Uneven chain-ring wear with Speedhub?

[ndt44]

Hi all,

I'm currently on a London - Tokyo tour with a Thorn Raven Nomad, and I've just noticed some odd behavior from my drive-train. I've done about 5000km so far, and am about to do the first oil change. When I back-pedal the cranks, I notice the chain goes slack-tight-slack with each revolution. I've pulled the chain off to give it a good clean, and I can't see any 'wobble' of the front chainring, side-to-side or off-centre. There does seem to be some significant wear on the chainring though. The crank axle isn't bent, and looking at the edge of the teeth as I spin the cranks, I can't see any variation in the profile of the chainring. The only solutions I can think if for this behavior are:

* Uneven chain-ring wear, making the gear somewhat elliptical - some teeth have been worn deeper than others, such that the 'rolling diameter' varies, despite the circular appearance

* Uneven chain stretch (highly unlikely, if not impossible in my opinion)

The chain is really quite slack at one point of a revolution, and then very tight at the other extreme (180 deg later). I have at least another 10,000km to go, so I'm a bit worried about this. I may try rotating the ring 180 degrees to see if I can wear the other side equally, or flip the chain-ring over and start 'fresh' - but I wonder how this could have happened in the first place...?

Edit: I should mention that I have tensioned the chain twice so far, being quite careful not to 'over-do it'. However I did notice some faint pedaling resistance after the initial tensioning - kind of like a stiff link in the chain - but every time I checked the chain it was no tighter than I was shown to do it at the Thorn shop. I have a feeling that I've had this lop-sided behavior for a while, and that my chain-checks were all at a point on the crank rotation where the chain was relatively loose.

Any ideas / advice welcome.

Cheers,
- Nigel

[Al Downie]

Hi,

Is yours an XT chainring, from about a year or two ago? I can't remember the exact model code, but there was one set of XT chainrings which was made from *very* soft alloy which wore out incredibly quickly. After a huge number of complaints, they could hardly give these rings away. I'm wondering if yours has worn corresponding to the two 'maximum force' segments of the revolution? (the same segments that 'Biopace' was meant to affect)

[ndt44]

Hi Al,

No, my ring is the Thorn 38T super-thick item, and the tension only occurs at one point of a complete revolution, rather than at two (a-la Biopace).

I've combed the bike shops here in Istanbul for a new 38T ring, but came up dry (unsurprisingly)

I used to have a Biopace set on an old MTB from the late 80's - people used to hate them, but I never found too much to complain about...

Thanks for your input!

- Nigel

[stutho]

My first guess would be that the chainring isn't quite centred on the spider.  My second guess would be that there is some damage to the spider or chairing.  However from what you have said you have checked out these possibility.

I'd rotate the chainring by 2 chainring bolts.  If the tight spot moves with the chainring then you know the fault is with the chainring, If the tight spot stays aligned with the crank position then there is a problem with the spider or the bottom bracket alignment

Best of luck

Stuart

[ndt44]

Hi Stuart,

Thanks for your ideas - yes, I've had a pretty thorough look at the spider, axle, ring, chain & sprocket. Everything looks pretty good and centred, the only noticeable issue is the wear on the chain ring. I'll pull it off and rotate 2 holes as you say, and check where the tautness occurs, then report back here.

Cheers,
- Nigel

[ndt44]

Well, I've tried rotating the chain-ring 180 degrees, and the tension still occurs at the same point in the crank revolution. From this I take it that the chainring itself is not the culprit. When I take a very close look at the chain-ring / chain-gurad as I turn the cranks, I can see a slight off-centre wobble - about 1 mm or less. It seems much too small to account for the difference in chain tension I'm experiencing, but it's a possibility, I suppose...

The bike has had a bit of a rough time - I was hit by a car (quite deliberately) in Romania, heading into Bucharest, but the car struck the front LHS pannier and caused the bike to drop on the left side, not the right (crank) side. No damage visible - pedals, bar-ends and panniers were all fine (me too), so I can't imagine it would be enough to bend or distort a crank axle / spider.

I'll keep looking into it, but I think I'm just going to have to live with it...

- Nigel

[jonathan]

Have you checked the BB spindle? Could it be bent enough to be creating the problems. What kind of BB is it, if external bearings could one of the cups be misaligned? If normal bearings could the cups be misaligned.

[stutho]

If it is not the chainring then you are left with the spider or the bottom bracket (BB) as the problem.  If it is the BB spindle there is a quick fix.  

Remove both cranks; Remove the EBB from the frame; Put the EBB in the wrong way round (i.e right side to the left side); Replace cranks.

You will now be using the BB in the wrong direction. This WILL work as a temporary fix.  Long term you may loosen the BB in the EBB so you will need to keep an eye on this.

If this fix doesn't work then the fault must be in the spider  (there is nothing else to try).

Best of luck

Stuart

[Al Downie]

quote:When I take a very close look at the chain-ring / chain-gurad as I turn the cranks, I can see a slight off-centre wobble - about 1 mm or less. It seems much too small to account for the difference in chain tension I'm experiencing

I guess it depends on how much of the circumference has a 1mm smaller radius? If half of the chainring has a smaller radius than the other half, then that would account for a fairly big fluctuation in chain tension.

Although you've had a 'close look', it might be worth being a bit more rigorous? Stick a big blob of blu-tack on the seat tube, and stick a small allen key (etc) into that, so that the end of the allen key rests in space just about a millimetre from the extremity of the chainring, and then start turning the crank. That should make any irregularities more obvious and measurable...

[ndt44]

Thank you all for your ideas & suggestion, I'll certainly give them a shot if things get any worse. I've run out of time at the hostel I'm staying in, so I'll have to get moving...

I suppose the sub-mm off-centre wobble could add up to a few mm of tension over the circumference of the entire chainring. Reversing the BB and mounting the cranks on the (assumed) unbent end of the axle would be a solution (the BB is a cartridge type - 2006 XT cranks if I'm not mistaken, without the see-thru hollow axle)

I'll investigate further when I get a chance - probably Ankara, in a week's time.

Thanks again for your feedback!

Cheers,
- Nigel

[mountaincarrot]

Hi ndt,

I instantly wondered whether this might be related to some sort of "synchronous wear" (for want of a better expression).

If the chain were for example 105 teeth, and the front sprocket 42, then every 2.5 revolutions the chainring would encounter the same teeth. As the biggest power (and thus biggest wear) is with the pedals at "quarter to nine", which occurs twice per revolution, there would be constant wear points (and hence chain stretch pattern) along the chain, which would amplify wear pattern on the chainring. That could affect the apparent chain tension with relation to rotation position.

This would not occur if there was "slippage" (ie a ratio not exactly divisible by half the chainring)

If this were the case, I'd guess you may find the chain goes tight at the "half past twelve", and loose at "quarter to nine", and you might like to try moving the chain ~ 11 links forward.

BTW, I have never heard of this before, so it's pure speculation!

Rgds

[jonathan]

Somewhere on Sheldon Brown he talks about a way to avoid chainwear by always having the same chainlink in contact with the same chainring tooth.
Its is here: http://sheldonbrown.com/chain-life.html
It might add to your speculation Mountaincarrot.

[mountaincarrot]

Hi Jonathan,

Thanks for the link. Like a light went on - Yes obvious when you see it that way. By my way of thinking though, it only applies if both the chainring and the chain count are even. Sheldon doesn't seem to point out the latter.

I'll take care with my chainring and chain count when I get the Rohloff in then. - Or presumably using a half link chin overcome the problem.

But regarding my earlier post. This was assuming a very special case when half the chainring divides exactly into the chain. - So there is no overall circulation of the chain wear RWT the ring. I think this could still occur and lead to the loose/tight problem.

Rgds

[bike_the_planet]

As an aside, is anyone experiencing uneven sprocket wear for a similar reason - ie having a setup whereby the number of teeth on the chainring is exactly divisable by the number of sprocket teeth? An example would be a 40/16 setup - the sprocket will rotate exactly 2.5 times for one rotation of the pedals.

This would be a good indicator of exactly how wear-resistant the sprocket is.

Cheers

[graham]

If the chainring is 0.5mm off centre, I think that'll cause a difference of 1mm from lowest to highest point, and since the chain comes off top and bottom of the ring, that'll introduce 2mm of slack into it at the worst point.
2mm of slack would give much more than 2mm of 'sag' in the chain. I think a relatively small off-centre will give quite an obvious effect.
I get my (Thorn alloy) chainrings machined out in the centre so I can put them on the middle position on the spider. My dad (who's tooled up for the task) had a lot of trouble centering one ring because it wasn't a true circle. And sure enough, on the bike, there was a noticeable slack-tight effect even with new ring and chain.
I think it likely that many chainrings from many manufacturers might not be perfect circles. On a derailleur bike the rear mech compensates for it so no-one ever notices.