chagzuki

Not sure if sliding key is the correct term, but I'm under the impression that the new 5 speeds with wider ratios have essentially reverted to the older, pre-ball-locking design.
Any thoughts as to the merits of each and why Sturmey would do this? And is there something fundamentally different about the new hubs that demonstrates an evolution / improvement of the older design?

I have noticed with the ball-locking system that it's possible for balls to ping in and out of the axle recess as the sun pinion rotates around it, and there's some amount of friction in this process. With copious amounts of grease the ball tends to stick in the axle recess preventing this from happening. But many hubs have never been dismantled to the stage where the sun pinions have been removed and axle re-greased. Obviously no such problem occurs with a sliding key system.

Airburst

Firstly, I can confirm that they have - the X-RD5 I bought last spring has the sliding key system not the ball-locking one.

As for why, if my memory serves, the ball-locking mechanism had an unfortunate tendency to allow the sun pinions to slip under hard pedalling, leading to issues I'd imagine these issues would either be the high gears or the low gears suddenly slipping into the direct drive gear, but I don't remember for sure.

Dan Burkhart 

Yup. Sliding key is the correct term. As to why they abandoned the ball lock, I don't really know. I think it works just fine, has for me for quite a number of years.
The new version also has the sun and planetaries integrated in a complete non serviceable sub assembly. The change to the sliding key might have something to do with that.
One benefit to this design is you no longer have to align the timing marks when you re-assemble. The downside is, if you break a sun gear, you buy a whole sub assembly.

chagzuki

I've just had a look at the (w) internals, I see the clutch is completely redesigned with 6 contact points. The springs are different.
Do you know if the shells are still interchangeable?

Dan Burkhart 

As far as I know, the shells are interchangeable. I am in a position to test it out, so I'll let you know, probably next week.
The clutch is redesigned as the planet pins do not protrude from the sub assembly, so the old style clutch that engaged the pins would not work.

chagzuki

Thanks. If you can experiment any further I'd be really curious to know which bits are interchangeable, particularly the driver assembly, since the new 'N' hubs have the shimano spline fit that allows greater chainline adjustment.

jrecoi

The other benefit to having the sun gears and the planetaries as one unit is that other ratios can be implemented without disturbing the rest of the hub design. Imagine an SRF5(C) with a tight gears, or an SRF5(M) with medium ratios.

Dan Burkhart 

I don't know how much demand there would be for that, but it could have been just as easily done with the ball lock type. The main benefit as I see it is ease of disassembly and assembly, and the elimination of the need to align the timing marks.

Dan Burkhart 

This is why I test things on the bench instead of just accepting what I read on the internet.
I wish I had my camera here, it's in the car which my wife took to work, but the shells are not interchangeable.
The old style internals will fit into the wide range shell, but the planet cage pawls stop short of engaging the ratchets.
The wide range internals jam up against the inside wall of the non drive side of the old style shell.
The inside measurement of the old style shell from the ball ring seat to the interior wall is 58mm, the new style is 60mm. The ratchets on the old style are about twice as wide as they are on the new one. If the new shell had the wide ratchets, I'm pretty sure the old internals would work in it.
I'll dig in a little farther later to see if the drivers are interchangeable.

Dan Burkhart 

The drivers appear to be identical. The outer dimensions are the same, as are the internal splines. The clutches swap back and forth between the drivers no problem.
The clutches of course are radically different, and not interchangeable, but we already knew that.

chagzuki

That sounds as though the new 5 speed shells are interchangeable with the 3 speed. The old 5 speed internals couldn't be used in a 3 speed shell for the same reason, the ratchets weren't long enough.

tcs

A little history:

Fichtel&Sachs introduced a 4-speed hub in 1912. Production was suspended during the Great War and the hub was never re-introduced.

Sturmey patented their first 5-speed in 1922. They didn't market it.

Sturmey debuted their first 4-speed in 1938, the FW. It was actually a 5-speed with a suppressed fifth gear, and garage tinkerers developed ways to modify them to liberate the highest ratio.

Sturmey finally decided the world was ready for a bicycle with five gears and debuted their first production 5-speed, the S5, in 1966. It had ratios of 2:3, 15:19, 1:1, 19:15 and 3:2. It was shifted by a pull chain on the RH side and a bell crank on the LH side, with two shift cables/shift levers.

Sturmey patented an inside-the-dropouts shift select in 1970, but did not market any hubs with this feature.

Sturmey's S5/1 (1977) and S5/2 (1983) were further developments (improvements?) of the basic S5 design.

Sachs brought out their H5100 5-speed in 1987, a dual pull chain design with identical ratios to the Sturmey.

Sturmey brought out their 5-StAr hub in 1991, their first NIG 5-speed. It was withdrawn from production by early 1992, and replaced by the Sprinter 5, a sliding key design with a single shift cable.

Sachs countered in 1992 with their H5213, also a single shift cable design with a bell crank operating a clever rod-in-tube actuator which maintained the strong sun gear lock up of the S5 design.

Shimano introduced their Nexus 7 in 1992, It had an aluminum hub shell, relatively even spaced gears and inside-the-dropouts shift selection, and it obsoleted both the Sturmey and Sachs 5-speeds overnight.

SRAM took over Sachs and in 1998 introduced their Spectro P5, a further development of the H5213 featuring wide range gears of 19:30, 57:73, 1:1, 73:57 and 30:19, matching the over-all range of the Shimano Nexus 7.

Sturmey's ball locking Summit 5 came out in 1999. It was to little and too late, and they went bankrupt just over a year later.

SunRace Sturmey-Archer returned the ball locking 5-speed design to production around 2001.

SunRace Sturmey-Archer came out with their 5(W), hubs in 2009. They returned to the sliding key design and matched the SRAM Spectro 5 with wider ratios of 5:8, 3:4, 1:1, 4:3 and 8:5.

SRAM discontinued all models of the Spectro 5 in 2011/2, and closed their 110 year old IGH factory in Schweinfurt, Germany.

Shimano introduced their SG-5R35 narrow range 5-speed in early 2013.

Dan Burkhart 

Thanks for the history lesson tcs. That's interesting stuff, to me at least.

Continuity

Wow - nice capsule IGH history there, tcs.

I always wonder just how did S-A end up making such *moronic* marketing decisions as they did? It's patently obvious that they had a competent, workmanlike engineering department, but the other divisions were total morons, apparently.

Just what reason could they have been thinking of to have had fully-working, really good world-first multi-speed hub designs, and then just sit on them for years? I mean - just WTF could they have been thinking of?

I've read some fascinating histories from people who used to work at Raleigh/SA over the years, and it sounded like a really weird place to work. It was *so* classist and divisory, even for the UK of that era - they had like 4-5 different 'grades' of break room, and woe betide you should you find yourself in the 'wrong' one. LOL

The stories also included that of some part of the paint dept. which had giant vats of cyanide in them. (and the workers who worked with them were provided with *no* protection and/or proper information on the hazards) One day a guy from another dept. walked in with his teacup, filled it from the tank and went to his car in the car park and drank it. Another occasion, a 17-year old trainee also tried to off himself using the tank solution, and only after this apparently did they put locks on the tanks.

Matariki

Sounds like a good read. Can you post the citations, please. Thanks

Continuity

Sure - you can find links to the stories I read here - https://thespace.org/items/e0000kjc

Dan Burkhart 

turns out, the dimensions of the wide range 5 speed are the same as the classic 3 speed. I measure them both at 57mm from the ball ring seat to the tip of the assembly. The ball lock 5 speed measures 46mm.
This shot shows the 3 speed on the left, ball lock centre, and wide range on the right.



The 3 speed fits and works in the wide range shell, but some machine work would be required to make the 5 speed wide range work in the AW shell. See that lip? It would have to go.



Compare to the new shell.



This shot shows why that lip doesn't interfere with the 3 speed internals.

Dan Burkhart 

I didn't try this. and the hubs are all back together, so I'm not going to try it now, but I think if you were to swap the ball ring from the ball lock internals onto the wide range, it just might work in the old shell.
Anyway, I gave the ball lock 5 speed an ATF dunk this afternoon, and took it on a grocery run.
Silent and smooth baby.

chagzuki

Very informative and a very handy resource, thankyou.

Interesting also to hear that Shimano have just brought out a new five speed.

Dan Burkhart 

Yes, inside the dropout shifting, that's a plus, but the ratio range is somewhat less. Don't remember the number off the top of my head, but I think it's about 225ish, which was about where the ball lock 5 speed was.

chagzuki

The ratios looked good to me, a huge jump from first to second and then all tightly spaced for cruising.