From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: three speed shifting
Date: 14 May 1998 19:24:09 GMT

Colin Kenny writes:

> I have a 1962 Schwinn Racer. I am rebuilding this bike to be nicely
> functional for my wife who loves cruisers. The two problems I have
> are with the three-speed hub and the kickstand.

Can the kickstand if you plan to do any longer riding or in the hills.
It weighs as much as two water bottles that I wouldn't volunteer to
carry for the pleasure of not leaning my bike against a wall or laying
it on the ground.  It can be extracted by removing the clip in the
back side of the mounting tube.  Don't worry, if you like bicycling,
you'll soon get a better bicycle.

> I greased the bearings on the hub and flushed out the innards with
> WD-40. It shifts somewhat stiffly, and not at all under load (I did
> replace the cables).

A Sturmey Archer hub gear will only shift under no load.  Pulling the
shift cable under load requires moving the internal driving cross,
that transmits all the torque, when it is fully loaded.  It doesn't
want t can be done because the spring moves the
cross whenever you let up on the pedals, which usually occurs once every
half revolution of the pedals.  That is, you can pre-shift and it will
drop in whenever your pedaling gives it a break.


What may not have happened yet, is that in top gear, even with a
perfectly adjusted and serviced hub, the driver can pop out of gear,
leaving you free wheeling forward.  If you are standing when this
occurs, it always causes a summersault.  This is why these gears are
not used in sports or hard touring.

> If I have to repair or replace parts on the inside of the hub, what
> should I know about them?

Pick a clean well lighted place to disassemble it.  There are fine
wire hairpin springs that easily vanish if you are not careful.  Other
than that, it is a nice mechanism that can be educational.  My
children had one that was clean and oil free that they used as an
assembly puzzle.  You should have a little mechanical adeptness to
discover the logic of it.  Use a hammer and flat blade screw driver to
unscrew the right hand cover that has a double lead right hand thread.
The unit is not liquid proof, so any excess oil will run out.  Use 30W
motor oil to lubricate it.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: Frank Krygowski <frkrygow@cc.ysu.edu>
Newsgroups: rec.bicycles.tech
Subject: Sturmey Archer dangerous?
Date: Fri, 31 Mar 2000 12:47:45 -0500

In a thread on a different internal gear hub, Jobst Brandt said:

> My experience with hub gears is limited
> to SA three and five-speed hubs, the three-speed hub being a dangerous
> and unrepenting misbegotten hub useful only for sedentary light
> pedaling in top gear.  Don't try sprinting in top gear.  That is its
> fatal flaw.

> Jobst Brandt      <jbrandt@hpl.hp.com>

This caught my attention. I've done only a little riding on a SA hub,
but I was considering building a SA three speed that I've got into a
folding bike.  This isn't a performance bike, but still: What goes
wrong, specifically?

--
Frank Krygowski        frkrygow@cc.ysu.edu

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 31 Mar 2000 18:43:48 GMT

Frank Krygowski writes:

> This caught my attention. I've done only a little riding on a SA hub,
> but I was considering building a SA three speed that I've got into a
> folding bike.  This isn't a performance bike, but still: What goes
> wrong, specifically?

The drive is transmitted from the sprocket through a four slot cup
through which a small cross protrudes to either drive the ring gear by
lifting ratchet pawls (low the gear that does not click when running),
performing the same operation with the pawls engaged (direct drive),
or fully extended against the four pins on which the planets ride to
drive the planets, hence the reciprocal relationship between low and
high to direct drive down 1/3 up 1/4 on the AW.

The four pins are free fit in the housing and when loaded are slightly
non perpendicular to the hub axis due to canting within the clearance.
This in itself has a disengaging bias to the driver cross that pushes
on them.  However, under load, the axle and mechanism bends slightly so
that depth of engagement of the driver cross to the pins varies during
each rotation.  These two effects disengage the driver from the pins
under high torque and drop the mechanism into free wheeling forward.
The result is that the rider, if standing, dives over the bars, the
bicycle following behind him.

This condition is apparent upon examining the driver and pins that
both become worn in a slant that enhances disengagement, however,
replacing these parts does not resolve the condition.  SA says that
the shift cable was mis-adjusted, a specious dodge if ever there was
one.  With the cable disconnected, the driver cross is free to make
perfect contact with the face of the planet carrier, the best
adjustment possible for top gear, and still disengage under load.

Had the driver been sloped and pins been made with matching flared
tapered ends, this would not have been an issue although misadjustment
could still leave one teetering at the edge of freewheeling forward.

Similarly, the springless ratchet of the SW (Silent) hub was sensitive
to lubricant viscosity and with anything more than 10W oil could
freewheel forward, the pawls clinging to the ramps by oil viscosity
while not engaging.  This hub was discontinued after a short run
probably because one could not place blame on user error.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 1 Apr 2000 01:05:28 GMT

Rick Paulos writes:

> I have ridden SA 3-speeds for years and I learned the hard way that
> you cannot trust them 100%.  Try shifting very slowly between 3rd
> and 2nd gear and you will find 'neutral'.!!!  Careful adjustment
> will not eliminate neutral while shifting from 2 to 3 or from 3 to
> 2.  Proper gear engagement relies on the indents in the shifter for
> proper positioning.  A improperly adjusted shifter will give you
> neutral in one of the gear positions.  One evening I was sprinting up
> a very steep hill and the hub went into neutral.  Probably due to
> excessive frame flex pulling on the shift cable.  Wound up in the
> emergency room.

As I mentioned, this is what SA would have you believe but, it isn't
true.  The mechanism WILL disengage under continuous heavy load no
matter how you adjust it.  With use, this tendency increases.

> Know this problem and respect it and you will never have any
> trouble.  Keep the hub oiled and check the cable adjustment often.
> AND DON"T STAND UP AND PEDAL.

That's a pretty confining restriction and one that reveals the basic
weakness of this hub.  As I said, had SA faced up to it instead of
blaming the victim, they could have prevented this.  Now they can't do
it or someone would sue them for sitting on this for 50 years.

> BTW, the older Shimano 3-speed hubs don't have the same problem.

I have no experience with them, but that doesn't absolve SA of
physical endangerment.

> I had a SW hub with it's springless pawls.  What an awful design.
> Never could get it to work well enough to even ride.

Use 10W oil.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 3 Apr 2000 16:52:52 GMT

Jim Adney writes:

>> The four pins are free fit in the housing and when loaded are
>> slightly non perpendicular to the hub axis due to canting within
>> the clearance.  This in itself has a disengaging bias to the driver
>> cross that pushes on them.  However, under load, the axle and
>> mechanism bends slightly so that depth of engagement of the driver
>> cross to the pins varies during each rotation.  These two effects
>> disengage the driver from the pins under high torque and drop the
>> mechanism into free wheeling forward.  The result is that the
>> rider, if standing, dives over the bars, the bicycle following
>> behind him.

> The cross is called the clutch and the pins are the pinion pins
> because they also serve as the axles for the planet pinions.

Thanks but nomenclature won't change the situation.

> Are you saying that the chain force causes the axle to bow thus
> leading these parts to no longer rotate about a common axis. Then the
> clutch walks slowly up the pinion pins until it escapes into the
> neutral between 3rd and 2nd?

That is exactly what I said and am aware from working in the auto
industry, that this is a cause of jumping out of gear in manual shift
cars, something that is well understood today and no longer occurs.

> I am very familiar with these hubs having serviced hundreds if not
> thousands. I have never seen this problem, but I suspect it is real.
> It is probably not common simply because you are probably the only
> person who has pushed them this hard.

I suggest you disassemble a few and look at the cross and pins, where
I am sure you will find the driving surface formed into a rejection
ramp.  I have not found a used hub that did not have this and this is
why components that I always replaced on hubs on which I worked.
These were hubs that I retrieved from bicycles discarded for lack of a
"10-speed".

> For how long, or for how may strokes, do you need to push an AW hub to
> experience this? I'm inclined to take your word for it rather than try
> the experiment first hand.

Just stand up and sprint over a short hill.  I did not use toeclips on
these.  After the second dive, I decided on a more cautious approach
after having placed an order for a derailleur bicycle.

> I was also interested in learning that 10W oil is the secret to
> making an SW hub work correctly.

Viscosity causes the pawls to adhere to the ratchet face and glide over
the teeth instead of residing in their cusps and "remembering" their
orientation from having passed under the previous tooth coasting.
When rotation again goes forward, the pawls follow the contour
backward through the engagement position.  That the pawl can pass
under a ratchet tooth coasting is apparent, that it can do this in
reverse is also apparent if one considers viscous attraction to the
wrong contour.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 3 Apr 2000 17:08:04 GMT

Mark Pace writes:

> S-A hubs have a potentially lethal "neutral" position between their
> top and middle gears.  If you are unlucky enough to 'find' this
> position, while, say, sprinting uphill, you could land on your ear,
> as if you had broken a pedal or a chain.

> HOWEVER, I've ridden them lots, and vigorously, and the only times
> this happened to me (twice) was: once when attempting to use a
> friction device instead of the official SA trigger (dumb!), and once
> when using the pathetically lousy SA cheesey plastic 2 cables-into-1
> control for their 5-speed.

That doesn't ring true.  A friction device can only relax control
cable tension, which ensures fuller engagement if anything.  As you
must be aware, disconnecting the control cable puts the SA hub fully
in top gear.  I find the Sturmey-Archer psychological warfare, in
which they have always placed the blame on the user, exceptionally
effective.  The more insulting their claims of user ineptitude, the
more the user cowers and takes the blame.  I think they are fully
aware of this and use it to best advantage.

> (Two 3-speed triggers together, instead of the plastic abomination,
> work crisply and reliably.)  Newer hubs by Sachs and Shimano have
> apparently eliminated this dead spot or neutral (freewheels both
> directions), and I like them too, especially the Sachs, but I still
> have confidence in the S-A, even though it does have a definite need
> for precise cable adjustment.

The dead spot is the essence of the SA hub and it requires it to
prevent having two gears locked against each other.  Low and second
cannot do this because this shift merely lifts one pair of over-running
pawls.  That's why there is no click when in low gear.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 4 Apr 2000 02:36:51 GMT

Jim Adney writes:

>>> Are you saying that the chain force causes the axle to bow thus
>>> leading these parts to no longer rotate about a common axis. Then the
>>> clutch walks slowly up the pinion pins until it escapes into the
>>> neutral between 3rd and 2nd?

>> That is exactly what I said

> No, but perhaps it is exactly what you meant. I was asking a
> question since I was not sure that I had understood your
> explanation.

I think you might have read something else into it, judging from your
belief that SA hub gears are safe.  I explained how automotive gears
disengage with axle sag to make it more clear.

>>> I am very familiar with these hubs having serviced hundreds if not
>>> thousands. I have never seen this problem, but I suspect it is real.
>>> It is probably not common simply because you are probably the only
>>> person who has pushed them this hard.

>>I suggest you disassemble a few and look at the cross and pins,

> Not necessary, I remember them well. They were amoung the most
> commonly replaced parts, worn as you described.

Well?  What's the point if you know they wear in this manner and the
wear marks indicate sloping contact and that this is not conducive to
stay in gear under load.

>>> For how long, or for how may strokes, do you need to push an AW hub to
>>> experience this? I'm inclined to take your word for it rather than try
>>> the experiment first hand.

>> Just stand up and sprint over a short hill.  I did not use toeclips
>> on these.  After the second dive, I decided on a more cautious
>> approach after having placed an order for a derailleur bicycle.

> "Short hill" is pretty subjective. Perhaps your short hill is longer
> than my short hill, but if this is all it takes I'm surprised that
> this does not happen to more people.

50 yards or how about a straight line sprint on the level, it makes
no difference.

> Are you aware that there is a "correct" way to adjust the bearings
> on the AW hub and that it may have an effect on your observation?
> With the left side cone loose, the sprocket side cone should be
> screwed down snug then backed off 1/4-1/2 turn, then the left side
> cone is adjusted in the usual fashion.  If this is not done
> correctly the neutral between 2 & 3 can become large.  If it is
> excessive, the clutch key can reach the end of the slot in the axle
> without fully engaging the pinion pins.  This would make the problem
> you describe worse, even if you left the indicator chain out.

The instructions are fairly straight forward and that the driver, when
the control cable is slack is completely engaged is also straight
forward.  I think it is also generally understood that there is a
freewheeling forward position between top ands middle gear.  Your
paragraph above comes right out of the SA intimidation in which it's
always the user's fault.  This gear will disengage from top gear with
brand new parts, adjusted to SA specifications and the control cable
slack.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 5 Apr 2000 18:42:22 GMT

Jim Adney writes:

>> I think you might have read something else into it, judging from your
>> belief that SA hub gears are safe.  I explained how automotive gears
>> disengage with axle sag to make it more clear.

> Actually, I never said anything that might imply that I disbelieved
> you. In fact, I have tremendous respect for your opinion, but in this
> case you seem to be the only person who has made this particular
> observation: that even new, properly adjusted clutches will inevitably
> walk off the ends of the pinion pins.

Well there's always a first time and the many people who discovered
this years ago did not have this forum to make their discoveries
known.  You will also find that the nature of stress distribution and
deflections in a bicycle wheel remained obscure for more than 100
years, partly due to the same problem.  For this reason I don't expect
anyone to take this on faith or someone's credentials, but rather on
observable characteristics of the mechanism and logic to explain what
you can observe.

>> Well?  What's the point if you know they wear in this manner and the
>> wear marks indicate sloping contact and that this is not conducive to
>> stay in gear under load.

> My point was that these are wear parts and that many of these hubs,
> without demonstrating the effect that you observe, wear to the point
> where they start to develop a problem with the same symptom but a
> different cause.

I don't understand what you mean by that.  They have sloping wear that
must be caused by sloping contact under load.  What is the "different
cause" and what is the effect?

>> 50 yards or how about a straight line sprint on the level, it makes
>> no difference.

> Thanks, that gives me a better idea.

>> The instructions are fairly straight forward and that the driver,
>> when the control cable is slack is completely engaged is also
>> straight forward.

> The driver doesn't move with gear changes; only the clutch moves. I'm
> not sure of this, but I think it is possible that adjusting the
> bearings in the wrong sequence might lead to the clutch key running
> out of travel in the slot in the axle before it is fully seated. If
> this happened then even running without a shift cable would not be a
> fair test. I'm just not sure how much extra room there is beyond the
> normal travel of the clutch key.

The slots are open ended and the "clutch" driving cross is free to
move until it bottoms against the face of the planet carrier, the
fullest engagement possible.  I've always found the instructions for
cable adjustment a bit misguided because this position its the most
important.  I've always adjusted the cable to just go slack in the
high gear position to insure that this tenuous position is given the
benefit of full engagement, the others far less being sensitive to
position.

>> Your paragraph above comes right out of the SA intimidation in
>> which it's always the user's fault.

> Funny, I was under the impression that I was the one who wrote that
> paragraph, and that you were the one who was trying to end this
> discussion by intimidation.

The point is that people who have had problems with SA hubs have been
generally told that it is user error.  I know the routine and it is
part of the hub gear fabric.  When something so consistently fails to
work properly, the design should be questioned.  The customer isn't
always right... but he usually prevails on the market.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 6 Apr 2000 20:40:52 GMT

Jim Adney writes:

>>> My point was that these are wear parts and that many of these hubs,
>>> without demonstrating the effect that you observe, wear to the point
>>> where they start to develop a problem with the same symptom but a
>>> different cause.

>> I don't understand what you mean by that.  They have sloping wear that
>> must be caused by sloping contact under load.  What is the "different
>> cause" and what is the effect?

> The cause is normal wear to these points due to shifting under load,
> which most cyclists do, at least occasionally.

Now you're making things up.  This hub cannot be shifted out of top
gear, the one in question, under load and when it does shift, it is a
disengagement, not a collision, as shifting under load might imply.
So called shifting UP under load is one thing this hub does well,
because it cannot move unless there is a torque interruption, as there
naturally is somewhere in the pedal rotation unless the rider is
working hard.  At that point the spring loaded driver cross snaps to
full engagement (if the cable is not adjusted to be tight in this
position).

Besides, my failures occurred with new parts, pins and driver.

> This naturally leads to a radiusing (the effect) of the sharp edges
> of both the pinion pins and the clutch. Eventually such a hub
> becomes unridable for even the most pedestrial of riders. In my
> experience, most riders ride for years without having any problems,
> finally start having problems, get the hub rebuilt, and the problems
> go away for another decade.

I believe you are so convinced of Sturmey Archer's excellence that you
will imagine almost any scenario to absolve them of their faulty
design.  Meanwhile you underscore that the shift mechanism "becomes
unridable for even the most pedestrial of riders".  You can't have it
both ways.  And "most riders ride for years without having any
problems" are exactly the most likely not to press hard enough to
notice, or if it does jump out of gear, not to care because it jumps
right back in leaving the rider with a little "clunk" in his stroke.
If, however, the rider were standing and exerting substantial torque,
he takes a dive over the bars.  I can assure you that I did that
twice, the first time I believed the assessment of the bike shop.
After the second time I was convinced that derailleurs have more than
a large gear selection going for them.

>>> The driver doesn't move with gear changes; only the clutch moves. I'm
>>> not sure of this, but I think it is possible that adjusting the
>>> bearings in the wrong sequence might lead to the clutch key running
>>> out of travel in the slot in the axle before it is fully seated. If
>>> this happened then even running without a shift cable would not be a
>>> fair test. I'm just not sure how much extra room there is beyond the
>>> normal travel of the clutch key.

>> The slots are open ended and the "clutch" driving cross is free to
>> move until it bottoms against the face of the planet carrier, the
>> fullest engagement possible.  I've always found the instructions for
>> cable adjustment a bit misguided because this position its the most
>> important.  I've always adjusted the cable to just go slack in the
>> high gear position to insure that this tenuous position is given the
>> benefit of full engagement, the others far less being sensitive to
>> position.

> The slot is NOT open ended.  It is finite in length.

The slot in the axle is limited but the driver in which the cross
moves is open ended.  I don't have a nomenclature manual at hand so I
can't dispute the SA names for the components.  That's why I describe
them with more words than their proper name (that I don't know).

> You may be right in that it is long enough toward the left side that
> that the left end of the slot can never get in the way, but if SA
> designed this part closely, expecting that it would always be
> correctly adjusted, then adjusting it incorrectly, and allowing the
> axle to shift to the right would both widen the neutral AND keep the
> clutch from fully engaging the pinion pins. There would be good
> mechanical engineering reasons not to make the slot any longer than
> it really needed to be, but it might be a "human engineering"
> mistake. To be honest I have never checked this out and it's not
> particularly convenient to do so right now, but I will keep it in
> mind the next time I have one of these apart.

Your creating a smoke screen.  The drive disengages under load and the
flanks of the driver cross and pins show wear over their full
engagement length.  Lateral adjustment of the internal parts is not
great enough to prevent full engagement of top gear.  All this aside,
you seem to gloss over that clearance and elasticity of the axle is
enough to cause a rejecting angular engagement, the design having no
countermeasures to prevent disengaging creep under load.  This is a
design flaw and has been so since the beginning.

>> The point is that people who have had problems with SA hubs have been
>> generally told that it is user error.  I know the routine and it is
>> part of the hub gear fabric.  When something so consistently fails to
>> work properly, the design should be questioned.  The customer isn't
>> always right... but he usually prevails on the market.

> At this point the AW hub is dead. I don't think it died because of any
> marketplace perception of a design flaw but rather due to lack of
> "coolness".

For many people this may be true but my initial contention remains,
that for a strong rider the hub is a dangerous mechanism that could
have been designed to not have the flaw that I discovered in using it.
I don't understand what moves people to defend this sort of design.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Sturmey Archer dangerous?
Date: 6 Apr 2000 20:57:40 GMT

Tho X Bui writes:

>> At this point the AW hub is dead. I don't think it died because of
>> any marketplace perception of a design flaw but rather due to lack
>> of "coolness".

> I think the main reason for its disappearance is its high cost of
> manufacturing. While it has its charm (enough for me to own a few),
> it's neither cheap to made, nor as easy to repair as an equivalent
> derailleur bike. I doubt that a typical shop today can do much
> repair work on one of these hub gears; the normal solution is
> generally a hub replacement/wheel rebuilt.

The hub is probably as easy to repair as any that I can think of.  My
children played with one as an assembly puzzle after I stopped using
it, completely dry and cleaned for them in their pre-teens years.

> While your experience may differ, I have yet to see one that can
> stand up to heavy use.  In my experience, the ones that last 10
> years usually see at most a few hundred miles per year of low
> intensity riding--nothing wrong with that, of course.

I used mine on 100 mile rides in the Santa Cruz CA mountains and soon
realized that this was not their forte.  I bought a derailleur
bicycle.

> For such riders, hub gears of that type are good solutions.  When I
> was regularly bike-commuting, I laced up a couple of hub bikes with
> cheap sew up rims.  They were good and cheap light weight
> all-weather bikes, until the mileage caught up to them.

The hub is not without merit.  It has the best labyrinth
non-contacting seals of any hub and uses double lead threads that
enable disassembly without special tools after long use that usually
makes threaded freewheels difficult to remove.  On the other hand, it
has an oil hole trough which the hub can be oiled, only to allow all
of it to run out down the spokes, there being no seal on the double
lead threads.  The seals are why these hubs still exist and run,
because even standing or riding in wet weather for years, no water or
dirt intrudes unless one undertakes submarine activities.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: 3-SPEED HUBS AND FRICTION SHIFTERS
Date: 30 Aug 2000 01:04:20 GMT

Sheldon Brown writes:

>>> I'm currently fixing a Raleigh Sports with a 3-speed hub. I'm
>>> intrigued by the hub's durability and have found many in good
>>> condition on trashed bikes in a local junkyard. I would like to try
>>> one of these 3-speed hubs on a mongrel *city bike* made from an old
>>> frame, parts, etc. Would any friction shifter (stem, thumb) work to
>>> move the 3-speed hub's inner gears sufficiently from one speed to
>>> the next?

>> No, this will not work because you have no feedback of how far you are
>> IN that gear.  When you disassemble and clean one of these hubs you
>> can see how the driver cross (clutch) is worn and how it can jump out
>> of top gear if under continuous load.  Top gear of the three is the
>> one that can pop into freewheeling forward

> Actually, it does work.  Top gear is easy, because in top gear the cable
> should be completely slack.  For low gear, the cable should be basically
> tight.  The slightly tricky one is the middle gear, but once you get
> used to the lever position, it's not a problem.

Well, yes and no.  The position between middle and top is the free
wheeling position and is the one that is worrisome.  Just the same, I
would not sprint in top gear, having taken a dive twice, long ago when
I still believed my SA 'expert' at the bike shop.  Once I realized the
credibility gap I analyzed what was going on and desisted from such
activities.  I got a derailleur bicycle.

> My Sturmey-Archer ASC three-speed fixed-gear hub came without the
> special trigger, and I've used it for several years with an old Sun
> Tour Barcon, no problems.

That doesn't count.  The ASC had no free wheeling position I'm told.
I've never seen one apart.

>> Don't sprint standing in top gear!

> That's good advice, for internal- and derailer-geared bikes.

Wait a minute, I've never seen a bike race in which the riders didn't
sprint standing on their derailleur bicycles, often in top gear.  They
can't all be wrong.  Besides, what is the hazard of doing so?  I've
done it for years without incident.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: 3-SPEED HUBS AND FRICTION SHIFTERS
Date: 30 Aug 2000 01:14:26 GMT

A Muzi writes:

> I know I risk your ire, but the clutch return spring is more than
> adequate to hold high gear securely if the cable's not kinked or too
> tight. I've commuted daily except for snow season on a '53 since
> 1971.  I'll admit that many 3-speeds I see in service have a kinked
> cable where it fell between the chain and chainring during a
> previous flat change.  I don't see that situation as a design fault.

You also didn't have one pop into neutral while sprinting on a new set
of planet pins and driver clutch, I did.  I also analyzed many hubs
that showed disengagement wear on the contact points that occurs
naturally when fretting away while engaged.  You can always back out
of your contention as so many SA diehards have when I say I was
standing pedaling hard, by saying "Ohhh it wasn't meant to be used
THAT way!"  Usually the claim is that the cable was too tight.

The cause for such disengagement is well known in gear box design
where the shift collar of a gear can disengage if the shaft is bent,
either by overload elastically or permanently from such an overload.
The engagement dogs (planet pins) oscillate axially once per
revolution because the shift collar (clutch 'cross) is rotating on a
skewed axis to the mating part.  If this is under high torque, the
parts push themselves apart.

It's the same story with broken pedal cranks, just because the shop
operator has never broken one, he is certain that the user must have
misused his to cause the failure.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: 3-SPEED HUBS AND FRICTION SHIFTERS
Date: 30 Aug 2000 15:11:19 GMT

Tim McNamara writes:

>> You also didn't have one pop into neutral while sprinting on a new
>> set of planet pins and driver clutch.  I did.  I also analyzed many
>> hubs that showed disengagement wear on the contact points that
>> occurs naturally when fretting away while engaged.

> <snip>

>> The cause for such disengagement is well known in gear box design
>> where the shift collar of a gear can disengage if the shaft is
>> bent, either by overload elastically or permanently from such an
>> overload.  The engagement dogs (planet pins) oscillate axially once
>> per revolution because the shift collar (clutch 'cross) is rotating
>> on a skewed axis to the mating part.  If this is under high torque,
>> the parts push themselves apart.

> Chris Juden of the CTC published a tech report on the SA Sprinter 7
> hub in which he had the same problem Jobst has described with the AW
> hubs- that the points of the driver clutch rounded off under load
> and thus the hub wouldn't stay in gear.  SA has since redesigned the
> hub with a "ball locking mechanism" (I am parroting this, I have no
> idea what it actually means; I've looked at the available technical
> manuals on the Web and can't make heads nor tails out of them) that
> has eliminated the problem in the 7 speed hubs.

> Perhaps Sheldon could comment on the new vs old arrangement, since
> he has these hubs for sale on his site and has some experience with
> them.

I don't see a stronger detent as a reliable solution.  That is much
like using a stronger return spring, it just moves the problem a
little farther out.  A retaining taper on the driver clutch and
engagement face would be a positive retention.  This would not be
difficult to incorporate, but I am not in any position to advise SA
who seem to be immune to change, possibly under the motto that change
would admit error.

> I suspect that for most users (read:  potterers, commuters and
> shoppers) of the old SA AW hub, the problems are minimal because they
> (1) don't sprint and (2) don't sprint standing up.  ;-)  I've always
> wondered about touring in England, with its many 1-in-4 and 1-in-3
> hills, as to whether grinding up such gradients resulted in this sort
> of failure.  Jobst, being a strong high performance rider probably
> climbing the Sierras on an AW in 3rd gear, was able to find that weak
> spot effectively.

Ah, but the problem does not exist in the low gear.  It is only there
for the top gear of the 3&5-speed.  This is one of the reasons they
can get away with this.  Most folks only use top gear with a tailwind
or when idling along.  Hence, the "You're doing it wrong" excuse can
generally be applied.  Hell, it's worked for about 100 years and
hardly anyone is the wiser.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: 3-SPEED HUBS AND FRICTION SHIFTERS
Date: 31 Aug 2000 16:33:30 GMT

Frank Krygowki writes:

>>>> The cause for such disengagement is well known in gear box design
>>>> where the shift collar of a gear can disengage if the shaft is
>>>> bent, either by overload elastically or permanently from such an
>>>> overload.  The engagement dogs (planet pins) oscillate axially
>>>> once per revolution because the shift collar (clutch 'cross) is
>>>> rotating on a skewed axis to the mating part.  If this is under
>>>> high torque, the parts push themselves apart.

>> I don't see a stronger detent as a reliable solution.  That is much
>> like using a stronger return spring, it just moves the problem a
>> little farther out.  A retaining taper on the driver clutch and
>> engagement face would be a positive retention.  This would not be
>> difficult to incorporate, but I am not in any position to advise SA
>> who seem to be immune to change, possibly under the motto that
>> change would admit error.

> Before I take apart a SA hub to see what you're describing, Jobst,
> let me ask: is this the kind of thing that can be cured by
> re-contouring parts with a Dremel tool, or something similar?

No, this would need to be a machining process and on parts that had
not been heat treated.  The parts in a three speed hub in question are
case hardened and cannot be modified readily.

> I'm thinking by analogy to some motorcycle transmissions many years
> back.  I recall reading a road test that claimed the previous model
> had a propensity to jump out of gear, and the engineers cured it by
> putting a slight taper (1 degree or so) on the dog-gear interface
> surfaces, so torque on the gear tended to keep the gear engaged with
> the dog.

Well I'm sure this was done by grinding and that it must be done with
an indexing mechanism to assure that all surfaces are in proper
circular pitch.  As I said, it is an old problem that has been
addressed in other gear boxes.

> I've got one of these hubs that I was hoping to use on a folding bike,
> but you've gotten me nervous about it.  Is it easily fixable in a home
> shop?

Just don't stand in top gear and you'll be OK.  My riding includes
sprinting in top gear pretty regularly.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: How to replace AW pinions?
Date: 22 Jan 2001 18:31:10 GMT

Tomek Liniecki writes:

> My thirty years old AW hub started to self switch gear (from 1 to 2) while
> riding upon rough, dirt/crushed rock roads.  The cable seems to be OK.  The
> SA manual advises to replace pinions.  I'm about to place a mail order for
> the new ones, but I'm not really sure which ones should I replace, since
> there are two sets of pinions in the hub.  Should I replace both?
> I was told, that the modern AW hubs are significantly less durable than the
> older ones.  Taking this into account, does it make sense to repair the old
> hub with the new, lower quality parts?

I'm no up on SA nomenclature but in a planetary gear there is a sun
gear and a ring gear with planets that connect them.  None of these
have any effect on shifting an SA hub, they being permanently engaged
with one another.  What you most likely have is a poorly adjusted or
kinked (springy) shift cable.  What wears out in these hubs are the
planet pins and the driver "clutch", the cross that engages the gear
positions.  Consider getting these anyway.  The ones in there are most
likely worn.

The correct gear adjustment is, in my estimation, not that which is in
the manual, the danger of popping out of top gear being the most
serious problem.  Popping out of top is the only condition that puts
the hub into free wheeling forward.  In contrast, up-shifting from low
to second can be done under power safely and without ill effect.  In
adjusting the shift cable, make sure that the cable is nearly slack in
top gear.  That is to say, that the shift mechanism is as far inward
as it can travel with no restraint from the cable.  If this is not the
case, the driver clutch will not be fully engaged in top gear and will
more easily pop out under load.  This is the action that damages the
clutch, the planet pins, and of course the rider, if he happens to be
standing when the mechanism fails.  He goes over the bars.  Most SA
users don't sprint in top gear so the popping out of gear merely
damages the driver and the user is admonished that he adjusted the
unit incorrectly.  Nevertheless, it is a basic flaw in the design.

> Maybe it would be more reasonable to install a complete new hub, and
> than dispose it after several kkms?

Nothing else in the hub wears significantly so you are just as well
off with new pins and clutch assuming the unit has not been run dry
for many miles.

> Is there any technically rational and affordable way to upgrade the
> three speed (110 mm dropouts) bike to four speeds?

Both the 4 and 5-speed hubs are exactly the same size and use the
same housing.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: How to replace AW PAWLS?
Date: 22 Jan 2001 19:11:23 GMT

Tomek Liniecki writes:

> I had on my mind pawls, not pinions. The SA manual advises, of
> course, to replace "gear ring pawls". I know quite well which ones
> are those, so my question should have rather be: How to evaluate,
> whether they're indeed worn?

The pawls have little chance of being worn, the detents being broader
and deeper than any freewheel on the market.  Besides, their failure
would only temporarily put the hub in a different gear until they
caught the next detent.  Either that or they would not work at all,
even for a short time.  Is this a time dependent effect or does it
occur immediately upon force on the pedal?

Since SA hubs are so easy to disassemble and reassemble, you ought to
look into it and see what's in there.  You need only remove the right
side cup of the hub for everything to come out (after taking of the
axle nuts and cones.  I think you'll find every thin in fairly good
shape except the cracked off ends of the planet pins and the wear
marks they made in the clutch (the cross that drives the gears).

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: left side S-A cup removal?
Date: 22 Jan 2001 19:24:43 GMT

Brad Upton writes:

> Anyone have experience removing the left side (non drive) ball cup from
> an older ('50s) Sturmey-Archer AW hub shell?  It looks to be threaded
> into the shell in the same fashion as the right hand ball ring.
> However, I don't know if the left side cup (is it called a ball ring
> also) is right or left hand threaded.  Also, Wonder how one could "grab"
> it to spin it off.  ( I bet it's on there pretty tight!)

Since both cups transfer torque to the wheel, they must both tighten
in the forward rotating direction, making the left side a left hand
thread.  So that these do not get too tight, they have a double lead
thread that makes removal reasonable.  I've never had trouble getting
them off even without special tools.  I just used a screw driver shaft
(one with the handle broken off) and drive it loose with a hammer.  It
takes a sharp blow, so just hitting a screwdriver may not work, the
handle absorbing the necessary shock to loosen the thread of the right
cup.  The left cup has wrench flats that can be used in a vise or a
large crescent wrench.  The non-threaded version presses out, having a
coarse knurl.  I've never tried that but it looks relatively easy with
a press.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jobst.brandt@stanfordalumni.org
Subject: Re: Raleigh 3 speed into fixed
Newsgroups: rec.bicycles.tech
Date: Wed, 28 Nov 2001 01:51:05 GMT

Keven Ruf writes:

> However, it seems the rear dropouts are too narrow to accept a normal
> axle.  I don't mean the rear spacing, I mean the space in the drop
> out where the axle slides in.  Is this true?  If so, how can I get a
> standard wheel in there so I can switch out my 3-speed wheel with a
> fixed wheel?

Unless the dropout is bent (hammered shut) it should fit.  If not, use
a rat tailed file of about axle cross section and open it up,

> Also, the hub leaks oil terribly, is there something I can do about
> that?  The pawls are loud if I don't put oil in there, but it all
> comes out on the floor over night when I do.

That's one of the reason I find the SA hubs designed to go out of
business, like English motorcycles that always dripped and had other
ills that were known but never fixed.  The hub is not oil tight, both
in the threaded side and the other one that was threaded in earlier
models, but even this side leaks.

Beyond that, the drive cross (clutch) drives four pins, neither of the
two elements having a reentrant (sloped) contact surface, and under
continuous pedal force, disengage, leaving the gears in free rotation
forward.  This is a well understood concept in transmissions where
loads deflect the axle so that a drive4r spline and the driven spline
are running off axis.  This invariably causes them to separate if
there is no other holding force.  That force on the SA hub is a weak
spring at best.  Standing sprinters going over the bars in top gear is
a known mode of failure, always attributed to improper cable
adjustment, an excuse that most accept.

Even if misadjustment were the cause, although it isn't, the ability
to so easily misadjust the cable and cause serious injury, is a
design flaw.  That's the way product safety works.

Jobst Brandt    <jobst.brandt@stanfordalumni.org>

From: jobst.brandt@stanfordalumni.org
Subject: Re: Sturmey-Archer lubrication
Newsgroups: rec.bicycles.tech
Message-ID: <lIN38.12463$TI3.122382@typhoon.sonic.net>
Date: Thu, 24 Jan 2002 06:20:01 GMT

Andrew Muzi writes:

>>> One thing I can say for sure.  Back in the early 1970's I used
>>> some 10-weight non-detergent motor oil on a Sturmey 3-speed.  It
>>> ran like a deer, but the oil ran right out of it -- kind of like a
>>> Harley of the same vintage.  I'd have to park the bike on a pad of
>>> newspapers.  I eventually stopped that and went back to a can of
>>> genuine Sturmey Archer brand oil, and I didn't leak a drop after
>>> that.

>> As I said at the beginning of this thread, Sturmey-Archer hubs are
>> not sealed in any way and are not designed to run their gears in an
>> oil bath, although that might have been a good idea.  Threads
>> (unless pipe threads) do not retain fluids.

> Jobst, I know you have some antipathy to SA hubs, but really, the
> hubshell will hold two tablespoons of oil well below the dustshell
> level.  A bit does seep out the ends but that's just the oil flung
> about by the mechanism in operation that happens to land at the
> ends.  It's not much and Sturmey's "a drop per fortnight in daily
> use" is about right.  Hardly a gusher like classic Harleys or MGBs.

It's not an antipathy but fact.  The hub is a tube with threads on at
least the right side although older ones have them on both sides.  The
threads are cut into the cylindrical housing, so they are lower than
the 'shell' diameter.  At least the right hand cup has a double lead
thread so it doesn't get "welded" on like single lead freewheels of
old.

Neither the threads nor the steel cups make an oil seal that does not
all bulk oil in the housing to weep out and run down the spokes.  That
is why most well maintained hubs have a halo of oily dust on both ends
unless the user is on it with a rag regularly.  No oil comes out of
the labyrinth seals in the end caps.  It is seepage at the threads.

The two most notable features of these hubs are the excellent
labyrinth seals on the cones and sprocket carrier,and the double lead
threads on the right hand cap that allows the hub to be opened easily
with simple tools.  The 'sealed' hubs, more than 40 years old, can be
opened to reveal clean working parts, even if worn for lack of oil.

Since SA popularized the three speed, much has been improved but SA
was not willing to change anything.  All failures of popping out of
gear were thrown in the user's lap when in fact there is an obvious
design flaw known to anyone in the motorized gear business.  Even if
it weren't a design flaw, the customer is falling on his snout, so the
design needed change, change that was not forthcoming.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA

From: jobst.brandt@stanfordalumni.org
Subject: Re: Sturmey-Archer lubrication
Newsgroups: rec.bicycles.tech
Message-ID: <1cJ38.12379$TI3.121247@typhoon.sonic.net>
Date: Thu, 24 Jan 2002 01:12:29 GMT

Andrew Muzi writes:

>> One thing I can say for sure.  Back in the early 1970's I used some
>> 10-weight non-detergent motor oil on a Sturmey 3-speed.  It ran
>> like a deer, but the oil ran right out of it -- kind of like a
>> Harley of the same vintage.  I'd have to park the bike on a pad of
>> newspapers.  I eventually stopped that and went back to a can of
>> genuine Sturmey Archer brand oil, and I didn't leak a drop after
>> that.

> As I often ask, "What are we measuring?". Here, is it the material
> or the volume of it?

> Speaking of measurement, I strongly doubt that exactly two
> tablespoons (SA's recommendation to fill a dry hub) of anything will
> be enough to run out then ends unless the bike is laid flat and even
> then not much.

It makes no difference.  Anything more than an oil film on the parts
will run out.  The ends of the hub are essentially open to seepage,
and that goes for the later ones that had the left endpiece knurled
and pressed in.  These are not fluid tight covers although a small
O-ring could have done the job.  It was partly this attitude of SA
that made their hubs vanish from the market.  Like British motorcycles
that all leaked and made no technical progress, this is another one of
those "who knows more about this than we who made thousands of them".
That is a primary reason why there are no new British motorcycles.

> I believe you saw oil running out but how much was in?

I'm sure it was all of it came out except what stayed behind as an oil
film on the innards.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA

From: jobst.brandt@stanfordalumni.org
Subject: Re: Sturmey-Archer lubrication
Newsgroups: rec.bicycles.tech
Message-ID: <MON38.12465$TI3.122382@typhoon.sonic.net>
Date: Thu, 24 Jan 2002 06:26:52 GMT

Andrew Muzi writes:

>>> I believe you saw oil running out but how much was in?

>> I'm sure it was all of it that came out except what stayed behind
>> as an oil film on the innards.

> Without pursuing that, what of his report above that motor oil
> flowed like wine yet SA oil stayed inside??  That was the part I
> found incredible.  Any comments on that topic?

You've been around wreck.bike long enough to know that writers often
fail to get cause and effect linked and also often miss what others
might find obvious.  SA oil is not magic even though it is made for
bicycles.  The only thing that changes leak rate is viscosity, and SA
oil has a lower viscosity than conventional motor oil.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA

From: jobst.brandt@stanfordalumni.org
Subject: Re: Sturmey-Archer lubrication
Newsgroups: rec.bicycles.tech
Message-ID: <AZN38.12474$TI3.122382@typhoon.sonic.net>
Date: Thu, 24 Jan 2002 06:38:24 GMT

Frau Sheldon Brown writes:

> Actually, they converted to "lifetime" grease lubrication several
> years ago, and stopped providing the oil port.

I guess that was hard for them at SA to swallow... and explain.

> The engineering efforts of the British bicycle industry were devoted
> to making the products better up until sometime around the late
> '50s/early '60s.  After that, they turned their energies into
> finding ways to make stuff cheaper.

> All I know about British motorcycles is Richard Thompson's "Vincent
> Black Lightning, 1952" and but it is my impression that a similar
> arc occurred in that industry.

It's interesting that you should mention "The Vincent" because I spent
much time ministering to a black lightning.  It had an elegant engine
coupled to a disastrous clutch and gearbox.  The clutch was a tilting
dual leading shoe servo device, actuated by a single plate clutch.  At
high RPM, the time when one should shift to the next gear, the clutch
remained locked and blew out the next gear in the shift process.

Fortunately the gearbox part of the engine housing was so spacious,
for some reason, that the broken parts could fall to the bottom
without damaging other gears.  This was my last endeavor in the
motorized two wheeler.  By then I had already switched to a Cinelli.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA

From: jobst.brandt@stanfordalumni.org
Subject: Re: Sturmey-Archer Hub Dissasembly
Newsgroups: rec.bicycles.tech
Message-ID: <USTA9.48450$Ik.1231387@typhoon.sonic.net>
Date: Thu, 14 Nov 2002 20:40:20 GMT

Andrew Muzi writes:

> If the control cable is removed, and the clutch return spring slides
> the clutch all the way to the left in high gear, what would cause
> the clutch to move to the right and disengage its edges from the
> gear (pinion) pins?

Good point.

Two features that cause disengagement occur only under continuous hard
torque.  The four pinion pins, that fit loosely in the planet cage,
cant slightly under pressure from the clutch (driver cross), to an
off-perpendicular angle so that the bearing surface with the clutch
slopes toward disengagement.  Meanwhile, chain loads on the hub cause
the axle to bend slightly so that the active parts (planet cage and
clutch) displaced along the axle, rotate about separate skewed axes to
each other.  During rotation, the four bearing faces of the clutch
each experience reciprocating motion of the planet pins that, through
their slant, generate disengaging creep.  The engagement spring is no
match for these forces.

In addition to the disengagement forces, caused by pin skew and axle
flex, the clutch faces develop indentations from the slanted pins that
enhance disengagement.  This is clearly visible on any used clutch.

This problem could have been resolved by putting a slight inward taper
to the ends of the planet pins and a similar matching slant on the
bearing faces of the clutch, giving their engagement a preferential
retaining force instead of the opposite.  Most motorcycle gear boxes
use such features, especially in older non-synchronized sliding gear
boxes... the classic clunk of BMW boxes for instance.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA