Free Wheel Hub Tool

[Rowan]

My usual trick of sorting a troublesome freewheel hub by giving it a good boil in some oil didn't realy work this time.  So I finally got round to making a specia tool out of a 7/16 socket, it wasn't as difficult of a job as I thought it might be, 



when I worked out it was a left hand thread I soon got into the hub.  it was pretty full of gunge.

Simple little device, two pawls held in place with a little wire  (a bit of muck had got lodged beneath the wire (god knows how) and that had may be removed the springyness (i think?)

what a lot of effort gets delivered through these tiny pawls,  and curious they are not equally spaced out ?

[JonBuoy]

Shouldn't there be a third pawl?

[Paul]

Shouldn't there be a third pawl?

Most definitely.

[Paul]

So I finally got round to making a specia tool out of a 7/16 socket, it wasn't as difficult of a job as I thought it might be,

Nice work, by the way.

[Jurek]

Shouldn't there be a third pawl?

I remain to be convinced that there should.
Look at the top surface of what I am going to call the pawl holder.
Where the bevel falls away from the top surface, there are two bite shaped chunks taken out of the bevel (presumably to allow each pawl to be lifted out vertically and removed from its recess).
No such bite shaped chunk exists in the bevel above the third recess, ie: you'd not be able to drop a pawl into the recess - that said, I'd be curious to know what purpose is served by the third recess.

ETA: Nice tool.

[Rowan]

Shouldn't there be a third pawl?

that was my initial though, had I somehow lost it when taking it apart.

but on further examination the third recess was never designed to take a pawl, not sure why it is there but it certainly would not take a pawl

a couple of images showing the recesses where the pawl sits and the empty recess - it seems a curious set up wondered if there was an eplanation.

(I may have 10,000 mile on this hub, so it was working properly.  don't see any reason it won't do another 10k)

[Ian H]

I suspect the casting or forging could be machined for either two or, for a premium, three pawls.  You'd specify three for a tandem.

[Rowan]

I suspect the casting or forging could be machined for either two or, for a premium, three pawls.  You'd specify three for a tandem.

as good a reason as any.  I'm now wanting to upgrade to a 'premium three' 

[rogerzilla]

Shimano BMX freewheels are quite fun because you can reassemble them for left-hand drive.  Whether you could completely stop the thing unscrewing from the hub is another matter, although you can put it on harder with a chainwhip than the pedals can unscrew it through the gearing. 

Proper LH drive BMX kit does exist (for grinders* who prefer to use the RH pegs).

*this doesn't involve using smartphones to find local gay sex.  Or, if it did, you could get there quickly and more stylishly.

[Rowan]

With my new tool I was able to dismantle this old freehub that failed many years ago  (I knew I would one day get round to making the tool)

This failed by locking up (as in it would no longer free wheel)  quite a task having to ride a bike in this way, you have to remember to keep pedalling all the time or the deraileur pulls forward and I would imagine would soon destroy itself,   

So here it is, the inner body broke up behind the pawls, presumably the broken off pieces jammed the mechanism,  spring was well and truly bent aswell

[Brucey]

Shouldn't there be a third pawl?

Most definitely.

Definitely not. Most freehub bodies made since about 1980 are made this way, with two offset pawls.

The idea of having two pawls offset like that is so that both engage and then pull the ratchet ring off-centre. This way even if there is slight free play in the bearings, the freehub body won't necessarily be 'squirming' in every case, i.e. the applied torque often creates enough side load to prevent the bearing slop from manifesting itself.

Those with long memories may remember riding around on freewheels with two pawls which were set 180 degrees apart and were both meant to engage at the same time. If there was any slop in the bearings with this arrangement, the usual consequence was that the bearings would go 'clunk' twice per wheel revolution if you were tapping along on the small sprockets.  They still make freewheels which do this even today.

SunTour changed from 180 degree two-pawl drive to an offset arrangement like this in their freewheels, starting with the 'Winner Pro' model introduced in the mid 1980s. However it costs more to make freewheels this way so most makers don't bother even today. So much for progress.

BTW a modified socket is a good tool provided the socket is made from the right stuff; plenty of sockets are too brittle and the tool won't stand hard use. I made my own tool tears ago with a replaceable set of dogs.

cheers

[Brucey]

With my new tool I was able to dismantle this old freehub that failed many years ago  (I knew I would one day get round to making the tool)

This failed by locking up (as in it would no longer free wheel)  quite a task having to ride a bike in this way, you have to remember to keep pedalling all the time or the deraileur pulls forward and I would imagine would soon destroy itself,   

So here it is, the inner body broke up behind the pawls, presumably the broken off pieces jammed the mechanism,  spring was well and truly bent aswell

Quite probably one of the pawls broke first, and the rest followed that. Having said that, maybe a pawl seating did crack; the two pawls won't always share the load when the freehub is brand new, and they also won't share the load reliably should the bearings develop much free play.

BITD some racers would only race on well-used freewheel bodies, because until they were 'run in' the pawls wouldn't both engage at the same time (no reliably) and this could lead to breakage. I thik many freehub bodes are the same way.

IMHO it is certainly  a PITA to maintain freehub bearings but if you do it once, and do it well, it rarely needs to be done a second time. Thereafter a drop of lube every now and then and all should be well for intergalactic mileages, on a road bike, anyway.

cheers

[Jurek]

If memory serves from the last time I took my hubs apart, Mavic Ksyrium hubs have 180° opposed pawls.
I have such a hub sitting on my desk at work and yes, if you rotate it slowly enough, it'll click twice as the pawls do their thing.

[quixoticgeek]

I bought one of these on ebay, it was about a tenner shipped from china. I have yet to try it in anger.

J

[zigzag]

impressive! it looks like both pawls are intact and to break off a chunk of hardened steel it required an impact, not just torque. my guess is that the freehub skipped (common behaviour for shimano freehubs) and the momentum of the misengaged ratchet slammed into a pawl which transferred the force and broke two chunks of freehub off.

[rogerzilla]

If memory serves from the last time I took my hubs apart, Mavic Ksyrium hubs have 180° opposed pawls.
I have such a hub sitting on my desk at work and yes, if you rotate it slowly enough, it'll click twice as the pawls do their thing.

I had the misfortune of taking apart a Ksyrium hub for someone else last year and ISTR that you're right.  Mind you, Mavic is not renowned for its engineering prowess.

[zigzag]

If memory serves from the last time I took my hubs apart, Mavic Ksyrium hubs have 180° opposed pawls.
I have such a hub sitting on my desk at work and yes, if you rotate it slowly enough, it'll click twice as the pawls do their thing.

I had the misfortune of taking apart a Ksyrium hub for someone else last year and ISTR that you're right.  Mind you, Mavic is not renowned for its engineering prowess.

older mavic freehubs skip too, although less often if aftermarket pawls are installed (i'd say once in a 1000km). they've moved away from that design in their newer freehubs.

freehub should _never_ever_ skip, it's unnerving at best and can lead to a serious accident at worst.

[Foghat]

freehub should _never_ever_ skip, it's unnerving at best and can lead to a serious accident at worst.

Would much prefer it if they didn't, but every freehub I've ever owned since 1989 has skipped from new.  Admittedly only occasionally, but every single one has done it.  It is unnerving, and it's amazing they have survived those excruciating clunks over the years.

Dura-Ace 7400 x 2
Dura-Ace 7402 x 2
Dura-Ace 7403 x 1
Dura-Ace 7700 x 2
Dura-Ace 7900 x 1
Dura-Ace 9000 x 4
Ultegra 6600 x 1
Ultegra 6800 x 5
Deore XT M737 x 1
Mavic Ksyrium SL x 1
Hunt 4Season x 2
Giant Performance Tracker x 2

[Brucey]

impressive! it looks like both pawls are intact....

actually now I look at it more closely you are right, the second pawl appears to be lurking intact beneath the centre in the picture.

However I don't think it is a coincidence that the seating failed in the way it did; if there is the slightest free play in the bearings then this

a) increases the risk of one-pawl-engagement and
b) causes the pawl seating to be unevenly loaded, so that failure at one end is more likely.

If both things apply then locally the loads could be at least x4 more than normal, more again of course if there is a shock load of any kind.

BTW if you keep freewheel mechanisms of this kind bone-dry, this also increases the risk of one-pawl-drive and other problems; what appears to happen is that the (dry) pawls can get dragged (partially or wholly) out of their seat and can even cause the freewheel to jam up entirely.  One-pawl-drive and slippage (often one-half-engaged-pawl-drive...) are both more likely in new freewheel mechanisms than old ones.  If the ratchet ring is made concentric (not all are) and you keep it lubricated and adjusted to be play-free, a good freehub body is usually run-in after 5 or 10K miles. 

cheers

[Paul]

Shouldn't there be a third pawl?

Most definitely.

Definitely not. Most freehub bodies made since about 1980 are made this way, with two offset pawls.

Is there a rule (possibly by Kim) along the lines of The Minute You Think You Know Something, You Probably Don't?

[Rowan]

I have always thought the skipping problem is very temperature and mileage related, never seems to happen in the summer months but it always seems to happen on cold winter days, my theory is after big mileage they start to fill up with water, road gunk and excess oil off the chain (I use far too much oil), this then becomes too viscous in cold weather and the occasional skip then happens.?   

The one featured above that broke - this was on a cold January day and the hub had been skipping prior to it locking up.  I had just been on a steep bit of road (up on the pedals in first)  if it had skipped then, then all that energy being released released in that way would very likely cause damage.

I now have four of these free hubs, one broken two skipping and the one I have rebuilt - good news is I have done a couple of hundred mile so far on the rebuild hub, mostly in cold weather, and so far so good, no skipping.

The rebuild is straight forward enough  to do once a year, so hopefully I may have found a solution to this annoying problem.

[quixoticgeek]

I have always thought the skipping problem is very temperature and mileage related, never seems to happen in the summer months but it always seems to happen on cold winter days, my theory is after big mileage they start to fill up with water, road gunk and excess oil off the chain (I use far too much oil), this then becomes too viscous in cold weather and the occasional skip then happens.?   

Metal expands and contracts with temperature. At -5°C, things will be smaller than +40°C. Given the tolerances of modern manufacturing, this could be enough to explain the behaviour you describe.

J

[Brucey]

]

Metal expands and contracts with temperature. At -5°C, things will be smaller than +40°C. Given the tolerances of modern manufacturing, this could be enough to explain the behaviour you describe.

J

All the parts are steel, so they expand and contract together, in the absence of a large temperature differential within the assembly.

I have always thought the skipping problem is very temperature and mileage related,....

It is entirely bound by the viscosity of whatever you have inside your freehub body.  IME if you run your freehub body in oil (only possible if the seals are super) or a fairly runny  SFG (semi-fluid grease) then provided the freehub body is correctly adjusted (so that there is just no free play, with no preload on the bearings) then all will be sweetness and light for extended mileages.

Things that don't work include

a) using a #2 grease in the RH bearing and hoping it will get into the freehub body. It probably won't do, and if it does, it will gum the pawls up (=  slippage)
b) leaving the freehub body bone dry so that you can hear the pawls ticking. This can cause one-pawl engagement through pawl dragging and of course this leaves plenty of room for water to take up residence within.
3) using a #2 grease on the seals. Very few such greases are sufficiently immune to drying out or thickening when road salt gets at them and the result is that the seal runs dry, then wears, so that water gets in very easily.
4) having any free play in the freehub bearings; this causes the seals to wear out prematurely, and often causes gaps that are big enough for the water to get in even if the seals are still intact.

If the seals are Ok and the adjustment is good ( if the thinnest shim won't do it, I recommend lapping the lockring about 5-10um at a time until the free play is gone) then all that is needed is to introduce more SFG at the main hub bearings (say once a year) and that ought to be it.

One method of doing this without bothering with cone spanners is to drill a small hole in the RH dustcap, then using a small seal ring (eg made of closed cell foam) which is held in position (with low force) by the cassette lockring.  More SFG can be introduced through the hole using a nozzle-type grease gun.

It isn't a bad idea to experiment with adding oil to SFG until you get the required consistency for your conditions of use.

cheers

[Rowan]

can you explain "#2 Grease" ?

I just use multipurpose for anything like this.  Is there a more suitable?
and I tried to put as little inside the freehub as possible,  greased the pawl bearing (but kept the recess clean), and a smear overtop of the ratchet  and of course the bearings,


and what do you mean by "lapping the lockring about 5-10um"
I just tightened it as tight as poss and and it felt OK

[Brucey]

greases are graded by approximate viscosity, most usually by a scale known as NGLI. Most car/bicycle wheel bearing greases are NGLI#2. Some very easily dry out or thicken (eg by chemical reaction) to become much thicker than that.    If you use a #2 grease in a freehub body  and it gets into the pawls, it will almost certainly cause the pawls to be slow to engage and this will cause skipping and slipping, and consequent damage. The grease will only thicken over time unless it is well diluted with oil.



  NGLI#00 or preferably NGLI#000 would be nice runny SFGs which won't cause problems with the pawls.

When reassembling a freehub body it is easiest if the LH balls are set into position using a #2 grease. If you add oil the freehub body from the RHS then at worst you will have a mixture of mostly oil and a little grease in the region of the pawls. You will soon find out if out if your freehub body LH seal is good enough to hold a runny lubricant like that or not; if it is, great, but if it isn't you can add SFG the RH hub bearing area and it will work its way into the freehub body and keep the pawls happy, without leaking out too fast.

Good SFGs have shear dependent viscosity, i.e. they will stay put (instead of running away like an oil) under the forces of gravity but once sheared they are almost as inviscid as an oil, so create very little drag and allow pawls to move freely.  The best greases also have a time dependent element to their viscosity, eg they stay runny for a few seconds after they were last worked, but thereafter congeal slightly and remain in position until they are disturbed again.  Oddly enough the crude food analogy above is more accurate than you might expect; tomato ketchup does both things.  You can usually make your own ersatz SFG by mixing gear oil with a #2 grease.

When you retighten the ball cup/lockring it sits on a bed of shims, the total thickness of  which controls whether or not there is free play in the freehub bearings. Removing shims takes out free play; adding them increases it.  The thinnest shims are about 0.05mm but not every freehub body has one fitted; some have shims of ~0.3mm at the thinnest. You can make new shims using beer can (normally about 0.1mm thickness) and such shims are (because they are usually softer)  best sandwiched between extant steel shims.  So you can reshim until there is minimal free play by messing around with the shims. If you get lots of spare shims you can achieve accuracy better than 0.05mm simply by swapping thicker shims for alternatives which differ in thickness by less than this. So 0.02 or 0.03mm is possible that way.

However if you want the highest levels of accuracy then instead of adjusting the bearing clearance with shims you can grind a little off the back of the cup/lockring and take out a little free play that way. Using ~400 grit paper on a flat surface you can take off about 5um (where um = microns, i.e. 5um = 0.005mm) in half a minute or so. If you do this until there is no free play (with the cup properly tight, don't ever feel tempted to leave it less than fully tight) then IME (provided you keep the lube in and the water out thereafter) the freehub body won't need touching again, not for a very long time. However if you leave any free play in the bearings, it will only ever get worse as time goes on.

cheers