's writing. Only that the shift mechanism
was surely not adjusted correctly. Even with the shift "chain"
removed, the driver (clutch) has only its engagement spring pressing
it to the home position so there cannot be any misadjustment. In that
condition these hubs also jump out of gear.

Jobst Brandt

This has happened to me, slipping in 3rd even with the indicator chain
removed!

I had things adjusted properly so the no-drive zone was avoided, and
then found the hub slipping in third. I took the chain out to make sure
it stayed in that gear, and sure enough it would spin as it should,
slip once, then re-engage. This would happen at random, and always
under more pressure. Not very confidence inspiring when trying to clear
an intersection in a hurry...

The hub stays in 2nd without issue - no slipping into the no-drive
zone, but then into 3rd and it's a guessing game. I'm told this is due
to wear on the inside of the hub's shell, so it's not just a matter of
replacing the guts, but the whole hub, which means a wheel rebuild.
Arghh...

It's almost enough to make me rebuild it with a old Shimano 3-speed!
(Too cheap to buy a new internal hub, and I like recycling whenever
possible.) But everything I've read says those hubs can be problematic.
Too bad, because they don't have the no-drive zone, and the shifter is
easier to use (this is my winter bike, and of course I wear thick
gloves). So I guess I'll install another AW and hope for the best. Or
try the Shimano. Hmm...

Ideally I would go single-speed or fixed for winter use, but being
car-free I need to tow a trailer every now and then, and need a few
extra gears.

Mark

email hidden said:

Borged? A misleading representation of how Sun recycled Sturmey Archer
from a dust bin after SA was neglected for many years and robbed by con
artists. Sun has done for SA what SA needed for 75 years. Put money
in to engineering new and improved designs.

That's one view. The other is that Sun, having no rep in gearhubs,
didn't think that their image as a supplier of second- and third-line
components would give them any leverage in the gearhub market if they
took the plunge on their own. But by assimilating the corpse of SA
and acquiring its name in the process, they could cherry-pick what
could be reproduced of the SA line to gain immediate entry as a
gearhub supplier with kit that was known if not universally loved.
I'm sure that their plan was to bank on SA's name recognition to make
it easier to get their own designs into the market once the legacy
products had paved the way. All that's left of the old SA is the name
and a few bits that will likely get cast off as Sun gets its own hubs
out. We can only hope that they will avoid duplicating (or adding to)
the errors made in SA's history.

Quoted message said:

An all overdrive design may be more efficient or cheaper to produce.
Anyhow I must rejoice that SA is born again and not another great
company gone for good. I have 1 new AW 3 speed hub on a bike. It has
a beautiful aluminum shell, sealed bearings and the dreaded neutral
between 2 & 3rd is no more. Works perfectly.

--
Typoes are a feature, not a bug.
Some gardening required to reply via email.
Words processed in a facility that contains nuts.

Quoted message said:

If the SA planet shafts had a slightly funnel shaped tip and the
driver a matching slant, torque would force a more secure engagement
rather than the driver weaseling its way out of engagement because its
rotation axes was skewed a small fraction of a degree from the parts
it was driving.

Is this the kind of thing that could be done by a person with a Dremel
tool? It doesn't sound like it would have to be super-accurate.

- Frank Krygowski

Mark who? said:

This has happened to me, slipping in 3rd even with the indicator
chain removed!

Quoted message said:

I had things adjusted properly so the no-drive zone was avoided, and
then found the hub slipping in third. I took the chain out to make
sure it stayed in that gear, and sure enough it would spin as it
should, slip once, then re-engage. This would happen at random, and
always under more pressure. Not very confidence inspiring when
trying to clear an intersection in a hurry...

Quoted message said:

The hub stays in 2nd without issue - no slipping into the no-drive
zone, but then into 3rd and it's a guessing game. I'm told this is
due to wear on the inside of the hub's shell, so it's not just a
matter of replacing the guts, but the whole hub, which means a wheel
rebuild. Arghh...

Not true. It is the interface between the driver cross (clutch) and
the planet pins. As I said, even new ones will pop out of engagement.
There is no way for the hub to disengage in low and middle gear.

Quoted message said:

It's almost enough to make me rebuild it with a old Shimano 3-speed!
(Too cheap to buy a new internal hub, and I like recycling whenever
possible.) But everything I've read says those hubs can be
problematic. Too bad, because they don't have the no-drive zone,
and the shifter is easier to use (this is my winter bike, and of
course I wear thick gloves). So I guess I'll install another AW and
hope for the best. Or try the Shimano. Hmm...

I've never gone back to hub gears after my experience with SA so I
never took a Shimano hub apart to see how they solved the problem...
if in fact they did so.

Quoted message said:

Ideally I would go single-speed or fixed for winter use, but being
car-free I need to tow a trailer every now and then, and need a few
extra gears.

I'm satisfied with a derailleur.

Jobst Brandt

Frank Krygowski said:

Quoted message said:

If the SA planet shafts had a slightly funnel shaped tip and the
driver a matching slant, torque would force a more secure
engagement rather than the driver weaseling its way out of
engagement because its rotation axes was skewed a small fraction of
a degree from the parts it was driving.

Quoted message said:

Is this the kind of thing that could be done by a person with a
Dremel tool? It doesn't sound like it would have to be
super-accurate.

No. Two reasons are that the pins are case hardened and would need to
be modified on a lathe (possibly by grinding) and the driver cross
would take some careful mill machining to make all four faces exactly
located and slanted (ever so slightly).

Jobst Brandt

On 08 Dec 2005 07:13:52 GMT, [email hidden] wrote:

First off, I'd like to apologise for my previous postings. I wasn't
thinking very clearly, seeing as how '8 am' and 'late at night' as opposed
to 'early in the morning' are really not things that should go together.

Quoted message said:

From that I have heard and read, the cause for disengagement under
load is not understood. I've spent much time discussing this with
people claiming to be SA experts and found that the concept of a
skewed driver causing disengagement is not understood. I have not
seen it mentioned in anyone's writing. Only that the shift mechanism
was surely not adjusted correctly. Even with the shift "chain"
removed, the driver (clutch) has only its engagement spring pressing
it to the home position so there cannot be any misadjustment. In that
condition these hubs also jump out of gear.

Well, the mechanism that's generally suggested (and what I was riffing
off) is that from inevitable wear due to slight misadjustments and various
other things, the star clutch teeth wear until they are, at least to a
certain extent, not straight edged but diagonal (in the direction that
would tend to force them off the planet wheels). I've seen pictures of a
star clutch whose teeth were worn at probably a 20 or more degree angle,
though I never saw one of those in the flesh.

I suppose it's possible that under extreme pedaling loads, the axle flexes
in such a way so as to cause the clutch to ride up (incidentally, wouldn't
strengthening the engagement spring cause the problem to move to high
enough loads that it would no longer be a problem?). Regardless of whether
theory one or two is the culprit, design changes that fix one should
pretty much also fix the other. Making the notches (Planet axle ends) that
the clutch engages with higher relative to the driver and/or giving a
positive engagement bevel to star clutch and/or notches should work, no?

Incidentally, returning to the problem of the no-drive gap between 2 and
3. If you make the second gear notches extend farther left, without
changing anything else, and give them an angle at the end, it would be
possible for the star clutch to momentarily be in between gears 2 and 3
and engaging them both, but wouldn't the faster-turning 2nd gear notches
immediately press the star clutch down into the third gear engagement,
thus removing the lock-up situation? Given that you're supposed to shift
with the pedals stationary and that if you do that the entire innards are
standing still, I think you might even be able to shift back into 2.

Jasper

Jasper Janssen said:

Quoted message said:

From that I have heard and read, the cause for disengagement under
load is not understood. I've spent much time discussing this with
people claiming to be SA experts and found that the concept of a
skewed driver causing disengagement is not understood. I have not
seen it mentioned in anyone's writing. Only that the shift
mechanism was surely not adjusted correctly. Even with the shift
"chain" removed, the driver (clutch) has only its engagement spring
pressing it to the home position so there cannot be any
misadjustment. In that condition these hubs also jump out of gear.

Quoted message said:

Well, the mechanism that's generally suggested (and what I was
riffing off) is that from inevitable wear due to slight
misadjustments and various other things, the star clutch teeth wear
until they are, at least to a certain extent, not straight edged but
diagonal (in the direction that would tend to force them off the
planet wheels). I've seen pictures of a star clutch whose teeth
were worn at probably a 20 or more degree angle, though I never saw
one of those in the flesh.

I wouldn't call the "teeth" and as you can see this is a cross with
flat faces that drive against round pins, which already presents a
contact problem in itself. These pins are not a press fit in
their housing so they already have the ability to skew with respect to
the hub axis to which they are theoretically parallel.

http://www.users.globalnet.co.uk/%7Ehadland/sa/saaw.pdf

Part 510 Planet pin is driven by part 505A Sliding clutch which is
driven by K462, on which the drive sprocket is mounted.

Quoted message said:

I suppose it's possible that under extreme pedaling loads, the axle flexes
in such a way so as to cause the clutch to ride up (incidentally, wouldn't
strengthening the engagement spring cause the problem to move to high
enough loads that it would no longer be a problem?).

Disengagement forces are large enough to overcome any reasonable spring
load but only if torque is continuous to prevent allowing the designed
spring from re-engaging the clutch to its home position. As it turns
out, an athletic pedaler is able to do this and that is where the
failure occurred most reliably... with brand new parts.

Quoted message said:

Regardless of whether theory one or two is the culprit, design
changes that fix one should pretty much also fix the other. Making
the notches (Planet axle ends) that the clutch engages with higher
relative to the driver and/or giving a positive engagement bevel to
star clutch and/or notches should work, no?

Notches will not do because that would make exiting that gear position
impractically difficult. As I said, a slight taper to the contact
faces would have done it but then driving round pins against a flat
surface also does not guarantee durability. As you suggest, the
clutches of used hubs have wear marks at these contacts, some of which
are rounded from having disengaged under load. That is the only way a
rounded ramp (exit) can be generated. Just that alone should have
kicked the hubris out of the SA engineers sails, but it didn't. They
were so sure of the correctness of their design that they attributed
the failures to faulty user adjustment of the shift chain.

Quoted message said:

Incidentally, returning to the problem of the no-drive gap between 2
and 3. If you make the second gear notches extend farther left,
without changing anything else, and give them an angle at the end,
it would be possible for the star clutch to momentarily be in
between gears 2 and 3 and engaging them both, but wouldn't the
faster-turning 2nd gear notches immediately press the star clutch
down into the third gear engagement, thus removing the lock-up
situation? Given that you're supposed to shift with the pedals
stationary and that if you do that the entire innards are standing
still, I think you might even
be able to shift back into 2.

Not possible in this mechanism. The 5-speed hub was better in this
respect, not trying to drive the ends of planet pins. Just the same,
amazingly many parts are the same for both the 3 and 5-speed hubs
especially the housing.

Jobst Brandt

Thakjs for the clarification as to why mine slips in 3rd. I replaced
the clutch twice, and when that didn't solve the problem I got [censored] enough to yank the innards and replaced them from another hub. Of
course, those had wear as well...

I'd be happy with a derailer as well, but where I live (Ottawa, Canada)
the winter temps are a mite brisk (I've cycled at -25C). Derailers tend
to get gunked up with slush, and sometimes freeze in place, as do the
hub's pawls (they stick so that you get freewheeling in both directions
- not fun).

I've never had the AW hub not shift in the cold, and never had the
pawls stick either. I guess I have three choices: either try one more
clutch and pin replacement, try the Shimano hub, or finally give in and
buy a new model. I have access to lots of the old hubs (I volunteer at
our local bike recycling place), so it comes down to spending time or
spending money...

Mark

Werehatrack said:

email hidden said:

Borged? A misleading representation of how Sun recycled Sturmey
Archer from a dust bin after SA was neglected for many years and
robbed by con artists. Sun has done for SA what SA needed for 75
years. Put money in to engineering new and improved designs.

That's one view. The other is that Sun, having no rep in gearhubs,
didn't think that their image as a supplier of second- and
third-line components would give them any leverage in the gearhub
market if they took the plunge on their own. But by assimilating
the corpse of SA and acquiring its name in the process, they could
cherry-pick what could be reproduced of the SA line to gain
immediate entry as a gearhub supplier with kit that was known if not
universally loved.

Well, you had your Cynic-Os for breakfast. They not only bought the
name, they bought the tooling and hired some of the staff according to
news reports at the time. And so what if your version is accurate?
It's a product in some demand, serves a useful purpose, and should be
carried on.

Quoted message said:

I'm sure that their plan was to bank on SA's name recognition to
make it easier to get their own designs into the market once the
legacy products had paved the way. All that's left of the old SA is
the name and a few bits that will likely get cast off as Sun gets
its own hubs out. We can only hope that they will avoid duplicating
(or adding to) the errors made in SA's history.

By reputation they have fixed the neutral gear problem by replacing
the old clutch mechanism with a ball-locking mechanism in the S-A 3
speed. That's a good thing; S-A resisted this simple fix for decades,
preferring to blame the user. IIUC they replaced the steel hub body
with an aluminum alloy body, also a nice touch. And Sun-SA have
pursued new products and potentially new markets. And, for those of
us still with freewheels, they remain one of the few makers on the
planet to bother with making freewheels.

Quoted message said:

This has happened to me, slipping in 3rd even with the indicator chain
removed!

I had things adjusted properly so the no-drive zone was avoided, and
then found the hub slipping in third. I took the chain out to make sure
it stayed in that gear, and sure enough it would spin as it should,
slip once, then re-engage. This would happen at random, and always
under more pressure. Not very confidence inspiring when trying to clear
an intersection in a hurry...

The hub stays in 2nd without issue - no slipping into the no-drive
zone, but then into 3rd and it's a guessing game. I'm told this is due
to wear on the inside of the hub's shell, so it's not just a matter of
replacing the guts, but the whole hub, which means a wheel rebuild.
Arghh...

It's almost enough to make me rebuild it with a old Shimano 3-speed!
(Too cheap to buy a new internal hub, and I like recycling whenever
possible.) But everything I've read says those hubs can be problematic.
Too bad, because they don't have the no-drive zone, and the shifter is
easier to use (this is my winter bike, and of course I wear thick
gloves). So I guess I'll install another AW and hope for the best. Or
try the Shimano. Hmm...

Ideally I would go single-speed or fixed for winter use, but being
car-free I need to tow a trailer every now and then, and need a few
extra gears.

Although Jobst had a less than positive experience I've
never had that happen in a literal lifetime of commuting on
Sturmey hubs. Not once.

I am still unclear about the syndrome.

Let's rule out any worn bits such as the clutch and pins (
pedaling while shifting can chip the edges). And Jobst
agrees the easily-kinked control wire is not at fault here.

Jobst says a generally sloppy fit and axle deflection can
allow the driver and clutch to move - where? Not much room
inside!

A Shimano is smaller and the parts fit more closely so I
suppose that's an alternate. Jim Adney's opinion of Shimano
hubs is that if they saw any care or lubrication they might
have worked. Since 'better' bikes all had Sturmey in the
day, Shimano hubs on XMart bikes were poorly adjusted and
seldom lubricated. My own opinion is not as generous.

[I haven't talked with Jim about this in a good long while.
He may have further thoughts]

You were told, "this is due to wear on the inside of the
hub's shell". Since the high and direct gears are driven
through the same pawl/ratchet set so that can't be true.

I can accept that there is something wrong in there (no more
credible observer than Mr Brandt) but I still don't get it
and this is the third iteration in as many years.
--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971