Sturmey Archer S2C Duomatic; Peek inside a friction box

For discussions about bikes and equipment.
Post Reply
Brucey
Posts: 43116
Joined: 4 Jan 2012, 6:25pm

Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by Brucey »

'Friction Box'.... that is what the unbelievers used to call all hub gears.... I didn't really subscribe to it myself, and delighted in going as quickly as possible on my various hub gears....

Well, now I've found a real one. Alan kindly sent me an S2C Duomatic to play with that he had rejected. As soon as I got it out of the box I could see why; with the sprocket fitted, turning it forwards took both strength and a masochistic streak as the teeth dug in, the force was that great.

So I set to as promised and started to strip it down. My first action was to question the parentage of the designer for his choice of 22.6mm AF locknuts on the left side of the hub, and no tab washer. This makes adjustment of the hub rather tricky as you need two of the piggin' spanners to adjust the hub bearing in situ.

However having undone the locknuts the brake plate comes off, then the centre pulls out in one piece from the right side, brake shoes and all.

SA S2C duomatic 1v2.jpg


The SA hub has a screw-in right side ballrace, but this can stay put unless it is worn; there is no need to disturb it to remove the centre.

The right side cone and locknut also don't have a tab washer which may lead people to suppose that this is where you should adjust the hub bearing. This would be a mistake. There is typically about 0.5mm 'float' in the centre parts, so the traditional SA scheme of snugging the right side (with the hub dead loose on the left), backing off half a turn, then locking these and leaving them alone for ever is the right one here, too. Unless you are dismantling the entire centre, the only gain to be had by removing the right side cone is an inspection of the bearings; it affords no further view inside and is not a good place to add lubricant.

The hub bearings adjust on the left side; as usual the locknut will preload the bearing if care is not taken with the adjustment. Unless you happen to have two thin 22.6mm spanners the bearing adjustment is probably best done with the axle clamped in a bench vice.

As supplied, the hub cannot be fitted with tab washers to facilitate cone adjustment and locking, because the axle flats are (amazingly) about 3mm too short on both sides. I plan to grind the flats a bit longer and fit tab washers. Fortunately the axle threading is the traditional 13/32" size long favoured by SA for three-speed hubs etc so AW tab washers and locknuts can be fitted OK. This may mean I can even avoid having to make an extra-thin 22.6mm AF cone spanner... we'll see...

SA S2C duomatic 2v2.jpg


Inside there is a fair amount of grease and everything looks quite substantial. In the pictures you can see some of the hub's inner workings. The planet ring is in constant mesh and turns 138% of the sprocket speed. It is fitted with eight detents; four shallow 'low gear' detents and four deeper 'high gear' detents. In the low gear, the high gear pawls are defeated by the shallow detents and drive is transferred by the low gear pawls. In the high gear, the high gear pawls are free to engage with the hub shell and transmit drive; the low gear pawls are overrun backwards and will emit the classic sturmey-archer 'tick-tick' sound.

The planet cage is a bolt-up affair; I didn't disturb it but I would assume that the bolts are loctited in for the duration. I'm not usually a fan of this arrangement, but this looked OK to me, pretty robust. The gear meshing seemed snug and confidence inspiring, but it may have just been an illusion created by the grease they use which is rather sticky.

A small turn backwards on the pedals will engage the coaster brake via a roller/ramp mechanism. This is helped to disengage via a drag spring labelled 'DS2' in the pictures. The brake shoes are held together and thus away from the hub shell by a spring clip 'DS3'. This allows the hub shell at least to turn very freely when coasting. A short backpedal will also change gear, although it takes a slightly different length backpedal going from high to low than low to high. This means that whenever you brake, you will also change gear.

It turns out that when stationary with the hub in question, it should change from low to high, but not high to low. To make the high to low shift, the hub shell needs to be turning forwards (as well as the pedals backwards) i.e. it seems that you need to be moving.This is the exact reverse of what is desired; if pedalling in low gear, working the brake shifts into high; if you then stop without a second burst of braking you will stop in high gear, not low gear, and you cannot change again until you are moving forwards once more. It may be that once the brake shoes have worn a little, the length of backpedal possible when stationary (which is limited by the brake coming on) will lengthen slightly and allow a stationary high-to-low shift. Until then it may be rather annoying. [edit; after more fiddling, it now does the high-low-shift when stationary, but not every time. I am somewhat optimistic that it will settle down in this respect]

[With the standard S2 hub, this is not going to happen in the same way; there is no reason to backpedal whilst slowing down, and a gear change (or several by pedalling backwards further) can be made whilst stationary. The 'click' of the high gear pawls is different depending on which gear is engaging, so it is quite possible that you will be able to select the 'gear by ear' as it were.]

I stripped down the centre to its main parts, then removed and reworked the three drag springs. All were too tight and needed various forms of reshaping. I guess these should ease with use anyway, but they were stupidly, way-too-tight. The drag force seemed worse in the high gear, but this may have been illusory.[edit; I have since learned that this high drag has been 'explained' in some cases as 'bearings shipped adjusted tighter than normal to avoid damage in transit'. What a crock...]

DS1 effectively preloads and therefore allows the planet ring not to chatter when it isn't being used for drive. This helps to keep the chosen gear in when pedalling forwards. I'm not sure if it has a great effect on the brake operation or not. I made it about half as draggy.

DS2 is the brake drag spring, and needs a slightly higher force I believe. It is stiffer and thicker than DS1. I again made it about half the force.

DS3 is the brake shoe retainer. I'm not even sure it is really necessary, but the axial float in the shoes seems quite large, so maybe they can go skew without this spring and cause the hub to jam. This I'm not sure of, but what I do know is that they could have made it so that the shoes are held against one another but away from the both the inner and outer part, but they didn't. They made it so that the shoes are held tight by DS3 against the roller assy, and this creates yet further drag whenever the hub is driven. I again eased the tension in this spring.

I put everything back together again with a load more semi-fluid grease, and was pleased to note that I can now turn the hub drive without my fingers bleeding from the effort. Only time will tell if this is a good fix or not. There is a good spot for adding a lube port in time, where the Al shell is the only thing that needs drilling through. I noted that the planet gears are not easy to lubricate per se, but should water get past the right side ring bearing, it will head straight in there.

[edit; the axle on my hub is 160mm long. I reckon this is long enough for a 130mm frame, provided you put thin (say1.5mm) washers on the outside and use the NTWs on the inside of the dropouts. My hub has a serial ID of JFA8, and the internals are of rather different design to those you will see in the McCraw blog (link in a following post). I don't know if these internals are likely to suffer problems similar to those seen by others. I guess only time will tell...]

Comparisons with the SRAM A2 Automatix coaster brake are inevitable. Despite its steel shell, the SRAM is lighter (by 7 Oz, ~200g), less draggy, and easier to use. Whilst I'm a sceptic for automatic gear changing in general (especially electronic, and it'll be here soon, trust me....) with two gears I can live with it, it really is a no-brainer for an urban bike.

By comparison the SA S2C Duomatic is heavier, more draggy and fussy in use, especially in this coaster brake form. It does have a nice aluminium shell which can be had in pretty colours, and it does look a little more robust inside (18 balls in the main bearings rather than 12, bigger pawls, very chunky anti-turn washers etc). Also, it will hold the gear you are in regardless of road speed if you want it to.

Sealing and lubricant retention looks about the same for both hubs, they both have about the same gear ratios, and both came with a 19T sprocket (suggesting 36x19 gearing on a 700c wheel maybe). Long term reliability of either hub is not exactly proven; thus far the S2C has had a fair amount of bad press but maybe they have sold a load more of them, had teething troubles etc..... the centre on both can be replaced easily, and in theory the S2C is more repairable because of it's screw-in RHS ball race.

Overall however the SRAM does look to be the easier to live with, even assuming that you can get the S2C to work without the massive drag penalty that my example was saddled with out of the box.

[edit the tooth counts are as follows; 17T planets, 21T sun, 57T ring gear. This gives a ratio of 1.3684.]

Thus far I've not ridden the S2C, only bench tested it, but it has been quite educational. As with the notes re. the SRAM A2 hub I prepared last week, the usual disclaimers apply; none of this is endorsed by SA, you may void your warranty, you carry out such at your own risk etc. However I hope folk find it interesting/useful, anyway.

cheers
Last edited by Brucey on 31 Mar 2020, 2:56am, edited 4 times in total.
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
catsnapper
Posts: 62
Joined: 19 Sep 2010, 8:29am

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by catsnapper »

Thanks Bruce, good comprehensive first look at the S2C. Now I know why I wasn't too keen on them:)

Alan
User avatar
hubgearfreak
Posts: 8210
Joined: 7 Jan 2007, 4:14pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by hubgearfreak »

oddly, when i put S2C into the forum's search function this thread (nor any others) came up. i'm glad to have found the thread though. thanks Brucey for doing a strip down. 8) i'll follow it at the weekend :?
User avatar
[XAP]Bob
Posts: 18709
Joined: 26 Sep 2008, 4:12pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by [XAP]Bob »

hubgearfreak wrote:oddly, when i put S2C into the forum's search function this thread (nor any others) came up.

Probably a minimum of 4 letter words in the search function.

Just try it on a computing forum:
cpu, psu, vga, gfx, hmm...
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
Brucey
Posts: 43116
Joined: 4 Jan 2012, 6:25pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by Brucey »

Last edited by Brucey on 28 Jul 2012, 1:20pm, edited 1 time in total.
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
pioneer
Posts: 1699
Joined: 13 Feb 2007, 10:39am

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by pioneer »

Thanks Brucey, very interesting. An S2 hub is on my shopping list for when I actually get around to having a bike built to my very own specs'. (i.e. one day in the future, probably many years hence!). I have got a Fitchell and Sachs Duomatic that I had built into a 20" wheel for the folder. This hub had lain unused and unloved for years in a Moulton Mark 1, mouldering away in a back garden. With a clean up and fresh grease, it works fantastically well and has been a revelation. Very impressed.
Nettled Shin
Posts: 644
Joined: 1 Jul 2010, 10:01am
Location: Brigadoon

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by Nettled Shin »

hubgearfreak wrote:oddly, when i put S2C into the forum's search function this thread (nor any others) came up


The best way to search the forum is to use Google & the keyword site
so typing S2C site:forum.ctc.org.uk would bring up what you needed.
Brucey
Posts: 43116
Joined: 4 Jan 2012, 6:25pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by Brucey »

hubgearfreak wrote:oddly, when i put S2C into the forum's search function this thread (nor any others) came up. i'm glad to have found the thread though. thanks Brucey for doing a strip down. 8) i'll follow it at the weekend :?


Further to my PM- On further thought I don't think there is anything inside the Roller assy other than the low gear pawls that could give your fault. If there is a fault with the low gear pawls you may be able to partially diagnose it simply by listening carefully to the sound of the pawl override in the high gear. No sound at all means that the pawls are not sprung, or there is too much grease.

On my hub you can hear a 'double click' as both pawls click over at fractionally different times in the high gear.

Good luck with the stripdown

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
Brucey
Posts: 43116
Joined: 4 Jan 2012, 6:25pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by Brucey »

Recently some discussion of S2C failure in the Sram A2 Automatix thread;

Re identification of internals: the ones here

http://mccraw.co.uk/sturmey-archer-s2c/
are different to the ones I have.

I have not found a serial number on the internals, only a production code on the hubshell. The production code is just inkjetted onto the hubshell and can just rub off quite easily. I don't know how to interpret production codes to establish production date, but maybe it can be worked out; hubs with a cable operated brake have a separate PC for the brake plate and the hubshell. They also have a casting manufacture date inside on the brake plate.
[edit my hub has a code of 'JFA8' and has 'M103 I4' marked on the reaction arm. It also has the later style of Sturmey Archer logo. SA date codes have subsequently been examined in another thread.]

I have never seen a part number marked on any SA internal parts; I think you are meant to infer by context what you have in a hub, and new spare parts I suppose will come in little bags with a magic number on. It has been my observation that they will do various things in production;

1) change manufacturing route for a part (that interchanges forward and backward OK) without changing the part number
2) change the design of a (part or parts) significantly without updating the hub model reference.

When they do the latter, the parts listing is often updated to reflect the newer parts. Very often the older version 'disappears' from the available schematics. Sometimes parts are still available, sometimes not.

BTW the low gear pawl ring on my S2C has a fair amount of radial float in it. Arguably it needs some; during brake actuation the part must find its own centre. However if it has too much float I think there is a real danger that just one of the two pawls will engage on drive takeup and this might well be fatal; these pawls are clearly not as strong as the high gear pawls and just one isn't going to be strong enough.

Pawls are a problem in many hubs; it is easy enough to end up with pawls that are too hard and are also too brittle. Even Shimano get it wrong sometimes (recent XT freehubs spring to mind). Hub gears often have very much larger, heavier pawls than (say) freewheels. I think that they often need to be of heavier construction inside hub gears, especially if there is a 12 or 16 tooth ratchet or other backlash; both these things encourage the drive to be taken up with a bit of a bang. The low gear pawls in the S2C are larger than those in a freehub but smaller than the high gear pawls, and smaller than many other hub gear pawls.

I'd be interested to see any broken S2C parts.

cheers
Last edited by Brucey on 31 Mar 2020, 2:18am, edited 1 time in total.
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
EnquiringMind
Posts: 111
Joined: 6 Nov 2011, 1:32pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by EnquiringMind »

Your suggestion that the DD pawls might be able to half-engage is interesting. Mine had smashed at the point the spring sits against, i.e. into two half pawls if that makes sense (one half remained in situ with the spring and looked fine, the other was a bit more mangled as it found its way to something which caused the hub to stop rotating while riding)
User avatar
simonineaston
Posts: 4963
Joined: 9 May 2007, 1:06pm
Location: ...at a cricket ground

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by simonineaston »

Hat's off, Brucey. Really useful info and great piccies, too. :-)
I bought a Moulton TSR2 a couple of years ago, and the 2 speed hub wheelset's been hanging on my garage wall for a while now. :roll: Your post might encourage me to get it down and see if I can make things better!
(rides: Brompton nano & ever-changing Moultons)
scottrock
Posts: 1
Joined: 21 Jul 2013, 12:49pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by scottrock »

Excellent write-up - just what I needed!
I took delivery of a custom build (Esspresso wheelz) deep rim 700c with an S2C this weekend - and I am still in the middle of setting it up correctly!
The hub has been delivered set way too tight on the cone adjustment - lots of drag in high gear (and I don't have two cones of the correct size to adjust) so I am hoping it will settle down in time.
I also noticed that the drag increased when trying to clamp the frame down - I isolated this to the dropouts being forced to an unnatural angle and trying to twist the axle - so a quick coldset on the dropouts to make them parallel at the correct OLN dimension sorted it.
In use, on the short trips I have made so far, stopping at traffic lights on a hill is a bit of trial - the hub tends to change gear unintentionally using 45T-22T (the S2C was delivered with the 22T) on 700c-23 during free-wheel, so even if you try to set up a low gear stop sometimes you fail!

If anyone has any experience of the cone adjustment settling down - let me know, or I will have to adjust!
Brucey
Posts: 43116
Joined: 4 Jan 2012, 6:25pm

Re: Sturmey Archer S2C Duomatic; Peek inside a friction box

Post by Brucey »

I think you shoud adjust the cones if the bearing is too tight; however, before doing this, do be sure that the drag springs are not responsible.

Overly tight bearing cones will show as the hub being draggy when freewheeling. You can test for this by spinning the wheel forwards when it is in the frame.

Drag in gear (when pedalling) can be tight bearings or tight drag springs. S2Cs can suffer from either (or both...).

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
Post Reply