Weird Behaviour Campagnolo Potenza

[Graeme_FK]

Loads of guesswork going on here.
I'd go back to basics and check things carefully, one at at a time, starting with the fundamentals.

Check the cassette lockring is tight.
Check the rear wheel is in straight.
Check the hanger bolt(s) is (are) tight.
Check the hanger alignment in both planes. To do this with any degree of accuracy, you need a dropout alignment gauge (DAG).
Check that the lower jockey wheel is fitted "right way around" - they are directional. The Campagnolo logo and the direction of rotation arrow are on the outermost surface only - so you can see them when looking at the installed jockey wheel from the gear side.
Check the torques on the jockey wheel bolts.

That eliminates all the factors, apart from wear and tear, that might have changed since installation - the OP says this is a "recent" problem.

Check the lower jockey for side-to-side play - I wouldn't expect it to be significant after 1600 clean km but it might be an issue. The chain coming off the side and sticking against the inside of the cage might give you the issue you describe, as Yellow Saddle says.

Next, basic set-up checks, using Campag's spec.

We'll assume the rear hanger is to spec and not marginal in a way that might work OK on a brand new system but become a problem with wear and tear on the chain / cassette. Very few hangers in the market are problematic like that (though there are one or two).

Small ring to small sprocket - what is the space between the idler run of chain and the run of chain that passes around the "bottom" of the upper jockey wheel. It should be 8-15mm ideally, although 5 - 20mm is "unofficially" workable.

That sets the chain length and is Campagnolo's spec. No idea where that "45deg from the horizontal" of the rear derailleur comes from.

Check the chain carefully for any evidence that a link is starting to "open". We've seen it happen so seldom in a correctly-joined 11s Campag chain since 2009 that I'd almost discount it as a possibility ... but it might cause an occasional "snag" if it were to happen - and that could give your symptoms. Poor joining can also cause an issue but were that the case I'd expect you to have had the issue from day 1. Usually I'd expect a once-per-chain revolution "ticking" noise as well.

11s chain, we don't advise adding links in. General advice is to avoid breaking the chain if you don't need to and to only use the calibrated outer link plate when you do, 2x max.

So, if this test comes up with a chain that is too short, we'd advise a new Campag 11s chain. There is no functional difference between the Record 11 chain and the "Campagnolo 11" chain. Either will be fine. It's unlikely that you will need a new cassette as well is you have only 1600 "clean" km on the existing one and that chain wear is within the "0.75" marker on a Park or similar chain gauge / the Campag internal roller-to-roller measurement of 132.6mm

Rivet and peen the Campagnolo chain correctly using UT-CN300. Don't use a joining link like KMC etc. The fit may not be good and we do see problems quite frequently with them. Use a Campag chain. Third party chains are all a compromise between the exact requirements of Campagnolo, Shimano and SRAM, all of whom have slightly differing specs.

Small ring to biggest sprocket - ideally you need 5-7mm between the tops of the teeth of the upper jockey and the tops of the teeth on the biggest sprocket. That sets the tracking angle of the RD so keeping the top jockey as close as possible to the optimum distance from the cassette across the range. Set with the "H" screw at the derailleur cage. Leave the upper ("G") screw alone unless you can't, because of the design of the hanger, get the adjustment correct on the "H" screw (very rare). The G screw is alloy and difficult to move, Don't break or burr it, it's not sold as a spare part.

If all of that comes up trumps, there is something very odd going on and it may need professional help to diagnose as it'll need an eyes-on approach.

 

[Sillyoldman]

Ok lots to check here. I have a chain on order (Chorus 11 speed) and have just ordered a DAG tool to check the alignment. The lower jockey wheel is definitely ok with correct rotation direction arrow. I just serviced the free hub and the lock ring was tight to undo and is torqued back to spec. I will check the other things hopefully tomorrow. Going to swap the chains to get an extra link in and remove any doubt that there is anything wrong with the chain.

I have never seen torque specs for the jockey bolts. Do you have those Graeme?

Thanks everyone for the input.

 

[Yellow Saddle]

Loads of guesswork going on here.
I'd go back to basics and check things carefully, one at at a time, starting with the fundamentals.

Check the cassette lockring is tight.
Check the rear wheel is in straight.
Check the hanger bolt(s) is (are) tight.
Check the hanger alignment in both planes. To do this with any degree of accuracy, you need a dropout alignment gauge (DAG).
Check that the lower jockey wheel is fitted "right way around" - they are directional. The Campagnolo logo and the direction of rotation arrow are on the outermost surface only - so you can see them when looking at the installed jockey wheel from the gear side.
Check the torques on the jockey wheel bolts.

That eliminates all the factors, apart from wear and tear, that might have changed since installation - the OP says this is a "recent" problem.

Check the lower jockey for side-to-side play - I wouldn't expect it to be significant after 1600 clean km but it might be an issue. The chain coming off the side and sticking against the inside of the cage might give you the issue you describe, as Yellow Saddle says.

Next, basic set-up checks, using Campag's spec.

We'll assume the rear hanger is to spec and not marginal in a way that might work OK on a brand new system but become a problem with wear and tear on the chain / cassette. Very few hangers in the market are problematic like that (though there are one or two).

Small ring to small sprocket - what is the space between the idler run of chain and the run of chain that passes around the "bottom" of the upper jockey wheel. It should be 8-15mm ideally, although 5 - 20mm is "unofficially" workable.

That sets the chain length and is Campagnolo's spec. No idea where that "45deg from the horizontal" of the rear derailleur comes from.

Check the chain carefully for any evidence that a link is starting to "open". We've seen it happen so seldom in a correctly-joined 11s Campag chain since 2009 that I'd almost discount it as a possibility ... but it might cause an occasional "snag" if it were to happen - and that could give your symptoms. Poor joining can also cause an issue but were that the case I'd expect you to have had the issue from day 1. Usually I'd expect a once-per-chain revolution "ticking" noise as well.

11s chain, we don't advise adding links in. General advice is to avoid breaking the chain if you don't need to and to only use the calibrated outer link plate when you do, 2x max.

So, if this test comes up with a chain that is too short, we'd advise a new Campag 11s chain. There is no functional difference between the Record 11 chain and the "Campagnolo 11" chain. Either will be fine. It's unlikely that you will need a new cassette as well is you have only 1600 "clean" km on the existing one and that chain wear is within the "0.75" marker on a Park or similar chain gauge / the Campag internal roller-to-roller measurement of 132.6mm

Rivet and peen the Campagnolo chain correctly using UT-CN300. Don't use a joining link like KMC etc. The fit may not be good and we do see problems quite frequently with them. Use a Campag chain. Third party chains are all a compromise between the exact requirements of Campagnolo, Shimano and SRAM, all of whom have slightly differing specs.

Small ring to biggest sprocket - ideally you need 5-7mm between the tops of the teeth of the upper jockey and the tops of the teeth on the biggest sprocket. That sets the tracking angle of the RD so keeping the top jockey as close as possible to the optimum distance from the cassette across the range. Set with the "H" screw at the derailleur cage. Leave the upper ("G") screw alone unless you can't, because of the design of the hanger, get the adjustment correct on the "H" screw (very rare). The G screw is alloy and difficult to move, Don't break or burr it, it's not sold as a spare part.

If all of that comes up trumps, there is something very odd going on and it may need professional help to diagnose as it'll need an eyes-on approach.

Well, you have certainly added to the guesswork.

Where does the "45 degree to the horizontal come from?" Firstly, it is a pretty common and reliable way to check for a too-short chain in a photo or just at a glance.

Secondly, you can experimentally and mentally work it out for yourself. Experimentally, shorten a chain and you'll see that the RD bends forward by more than 45 degrees and displays all the symptoms the OP describes. Mentally, it comes from the fact that chainstays on racing bikes are all the shortest they can possibly be without a 23mm tyre fouling the seat tube. This sees to it that the bottom run of the chain is very short and, when the bike is in the big-big gear, the angle of the slack chain run (viewed from above or below or side-to-side if you like) is beyond the maximum that the average jockey tension pulley can handle. With the chain too short, the angle is even steeper (it travels the same lateral distance but over a shorter longitudinal distance) and the chain enters the jockey wheel with an unacceptable yaw. When pushed backwards or receiving a quick change on the front chainrings, this jams the chain in the jockey cage.

The 45-degree rule-of-thumb is a very easy way to diagnose such problems on all short-chainstay bikes.

I trust that clears things up for you and puts a new arrow in your quiver.

 

[Graeme_FK]

@Yellow Saddle - yes, it can give a reasonable approximation under some circumstances - but the "small to small" & check the gap method is
1) exact and
2) the way Campagnolo - who know how they want it set up - specify that chain length is set. Making that check, there's no "does it look right", it is either within the range or isn't - and is just as easy a visual check.
3) It works for Shimano & SRAM because of the way that they recommend chains are sized but would only have any validity in a Campagnolo system when the capacity is maxed out - because Campagnolo don't base their chain length recommendation around the size of the biggest sprocket, preferring to ensure that (providing chainring sizes stay the same), any cassette within the operating range of the RD can be mounted with no fear of running too short a chain for the size of the bottom sprocket.

The chain yaw issue is not unknown and is compensated for in design by the slight twist that you can see in Campagnolo 11 & 12s jockey cages and is the reason that the upper pivot also has a small degree of designed-in play, regulated by two o-rings.

Installation and testing, of systems once installed forms part of what I do ... for a living ... for Campagnolo ...

Jockey wheel fixing bolt torques are 2.7nm (24 in lbs) top and bottom, wet thread. We advise Loctite 248.

 

[Ajax Bay]

Rivet and peen the Campagnolo chain correctly using UT-CN300. Don't use a joining link like KMC etc. The fit may not be good and we do see problems quite frequently with them. Use a Campag chain.
Third party chains are all a compromise between the exact requirements of Campagnolo, Shimano and SRAM, all of whom have slightly differing specs.

Thank you for your comprehensive comment.
Please could you share the "exact requirements of Campagnolo" chains and how they differ from other (compromised) 11 speed chains, Graeme? I remember that there was something 'odd' about Campagnolo 10 speed but I thought Campagnolo had led the way (?in 2009) with 11 speeds and others had adopted the same chain specs:
Link = 1 inch (25.4mm)
Inner width (9-12 speed, between inner plates) = 11/128″ (2.18 mm)
11 speed outer width = 5.62mm

https://bike.bikegremlin.com/3555/bicycle-drive-chain-dimension-standards/

Leaving aside 'C-preciousness' and/or warranty (if any), why, in practice will a 'normal' (KMC say) 11 speed 'quick-link' not work entirely satisfactorily on a Campagnolo chain? What sort of "problems [do you see] quite frequently with [joining links]"? My experience is that riders/users joining a chain with a pins (eg CN-RE500 Ultra-link or Shimano equivalent) is more likely to result in a problem than joining it with a 'normal' quick-link.

 

[Sillyoldman]

Ok I have checked the items and my findings are as follows:
1)Check the cassette lockring is tight
- Yes it is tight
2) Check the rear wheel is in straight
- Yes it is
3) Check the hanger bolt(s) is (are) tight
- Yes they are
4) Check the hanger alignment in both planes. To do this with any degree of accuracy, you need a dropout alignment gauge (DAG).
- I checked with a Park Tool DAG and was really surprised that top to bottom (valve rotated to remove any slight wheel issue) measured 5 mm. The Park tool instructions state 4 mm is the limit. I have a brand new hanger which I fitted and got exactly the same reading of 5 mm. Front to back measures ok at 3 and 9 O'clock. Given the rear shifting is so smooth (other than this occasional issue) is this cause for concern?????
5) Check that the lower jockey wheel is fitted "right way around" - they are directional. The Campagnolo logo and the direction of rotation arrow are on the outermost surface only - so you can see them when looking at the installed jockey wheel from the gear side.
Check the torques on the jockey wheel bolts.
- Yes the Jockey wheels are correctly installed with the lower one showing the correct rotation direction orientation. By luck the torque was OK .
6) Check the lower jockey for side-to-side play - I wouldn't expect it to be significant after 1600 clean km but it might be an issue. The chain coming off the side and sticking against the inside of the cage might give you the issue you describe, as Yellow Saddle says.
- There is no significant play at the Jockey wheels

7) We'll assume the rear hanger is to spec and not marginal in a way that might work OK on a brand new system but become a problem with wear and tear on the chain / cassette. Very few hangers in the market are problematic like that (though there are one or two).
- Hangers (two have been tested) are from the frame manufacturer (Enigma) so I feel confident that these are ok.
8) Small ring to small sprocket - what is the space between the idler run of chain and the run of chain that passes around the "bottom" of the upper jockey wheel. It should be 8-15mm ideally, although 5 - 20mm is "unofficially" workable.
- I am on the limit at 20mm measured with a digital micrometer. I have not fitted the new chain yet as I am waiting to receive it.

9) Check the chain carefully for any evidence that a link is starting to "open". We've seen it happen so seldom in a correctly-joined 11s Campag chain since 2009 that I'd almost discount it as a possibility ... but it might cause an occasional "snag" if it were to happen - and that could give your symptoms. Poor joining can also cause an issue but were that the case I'd expect you to have had the issue from day 1. Usually I'd expect a once-per-chain revolution "ticking" noise as well.
- the chain appears to be ok, no stiff links. The chain measures 132.07mm measured in accordance to the Campagnolo 6 links method, so is within the wear limits (but will be replaced anyway).

To recap, the only possible issue I can see thus far is the Rear hanger and would welcome comment as to the relevance of this.

Many thanks

 

[Yellow Saddle]

To recap, the only possible issue I can see thus far is the Rear hanger and would welcome comment as to the relevance of this.

Many thanks

Poor gear changes from a bent hanger presents in a strange way in that it allows perfect shifting in one direction and poor shifting in the reverse. You have not reported that, so that's not the primary problem. I'm still with the chain-too-short theory, but time will tell.

If you get the same hanger angle reading on a new hanger as on the old one, the frame is suspect. Have a very good look at the hanger recess on the frame itself. Use straight edge to see if it is flat and, loosen the hanger bold a bit and see if the hanger rocks in its cradle when you hold it firmly in place but try and rock it. It has to fit squarely without ambiguity. Before you start bending the frame, also check the hanger for flatness, sometimes they are a bit concave at the mating surface.

Keep in mind that the ParkTool DAG for the hanger measurement is a crude instrument with plenty of play. When using it, you need to identify the play and work in such a way that you eliminate the play in the same direction for all measurements.

But in the end, wait till you've fitted a new, longer chain and let's see then. Try and use a quick link instead of the outrageously expensive Campag pin, so that you can go down in size without added cost. When you finally do use the pin, use a chain tool with special anvil for those pins which require peening.

 

[Sillyoldman]

Poor gear changes from a bent hanger presents in a strange way in that it allows perfect shifting in one direction and poor shifting in the reverse. You have not reported that, so that's not the primary problem. I'm still with the chain-too-short theory, but time will tell.

If you get the same hanger angle reading on a new hanger as on the old one, the frame is suspect. Have a very good look at the hanger recess on the frame itself. Use straight edge to see if it is flat and, loosen the hanger bold a bit and see if the hanger rocks in its cradle when you hold it firmly in place but try and rock it. It has to fit squarely without ambiguity. Before you start bending the frame, also check the hanger for flatness, sometimes they are a bit concave at the mating surface.

Keep in mind that the ParkTool DAG for the hanger measurement is a crude instrument with plenty of play. When using it, you need to identify the play and work in such a way that you eliminate the play in the same direction for all measurements.

But in the end, wait till you've fitted a new, longer chain and let's see then. Try and use a quick link instead of the outrageously expensive Campag pin, so that you can go down in size without added cost. When you finally do use the pin, use a chain tool with special anvil for those pins which require peening.

Yes the shifting both up and down the cassette on either chainring is superb so will await the new chain. Good idea re the split link until the length is finalised.
Agree about the Park tool DAG - it is a bit "wobbly" but I tried to mitigate the effect of the play as best I could.

cheers

 

[Graeme_FK]

Thank you for your comprehensive comment.
Please could you share the "exact requirements of Campagnolo" chains and how they differ from other (compromised) 11 speed chains, Graeme? I remember that there was something 'odd' about Campagnolo 10 speed but I thought Campagnolo had led the way (?in 2009) with 11 speeds and others had adopted the same chain specs:
Link = 1 inch (25.4mm)
Inner width (9-12 speed, between inner plates) = 11/128″ (2.18 mm)
11 speed outer width = 5.62mm

https://bike.bikegremlin.com/3555/bicycle-drive-chain-dimension-standards/

Leaving aside 'C-preciousness' and/or warranty (if any), why, in practice will a 'normal' (KMC say) 11 speed 'quick-link' not work entirely satisfactorily on a Campagnolo chain? What sort of "problems [do you see] quite frequently with [joining links]"? My experience is that riders/users joining a chain with a pins (eg CN-RE500 Ultra-link or Shimano equivalent) is more likely to result in a problem than joining it with a 'normal' quick-link.

Ummm, this is a pretty big subject.

Broadly, there are three defining differences that affect compatibility in modern chains.

Before we get to that, though, we have to understand what's meant when we talk about compatibility - because there is a difference between "will it work" and "will it work to specification"?

Even that isn't as simple as it looks from the outside because how well something works is a subjective judgement. I have retail customers who are very, very exacting in their requirements and who challenge me to get their bikes to work according to their subjective concept of how "well" is "well". Equally I have retail customers who assure me that their "shifting works fine" and I find that I have to downshift two and upshift one, or something of that nature, to select the gear required.

So when I refer to "working" I am generally thinking of "working to specification" - in the first example above, often those customers are looking for something that the system was never designed to deliver - say, front shift working within spec even where they are using incorrect chainring spacings, or a fully compatible system being asked to deliver good shifting at higher-than-designed-for applied torque loadings. In the second example - well, it should be clear that having to faff with shifting or treat it like it's made of glass, in this day and age, is definitely outside specification, even if the customer says they are happy with it.

Then, the defining characteristics are:
overall chain dimensions
specific shaping of chain components
tolerances in manufacture

Campagnolo launched 11s in 2009 and specified the overall outer link plate dimensions in terms of the over-plate and over-rivet width (bearing in mind that the rivets have a small protrusion outside of the link plate) and that dimension, very broadly, was picked up by Shimano and SRAM. They do have slight variations on it though. Third party chains also have small variations.
Campag over-outer link plate dimension is 5.25mm nominal, maximum internal pin protrusion (i.e. that "facing" the next-biggest sprocket) is 0.1mm. where the peened end should have that same maximum protrusion when the peening operation is correctly carried out, so an over-pin of 5.45mm.

At least as important in the context of joining links, though, is the exact over-inner link plate dimension and the tolerances that apply to it and the physical gauge of the outer link plates and the tolerances that apply there - more on which, later.

Link shaping matters because it is designed to do several things - confer sufficient strength (hence, we assume, SRAM's use of a "flat top" chain in the 12s system), to allow sufficient flexibility, to engage correctly with the tooth profiles that it will interface with and, in the case of outer links, to engage correctly with any ramping or pinning on the back of the big chainring. Here is one critical difference because there are differing methods used by the "Big 3" (or rather, the "Huge 1" & the "Smaller 2") to make that enagement happen. Campag's lift ramps are shaped specifically to pick up on the innermost edge of the outer plate. Different shaping (SRAM and Shimano) doesn't engage the ramp, so reducing the applied torque at which upshifting will occur & occasionally giving problems with the chain "riding" the tops of the outer ring teeth, momentarily. The shaping is also, in the case of KMC, for instance, problematic because their shaping tends to snag the small bumper plate that Campagnolo put inside the 10 & 11s FDs. In some cases, where users persistently upshift at very high torque values, testing in the lab and in the field has shown that this can strip the plastic plate away or, where it is bonded into place (as in EPS FDs) it can erode it to the point where it can no longer fulfil it's function.

Inner link shaping has to engage with the jockeys correctly to prevent the chain "falling off the side" of the lower jockey, to track correctly at an angle on the upper to facilitate over-shift and then smooth running, interface with the teeth of the sprockets and chainrings correctly to control noise. In the case of a chain made for a 1x system, it also needs to have correct engagement with a thick-thin chainring - that engagement is different to the engagement on a "normal" ring and is one reason why do-it-all chains are noisier on some multiple-chainring chainsets than are dedicated chains.

Manufacturing tolerances are of key importance. No manufacturing process is perfect and different makers will (as a function of cost and of functional design) allow a variation away from nominal dimensions. These manufacturing tolerances are usually amongst the more carefully guarded technical details. For example, Campagnolo's manufacturing tolerances, along with those of other companies I work with, all form part of the non-disclosure documents that I sign. This is done so that third parties will find it harder to make direct copies.

Where this becomes important, is that any maker of a joining link can only make assumptions about manufacturing tolerance. They can, of course, measure a great many chains and derive what they think is the likely average and make to that - but manufacturing tolerances don't always lie equally either side of a nominal value - they *might* be, something like 5 mm +/-0.05mm or, on the other hand, they might be 5 mm +0.1 / -0.00 mm and the third party would have no way to know, short of IP theft (and yes, it happens ...).

Where we have seen problems they are related to this and to another, related issue. I've got about 10 or 12 links here, all failed the same way, with one of the embedded pins torn out of the outer link plate. They were all supposedly "compatible" joining links from one manufacturer, all embedded into chains from another manufacturer (not all Campag / KMC). And these are just ones I have saved - I have seen other failures over the years. In most cases, I can surmise (but not prove as in the non-Campag cases I am as hamstrung as the link-maker) that this was a tolerance problem. The pins were just a fraction short, so whilst the link would "close" and was not over-stiff, repeated loading caused eventual failure.

OK, I see a lot of chains and I don't have a huge number of failed links - but, when one considers when a link is liable to fail (out of the saddle, loads of applied force), I'd rather know that the guy sprinting next to me was as safe as I can make him from a failure in my chain and I'd like to think he was, similarly, protecting my interests as well.

At a professional level, if I say "yes, that'll be OK" and it isn't, I can be held liable - so yes, I play safe.

The related issue is around wear and tear rates and materials. Another closely guarded secret is exact materials and expected wear rates and how that feeds into tolerance. Yes, materials can be chemically analysed but I have seen occurrences where joining links are in, or added into part-worn chains and that link no longer meshes correctly with the jockey wheels / and / or the sprockets / chainrings. I have one in my workshop now, in fact - a supposedly "correct" 10s joining link in a used Campag 10s chain that "ticks" where it fails to mesh correctly with the chainring under load. I've tried 3 different joining links, none of them mesh quite correctly. The chain wear is within tolerance. In this case, to cure the problem, as it's 10s Campag I can use an 10s UltraLink section in place of the joining link but in an 11s, I'd have a problem.

Chain joining has to be done right, for sure. It's a skill, you have to learn it - but if it's done correctly, the joined link is at least as strong as any other link in the chain (certainly in Campagnolo's case). I understand that it's easier just to flop a third party joining link in and 99.9% of the time, it'll be fine - but if I'd just paid out a couple of grand on a groupset, say, or even on a whole bike, I'd want it assembled by the book and working exactly to spec, now and once it had wear and tear on it, not a compromise solution that's easier for the mechanic ...

<Edit> I get the comment about C-preciousness ... I have worked with Campagnolo for 35 years now, on teams, in the field and in retail, in the service and warranty sphere (the last 12 years) and liaising with R and D. I've also worked with other makers including Shimano and SRAM. One thing I have learned, however, is that whilst all 3 have different approaches to what they do, is that they all have a good reason for it being the way it is. Campag are a much smaller company these days than Shimano or SRAM so they do tend to be more exacting, because they can't afford to throw huge amounts of money away in warranty or liability - they tend to be very cautious. They also have a very low tolerance of warranty acceptance. The 1st instance of any issue always generates an internal check (which is why I was interested in this case) and in many cases leads to what is called a "running change" - so if it looks like a design-based problem, once the issue is fully identified, it is often "seamlessly" corrected, even mid-season. If it's a problem found in the assembly information / advice, that is changed and the distribution network and key retailers are informed via the main Technical Service in each market (so in the UK, ourselves). </Edit>

 

[Ajax Bay]

@Graeme_FK I feel your pain ref subjective judgements by 'exacting' users and reasonable people. Many thanks for a most comprehensive answer.

8) Small ring to small sprocket - what is the space between the idler run of chain and the run of chain that passes around the "bottom" of the upper jockey wheel. It should be 8-15mm ideally, although 5 - 20mm is "unofficially" workable.
- I am on the limit at 20mm

This is indicative of a chain which is too short. The cage, in small/small, should 'wind up' more: enough to wrap at least one more 1" link.

View: https://www.youtube.com/watch?v=s6TodcAH948

 

[Graeme_FK]

Yes the shifting both up and down the cassette on either chainring is superb so will await the new chain. Good idea re the split link until the length is finalised.
Agree about the Park tool DAG - it is a bit "wobbly" but I tried to mitigate the effect of the play as best I could.

cheers

Unless they have changed the design, there used to be a screw that you could use to regulate the play.

I don't use the Park Tool in it's original form any more, I got annoyed with having to adjust the play & modified mine by line-boring it and installing a couple of sealed bearings, to eliminate any errors).

Campag hanger tolerance is +/-6mm at the rim (I've always considered that to be a bit "slack" but that is the published figure), so if you are at +/-5mm you "should" be OK but as Yellow Saddle says, the tool is not super-accurate and a couple of mm error either way is possible if that regulator isn't there any more.

Fore-aft twist is an important one for 11s systems - they are generally more susceptible to hanger error in this plane (than 10s systems, for example).

It's not that uncommon for hanger beds to be "off" and for the hanger to need to be gently and carefully "leaned on" to square it up. On one Team I worked with a few years ago, every frame delivered at the start of the year was "off" to the extent that we wedged all the hangers (to bend them fully into alignment was taking them further than I was comfortable with), as a temporary measure until the frame producers sorted the problem out.

Likewise, as Yellow Saddle says, hangers are made either by "cookie cut" (stamping) or machining out of solid. If the former, even out of the box / bag the hanger can be misaligned, if the latter, it's very unlikely. Off the top of my head, I can't remember what Enigma do.

In both cases, it's worth running an M10 x 1.0 cleaning tap through the thread - the Campag top binder is aluminium alloy and if you burr it in the thread on installation, it can either bind up or be hard to remove. Typically hangers are anodised or clear coated after the thread is cut and that reduces the thread clearance inside. The cleaning tap will correct that.

Don't set the chain length with a joining link! The only way that you can do that, is to cut out the link that carries the batch code for the chain - you can thereby invalidate your warranty ... In any case, the rivet drilling in the outer plates of the link assigned for joining is tested for exact compliance with the required standard when the chain is made. It's the only one that is. Use a pin, rivet and peen the correct way, with the correct tool, and test. Just start with the chain "long" rather than starting at the length you cropped the previous one to.

I suspect your issue *might* be chain length derived as if it's short, you will get the effect that Yellow Saddle referred to in his comment to me but potentially to a degree that is outside of the design parameters of the mech (which has a compensation for that effect built in) as the upper and lower spring tensions in the derailleur will be at higher-than-expected values.

Now that we are back from holiday I have checked back through the testing data (field and modelling) and we've not seen the behaviour that you describe before and in the UK, as far as I can see, we haven't warrantied for it, either.

I have put in a query to my colleagues at the factory to see if it's been reported in other markets.

 

[Yellow Saddle]

Ummm, this is a pretty big subject.

Broadly, there are three defining differences that affect compatibility in modern chains.

Before we get to that, though, we have to understand what's meant when we talk about compatibility - because there is a difference between "will it work" and "will it work to specification"?

Even that isn't as simple as it looks from the outside because how well something works is a subjective judgement. I have retail customers who are very, very exacting in their requirements and who challenge me to get their bikes to work according to their subjective concept of how "well" is "well". Equally I have retail customers who assure me that their "shifting works fine" and I find that I have to downshift two and upshift one, or something of that nature, to select the gear required.

So when I refer to "working" I am generally thinking of "working to specification" - in the first example above, often those customers are looking for something that the system was never designed to deliver - say, front shift working within spec even where they are using incorrect chainring spacings, or a fully compatible system being asked to deliver good shifting at higher-than-designed-for applied torque loadings. In the second example - well, it should be clear that having to faff with shifting or treat it like it's made of glass, in this day and age, is definitely outside specification, even if the customer says they are happy with it.

Then, the defining characteristics are:
overall chain dimensions
specific shaping of chain components
tolerances in manufacture

Campagnolo launched 11s in 2009 and specified the overall outer link plate dimensions in terms of the over-plate and over-rivet width (bearing in mind that the rivets have a small protrusion outside of the link plate) and that dimension, very broadly, was picked up by Shimano and SRAM. They do have slight variations on it though. Third party chains also have small variations.
Campag over-outer link plate dimension is 5.25mm nominal, maximum internal pin protrusion (i.e. that "facing" the next-biggest sprocket) is 0.1mm. where the peened end should have that same maximum protrusion when the peening operation is correctly carried out, so an over-pin of 5.45mm.

At least as important in the context of joining links, though, is the exact over-inner link plate dimension and the tolerances that apply to it and the physical gauge of the outer link plates and the tolerances that apply there - more on which, later.

Link shaping matters because it is designed to do several things - confer sufficient strength (hence, we assume, SRAM's use of a "flat top" chain in the 12s system), to allow sufficient flexibility, to engage correctly with the tooth profiles that it will interface with and, in the case of outer links, to engage correctly with any ramping or pinning on the back of the big chainring. Here is one critical difference because there are differing methods used by the "Big 3" (or rather, the "Huge 1" & the "Smaller 2") to make that enagement happen. Campag's lift ramps are shaped specifically to pick up on the innermost edge of the outer plate. Different shaping (SRAM and Shimano) doesn't engage the ramp, so reducing the applied torque at which upshifting will occur & occasionally giving problems with the chain "riding" the tops of the outer ring teeth, momentarily. The shaping is also, in the case of KMC, for instance, problematic because their shaping tends to snag the small bumper plate that Campagnolo put inside the 10 & 11s FDs. In some cases, where users persistently upshift at very high torque values, testing in the lab and in the field has shown that this can strip the plastic plate away or, where it is bonded into place (as in EPS FDs) it can erode it to the point where it can no longer fulfil it's function.

Inner link shaping has to engage with the jockeys correctly to prevent the chain "falling off the side" of the lower jockey, to track correctly at an angle on the upper to facilitate over-shift and then smooth running, interface with the teeth of the sprockets and chainrings correctly to control noise. In the case of a chain made for a 1x system, it also needs to have correct engagement with a thick-thin chainring - that engagement is different to the engagement on a "normal" ring and is one reason why do-it-all chains are noisier on some multiple-chainring chainsets than are dedicated chains.

Manufacturing tolerances are of key importance. No manufacturing process is perfect and different makers will (as a function of cost and of functional design) allow a variation away from nominal dimensions. These manufacturing tolerances are usually amongst the more carefully guarded technical details. For example, Campagnolo's manufacturing tolerances, along with those of other companies I work with, all form part of the non-disclosure documents that I sign. This is done so that third parties will find it harder to make direct copies.

Where this becomes important, is that any maker of a joining link can only make assumptions about manufacturing tolerance. They can, of course, measure a great many chains and derive what they think is the likely average and make to that - but manufacturing tolerances don't always lie equally either side of a nominal value - they *might* be, something like 5 mm +/-0.05mm or, on the other hand, they might be 5 mm +0.1 / -0.00 mm and the third party would have no way to know, short of IP theft (and yes, it happens ...).

Where we have seen problems they are related to this and to another, related issue. I've got about 10 or 12 links here, all failed the same way, with one of the embedded pins torn out of the outer link plate. They were all supposedly "compatible" joining links from one manufacturer, all embedded into chains from another manufacturer (not all Campag / KMC). And these are just ones I have saved - I have seen other failures over the years. In most cases, I can surmise (but not prove as in the non-Campag cases I am as hamstrung as the link-maker) that this was a tolerance problem. The pins were just a fraction short, so whilst the link would "close" and was not over-stiff, repeated loading caused eventual failure.

OK, I see a lot of chains and I don't have a huge number of failed links - but, when one considers when a link is liable to fail (out of the saddle, loads of applied force), I'd rather know that the guy sprinting next to me was as safe as I can make him from a failure in my chainand I'd like to think he was protecting my interests as well. At a professional level, if I say "yes, that'll be OK" and it isn't, I can be held liable - so yes, I play safe.

The related issue is around wear and tear rates and materials. Another closely guarded secret is exact materials and expected wear rates and how that feeds into tolerance. Yes, materials can be chemically analysed but I have seen occurrences where joining links are in, or added into part-worn chains and that link no longer meshes correctly with the jockey wheels / and / or the sprockets / chainrings. I have one in my workshop now, in fact - a supposedly "correct" 10s joining link in a used Campag 10s chain that "ticks" where it fails to mesh correctly with the chainring under load. I've tried 3 different joining links, none of them mesh quite correctly. The chain wear is within tolerance. In this case, to cure the problem, as it's 10s Campag I can use an 10s UltraLink section in place of the joining link but in an 11s, I'd have a problem.

Chain joining has to be done right, for sure. It's a skill, you have to learn it - but if it's done correctly, the joined link is at least as strong as any other link in the chain (certainly in Campagnolo's case). I understand that it's easier just to flop a third party joining link in and 99.9% of the time, it'll be fine - but if I'd just paid out a couple of grand on a groupset, say, or even on a whole bike, I'd want it assembled by the book and working exactly to spec, now and once it had wear and tear on it, not a compromise solution that's easier for the mechanic ...

Graeme, whilst I appreciate it when someone stands behind a brand, I also recognise when someone has been taken in by branding nonsense and turned it into brand religion. The last time I've read so much hyperbole was in a hi-fi magazine in the 1970s.

 

[Graeme_FK]

Graeme, whilst I appreciate it when someone stands behind a brand, I also recognise when someone has been taken in by branding nonsense and turned it into brand religion. The last time I've read so much hyperbole was in a hi-fi magazine in the 1970s.

It's not that at all, although I can 100% see why you might read it that way. I'm quite happy to admit when a company screws up, be it Campagnolo or anyone else. And they do.

I'm not employed directly by Campagnolo (which is why I am free to work in technical consultancy with other companies) and I've been doing this stuff a very long time and so have worked with Campagnolo for a long time - but also with Shimano, SRAM and a number of others, some of whom don't even exist any more.

I like things to be right. Not "about right", "more or less right" or "that'll do, right". I like them to be "right" in the sense of, this is the way that Campag, or Shimano, or SRAM or whoever say it should be, so that's what we'll do.

Had the question been about Shimano, or SRAM and were I up-to-date enough, I'd have answered in the same, slightly nit-picky, way.

One of my main faults is that I also like to understand the "why" when I am looking at an issue.
It was my favourite question as a kid :-D

So when people ask about an issue, I try to be as exact as I can be. Apart from anything else, I have seen enough issues caused by people being blase about technical specs (I've been an expert witness on both sides of liability cases in court and that's never a nice position to be in) that I'd rather not be there as a defendant myself - so I rarely take professional matters at face value.

It won't come as any surprise to you that my hobbies, outside my profession, are photography and yes, sound reproduction / hifi.

 

[Sillyoldman]

Thanks again everyone.

Graeme:
1) Campag hanger tolerance is +/-6mm at the rim (I've always considered that to be a bit "slack" but that is the published figure), so if you are at +/-5mm you "should" be OK but as Yellow Saddle says, the tool is not super-accurate and a couple of mm error either way is possible if that regulator isn't there any more.
-the regulator is still there and I am pretty happy I was getting consistent results measuring many times. I saw that 6 mm +/- in the tech docs after I posted earlier so currently plan to leave this alone at least for now until the new chain is on and I have some miles on the bike again - must admit I aren't too keen to start bending metal!

2) Fore-aft twist is an important one for 11s systems - they are generally more susceptible to hanger error in this plane (than 10s systems, for example).
- There was no real difference (about 1mm) for the fore/aft

3) It's not that uncommon for hanger beds to be "off" and for the hanger to need to be gently and carefully "leaned on" to square it up. On one Team I worked with a few years ago, every frame delivered at the start of the year was "off" to the extent that we wedged all the hangers (to bend them fully into alignment was taking them further than I was comfortable with), as a temporary measure until the frame producers sorted the problem out.
- as in point one this is scary to me so I hope the new chain fixes my issue

4) Likewise, as Yellow Saddle says, hangers are made either by "cookie cut" (stamping) or machining out of solid. If the former, even out of the box / bag the hanger can be misaligned, if the latter, it's very unlikely. Off the top of my head, I can't remember what Enigma do.
- they are machined from solid.

5) In both cases, it's worth running an M10 x 1.0 cleaning tap through the thread - the Campag top binder is aluminium alloy and if you burr it in the thread on installation, it can either bind up or be hard to remove. Typically hangers are anodised or clear coated after the thread is cut and that reduces the thread clearance inside. The cleaning tap will correct that.
- didn't do this but it went back easily so all ok

6) Don't set the chain length with a joining link! The only way that you can do that, is to cut out the link that carries the batch code for the chain - you can thereby invalidate your warranty ... In any case, the rivet drilling in the outer plates of the link assigned for joining is tested for exact compliance with the required standard when the chain is made. It's the only one that is. Use a pin, rivet and peen the correct way, with the correct tool, and test. Just start with the chain "long" rather than starting at the length you cropped the previous one to.
- That's a good point. Is it possible to put the pin in temporarily (i.e not fully insert it) until the idler run of chain and the run of chain that passes around the "bottom" of the upper jockey wheel is into the 8-15 mm range - which should be one full link?

 

[Ajax Bay]

the idler run of chain and the run of chain that passes around the "bottom" of the upper jockey wheel is into the 8-15 mm range

Watch the 5 minute video to which I linked: the method of getting the chain length correct ('within spec') is crystal clear. Felt tip pen required.