SRAM DualDrive; a common failure mode?

For discussions about bikes and equipment.
Post Reply
Brucey
Posts: 44454
Joined: 4 Jan 2012, 6:25pm

SRAM DualDrive; a common failure mode?

Post by Brucey »

I've recently been investigating the inner workings of a SRAM Dualdrive II hub.

As ever Aaron of 'rideyourbike.com' in seattle has posted some useful photos and diagrams including this;

Image

The Dualdrive hub has been around a while now and some folk have them that last for years and others seem to break them with monotonous regularity.

These are my impressions of the hub, how it might break and how to maintain it most easily.

Now there are many ways of making a three-speed hub and quite a few means of nailing on a cassette driver. SRAM must think they are doing something right because there are only detailed changes between the different models they have produced over the last few years. In the grand scheme of things some of the parts are rather small and weedy looking when compared with some other three-speed gears but they are clearly strong enough for most purposes. As with many IGHs I recommend that you continue to pedal forwards whilst shifting, but with little or no force on the pedals until the shift is complete.

If you want to lubricate the hub you can strip it down or you can lubricate it from the outside. If you select high gear on the IGH, and remove the LH axle plastic end plug (Phillips #2 point), you can add grease (not ideal if it is at all thick, it can jam pawls), semi fluid grease (a much better idea) and/or oil using a syringe or grease gun, via the hollow axle. Note that there is little in the way of an internal ridge to the hubshell, so if you use oil, most of it will come right out again; oil is good for flushing but not for long term lubrication.

-------------------------------------------------------
Control rods;

The OEM control rod is a bit flimsy (they can break off at the end) and as usual there is a crummy clickbox contraption on the RH side in most cases. If you want something different to this you can;

a) make a toggle chain pull rod (welding required) and use a radiused end nut on the RH end. You will need a shifter with a total cable pull of 10.3mm (or slightly over to get reliable gear 2 selection)

b) make a left side pushrod; (you will need to remove the LH plug). This can be a beefy 3/16" rod, because there is no spring on the LH side of the axle (unlike the other side). If you make the rod the right length (about 100mm) the rod elbow will bottom on the end of the axle before the rod is likely to break anything internal to the hub if/when the bike falls over or takes a knock.

c) make an SA-based toggle chain pull-rod. The key from an older SA AW hub can be ground to fit in the SRAM hub and then a long rod can be made up using SA parts and welding. Why bother? -well, the SA toggle thread is slightly bigger and stronger, and you are more likely to find an SA toggle chain on your travels I reckon, even if it needs modification to make it work OK. The rod needs to be longer than any standard SA rod and it needs to be a smaller diameter than most SA rods. It can be up to about 3mm diameter and still fit inside the axle spring on the RH side of the axle.
----------------------------------------------------------------
Internals;

If you remove the left cone assy the whole internal comes out in one go and removal of circlip #5 will allow most of the gear parts to be accessed. It is however less easy to reassemble the #14 clutch if you work from the left side only. It isn't impossible or anything but it is much easier if you work from the right when reassembling.

If you instead remove the RH locknut, you can remove and inspect about 75% of the internals, all without disturbing the bearing adjustment (the right cone butts up against a shoulder on the axle, so the adjustment is the same when the hub is reassembled provided you the LH cone is undisturbed) and reassembly of #14 is a fair bit easier. The driver #17 rides on a shielded deep groove cartridge bearing on its left side and a cup and one bearing on its right side. As is usual for SRAM hub there is stuff-all factory grease inside and some corrosion is commonplace; the hub is not well sealed.

Unfortunately you can't inspect the #7 pawls by disassembling on the right side only, but you can just about see if they are working or not. This end of the hub seems to be a trap for all kinds of crud, so if there have been any problems inside the hub the debris might foul the #7 pawl engagement.

The gear works in the same way as many other three-speed gears but there are some subtle differences. The DD-II gear defaults to the low gear when the control rod is removed. The pawls on the planet cage (#7) transmit the drive to the hubshell in gear 1 and gear 2 (in contrast to an SA hub where they are used in gear 1 only).

If you want to see how the hub works, you can assemble the hub internals (less gear ring #8) and see the various parts moving and how the clutches are operated.

In gear 1 the pawls on the driver (#17) drive the gear ring (#8) and thence the planet cage (#7) and thence the hubhell via the #7 pawls. The #8 pawls and the #10 pawls are defeated by part #12 and the toggle key assy (shown as part of the axle in both images) respectively.

In gear 2 the #10 pawls are allowed to drive the planet cage (#7) and thence the hubshell using the same pawls as in gear 1. The drive is transferred from the driver to #10 via a dog clutch (#14) that is in permanent mesh in all gears. The #17 pawls are freewheeling because #8 is going faster than the driver.

In gear 3 the #10 pawls work as in gear 2 (driving the planet cage #7), but now the #8 pawls are allowed to drive the hubshell directly.

So why do these hubs sometimes break? Well basically they have four pawls on #7 that are meant to engage two at a time. However this assembly has been designed for small backlash and given this feature it is not made accurately enough for two-pawl engagement to be guaranteed.

This means that for some of the time only one pawl will be transmitting the drive, which is A Bad Thing. This will put a large side load onto the planet cage and this will then wear against the axle. Obviously the one pawl can break as well as the axle wearing etc. Once the axle is worn, the chances of the pawls engaging two at a time (as intended) are further reduced; the worse it gets, the worse it is going to get. There are multiple reports on the internet of failed #7 pawls and worn axles/planet cages and I can see exactly how this will happen.

To test if your hub is likely to do this, clean the grease from the #7 pawls and ratchet ring and reassemble the hub, being sure to eliminate any slack from the hub bearings. How turn the hub slowly (as per backpedalling or freewheeling in gear #1 or gear #2) and listen carefully. If you turn the hub slowly enough, you will hear each pair of pawls as a 'click-click' sound. Then the second set of pawls will make a similar 'click-click' sound, with a different spacing, then the first pair again, and so forth. If you assess what proportion of the time spent is between clicks, and how much is between pawl sets, you can get a good idea of how likely single-pawl engagement is; I would say that any more often than one time in twenty would be cause for concern, and that this is quite likely to be the case.

The cure is likely to be a different pawl spring arrangement, and/or trimming the pawl lengths/numbers so that they are more likely to engage simultaneously in pairs as intended. The OEM #7 pawl spring is a coil type, similar to the kind of spring that is used on hydraulic shaft seals etc. I think that it is soft enough to extend lengthwise easily and that this may allow the pawls to move in an unhelpful fashion even if they are the correct length; a conventional solid wire spring may be a better idea, especially if it only has one pair of pawls to control. If one pair of (diametrically opposed) pawls is simply removed, the backlash will double, but so will the chances of proper simultaneous pawl engagement, too. This might well make for a more reliable hub overall.

[BTW if the planet gears were designed to run with minimal radial clearance to the ring gear and the sun, this would better support the planet cage against side loads, and this would mitigate the effects of single pawl engagement. However the gears are dead sloppy, with barn door clearances, meaning that any side load on the planet cage is borne between the planet cage and the axle, with predictable consequences.]

So overall, not a bad effort from SRAM but the poor pawl engagement in gears 1 and 2 could be a serious problem that can lead to a premature demise. As to whether this is likely, it is a pot luck job depending on the build tolerances of the hub. If you do the tests outlined above, you can at least see if it is at all likely.

I note with interest that they have used the same design feature on some other SRAM hubs such as the i-9; further photos here;

http://www.rideyourbike.com/sramIGH.shtml

suggest it can be a weak spot in some of these other hubs too.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
User avatar
squeaker
Posts: 4112
Joined: 12 Jan 2007, 11:43pm
Location: Sussex

Re: SRAM DualDrive; a common failure mode?

Post by squeaker »

Thanks :) Page saved ;)

IME this hub is dead easy to 'power shift' but probably not a good idea...

Just for devilment, where would you put your money - SRAM DD3 or the 100g heavier SA CS-RF3?

PS: I like the clickbox as it removes the need for a cable stop on the chainstay (or the, IMHO, messy SA pressed steel axle mounted thingy) but I understand why people don't like the clickbox every time I have to refit it - which is why I use a Marathon Plus on the wheel it's fitted too :lol:
"42"
Geoff.D
Posts: 1982
Joined: 12 Mar 2010, 9:20pm

Re: SRAM DualDrive; a common failure mode?

Post by Geoff.D »

Smashing. Thanks.
Brucey
Posts: 44454
Joined: 4 Jan 2012, 6:25pm

Re: SRAM DualDrive; a common failure mode?

Post by Brucey »

squeaker wrote:Thanks :) Page saved ;)

IME this hub is dead easy to 'power shift' but probably not a good idea...


actually on upshifts this will be OK up to a point and on downshifts I'd expect it not to go except in its own good time, i.e. once you have backed off some. On the 1 to 2 upshift the engaging pawls (#10) are small and set on a small diameter (bad; probably this is where the hub will fail first under extreme torque loads), but they do have a deep engagement so the chances of a mis-shift are not that great. The 2-3 shift should be pretty good too. If you were going to choose an IGH to powershift on upshifts with, you could do worse.


Just for devilment, where would you put your money - SRAM DD3 or the 100g heavier SA CS-RF3?


I think the DD3 has a few detail changes vs the DD2 ( and that DD2s in themselves vary). I've not had one in bits yet but I think the design of the CS-RK3 ought to be capable of withstanding more torque, but shifting is probably less slick and the manufacturing quality level can be suspect; but then so can it be with SRAM too. In fairness if there is a quality problem with either it ought be obvious early on and covered under warranty. The SA hub is a bit cheaper too IIRC. I don't think 100g is a big deal, either, so I guess the CS-RK3 would probably get my nod.

[ edit; it seems the failure of the #10 pawls in the DD hub is indeed a risk under very high torque loads. I suspect that if the #7 pawls are misbehaving, the planet cage will be skewed and the #10 pawls (which engage with the RH end of the planet cage) then see loads that they shouldn't, hastening their failure even at more modest loads.

If you want to improve the DD hub (at the expense of ease of reassembly) then, as in many other similar hubs, trading the clipped ballraces for loose balls is a pretty good idea. In fairness they don't normally give trouble unless the adjustment is a bit too slack, but on the DD hub the LH bearing is liable to be seriously overloaded if a disc brake is fitted in (say) a trike application. I'd regard using loose balls in this bearing as an essential upgrade in this case. It is a good idea to fit loose balls in the RH bearing too, and some folk even fit them in the main ring bearing on the driver, usually having had the clip break up first. If you do this, you go from having 12 balls to about 26 balls; much stronger but more of a faff to reassemble for sure.]

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
devdahon15
Posts: 2
Joined: 5 Jan 2015, 11:14am

Re: SRAM DualDrive; a common failure mode?

Post by devdahon15 »

Hi.

Having a problem with one of these units (DDII) - or rather my mechanic is! (Have forwarded a link for this thread to him.)

Anyone able to help on these points?

1. Apparently parts for them are hard to come by (I called C H White in Malmesbury). Is there another specialist repair place in the UK?

2. 2 out of three gears are working (flat and uphill icons on shifter) - is this significant?

3. A sticker warns that radial lacing of the hub will invalidate the warranty. At least one side is radially laced - relevant?

Thanks in anticipation.
Brucey
Posts: 44454
Joined: 4 Jan 2012, 6:25pm

Re: SRAM DualDrive; a common failure mode?

Post by Brucey »

devdahon15 wrote:Hi.

Having a problem with one of these units (DDII) - or rather my mechanic is! (Have forwarded a link for this thread to him.)

Anyone able to help on these points?

1. Apparently parts for them are hard to come by (I called C H White in Malmesbury). Is there another specialist repair place in the UK?


Your dealer ought to be able to deal with Fisher who distribute SRAM stuff in the UK. I think they have some clever chaps there and a heap of parts but this mayn't be any use to you directly; you can't talk to them unless you are a bike dealer and they are mostly concerned with warranty work on current hubs. SRAM's technical support documentation is woefully poor; there are not even any part numbers to reference, only the item numbers on dwgs like the one I posted above. This means the only way to ascertain if the parts from one hub will swap into another is to actually get them in your hands and see for yourself.

2. 2 out of three gears are working (flat and uphill icons on shifter) - is this significant?
barring a simple adjustment/external fault eg with the clickbox, (test by manually selecting high gear eg by using a pushrod from the left side) it most likely means that the #8 pawls are not working. If the fault has arisen when the bike has been standing, my best guess is that they are jammed, either with dried grease or through corrosion. You may well find that other SRAM three-speed hubs use similar components in this part of the hub, so don't assume that a repair is impossible until you have exhausted this possibility.

3. A sticker warns that radial lacing of the hub will invalidate the warranty. At least one side is radially laced - relevant?


Radial spoking tends to break flanges if they are not designed for this spoking pattern. It should not be relevant to the condition of the internals.

My local dealer is pretty good and they deal with Fisher on a regular basis. Even so getting spare parts can descend into the realms of comedy; when they asked for a cone (for a fairly common hub) a while back they sent a bag of five, and none of them were the right one.

I also needed a DD-II cone a few months ago, and I'm still waiting for it. In order to be able to use my hub in the meantime, I found an SA cone with a slack tolerance to it (the thread size is nominally 0.2mm smaller but with the same pitch) and I reground it to the correct profile.

With spare parts support like that, I am thinking that I need a very good reason to buy SRAM stuff, or it needs to be 'disposably cheap'.

Raleigh are now fitting bikes with SRAM kit so maybe the spares situation will improve a little; it seems to me that it could hardly get much worse.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
devdahon15
Posts: 2
Joined: 5 Jan 2015, 11:14am

Re: SRAM DualDrive; a common failure mode?

Post by devdahon15 »

Thanks, brucey.

Will pass on to my mechanic - very reasonably, he's treating this as an "education" experience. But am not sure how long his patience/my budget will last.
Cyckelgalen
Posts: 220
Joined: 21 Sep 2018, 11:29am

Re: SRAM DualDrive; a common failure mode?

Post by Cyckelgalen »

Having a couple of Dualdrive-equipped bikes in my harem, I had gone through this very instructive thread several times in the past but now, unfortunately, the time has come to put theory into practice.
After stripping and servicing the hub, I ended up with a broken pawl, one of the four pawls in the planet cage, #7.

The hub was in seemingly perfect condition appart from the dried up grease, no signs of wear or corrosion, pristine cone races.
I broke the pawl myself at the very last step of the job. The internal assembly should be inserted into the hub she'll with a twisting motion the last few millimetres to allow the pawls to retract and lodge into the hubshell's spline ring. I was too concerned with the bearings staying in place and thinking that the assembly was fully inserted, stated to thighten the cone and broke one pawl in two pieces

So beware anyone attempting to service a Dualdrive, quite a straightforward procedure otherwise.

I am aware that there are no spares to be had for any part of this hub, other than cannibalising a discarded hub, which I don't have. But going carefully through Bruceys's analysis I realise that I could remove the opposing pawl to the broken one and still keep a pair of functioning pawls. If the the hub is designed to have only one pair of #7 pawn at the time, it wouldn't have any negative effects and it may even be an improvement as a single pair of pawls reduces the risk of only one pawl engaging at the time.
And I will end up with a spare pawl on top of that! I hope all these assumptions are correct.
Attachments
image.jpeg
hemo
Posts: 1438
Joined: 16 Nov 2017, 5:40pm
Location: West Sussex

Re: SRAM DualDrive; a common failure mode?

Post by hemo »

I had an issue with my sram dual drive and also found the same snapped pawls, iirc I replaced them with some DTSwiss or might have been Sturmey generic pawls I bought off one of the cycling web sites. They were as near as a damn and worked, may have been SJScycles.
Last edited by hemo on 29 Dec 2020, 3:57pm, edited 1 time in total.
Brucey
Posts: 44454
Joined: 4 Jan 2012, 6:25pm

Re: SRAM DualDrive; a common failure mode?

Post by Brucey »

if you are buying new pawls, best to buy a set so that they all match one another rather than one at a time.

Some dual drive spares here

https://www.daszweirad.de/shop3/de/sram-getriebe-nabenteile/sram-dual-drive-ii/

but not pawls.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
hemo
Posts: 1438
Joined: 16 Nov 2017, 5:40pm
Location: West Sussex

Re: SRAM DualDrive; a common failure mode?

Post by hemo »

I now seem to think the ones I bought were a a mm or so too wide so I had to use the mini grinder on them but they appeared to do the job.
Cyckelgalen
Posts: 220
Joined: 21 Sep 2018, 11:29am

Re: SRAM DualDrive; a common failure mode?

Post by Cyckelgalen »

Thanks. Interesting that you can adapt pawls from other manufacturers. But to be honest, I had given up on trying to replace the cracked pawl. I intended to remove one more and keep a pair of diametrically opposed pawls. If it is so that only one pair of pawls engage simultaneously, as Brucey concluded in the original post (5 years ago), then the hub will be fully operational albeit with a bit more backlash.

It is intriguing however, that the highest torque gears rely on only one pair of these pawls that are the flimsiest of of the three sets of pawls in the hub.

That German site has got a few interesting dual drive parts. I heard that the bearing were non-standard size and difficult to source.
Post Reply