Chain length for a compact

henboy

Hi folks

I have a compact on one of my bikes and I have a 27 -12 sprocket on the back . I'm fitting a knew chain just wondering about how to figure the chain length . By the way the chain is Shimano ultegra 6700 10 speed.

I've been on the Shimano website but not to sure about what length with a compact . Thanks

blorg

There are a number of methods . If you are not changing anything else on the bike easiest just to cut the new one to the same length as the old one. I have also used the largest chainring and sprocket method. There is nothing peculiar about a compact in this regard (if you were switching to a compact from a standard you would probably want to remove two links.)

D!armu!d

If you went for the largest chainring & sprocket measurement would this mean you couldn't switch to a larger sprocket in the future?
Or would it affect how the chain ran if you left any slack in it to cover for this?

henboy

D!armu!d wrote: »

If you went for the largest chainring & sprocket measurement would this mean you couldn't switch to a larger sprocket in the future?
Or would it affect how the chain ran if you left any slack in it to cover for this?

I wouldn't imagine you could switch to a larger sprocket once you measure and cut it . (I could be wrong ). I'll do the method of large chainring and large sprocket and just add 2 links. With a compact and a 27 on the back I would be happy with that on the spare bike .

blorg

D!armu!d wrote: »

If you went for the largest chainring & sprocket measurement would this mean you couldn't switch to a larger sprocket in the future?
Or would it affect how the chain ran if you left any slack in it to cover for this?

Yes, if you measure it for chainring and sprocket you can't then go any bigger. You might get away with a tooth but I know a bike I measured with a 23 cassette I couldn't then put a 27 on. So measure for the largest combo you are going to use on the bike in question.

If the chain is too long you can have slack in it when in the small small (not that you should be) and a higher risk of the chain coming off. You should be able to leave slack for a bigger cassette without this being a problem but generally the shorter the chain the better the shifting.

Single Malt

blorg wrote: »

Yes, if you measure it for chainring and sprocket you can't then go any bigger. You might get away with a tooth but I know a bike I measured with a 23 cassette I couldn't then put a 27 on. So measure for the largest combo you are going to use on the bike in question.

If the chain is too long you can have slack in it when in the small small (not that you should be) and a higher risk of the chain coming off. You should be able to leave slack for a bigger cassette without this being a problem but generally the shorter the chain the better the shifting.

+1 for shimano. Here is the guide to chain length for shimano http://techdocs.shimano.com/media/techdocs/content/cycle/SI/Dura-Ace/SI_5X00A_002/SI-5X00A-003-ENG_v1_m56577569830700387.pdf

doozerie

D!armu!d wrote: »

If you went for the largest chainring & sprocket measurement would this mean you couldn't switch to a larger sprocket in the future?
Or would it affect how the chain ran if you left any slack in it to cover for this?

I don't think there is a clear answer to whether you could switch to a larger sprocket or not, I think a qualified "maybe" is the only answer. The "problem" is that the smaller chainring is not factored in at all when measuring a chain in this way. The method presumably assumes that your smaller chainring is not radically smaller than your large chainring, but these days that assumption seems a less safe bet when you have have compacts with a 50/34 as standard (which is a bigger jump than the much older/traditional 52/42, or the more recent 53/39 or 52/39). So for a compact I think there is a greater likelihood of being able to go for a larger sprocket later, though the only way to know for sure is to try it unfortunately (or do some creative working out with a measuring tape before breaking the chain).

As for whether extra slack in the chain will be a problem, that too depends. One of the things it depends on is whether you have a long cage or short cage rear derailleur. For a road bike with a double crankset you'd expect it to be a short cage, but it's worth double-checking. A long cage will take up more of the chain slack so minimises the risk of problems, but is usually unnecessary unless you have a triple crankset. Generally speaking, a slack chain can make for poorer shifting, and is more prone to falling off - it can also try to take chunks out of the paintwork on your chainstay on a particularly bumpy road too if there is a lot of slack.

blorg

The large chainring/large sprocket method of determining chain length basically gives you the minimum chain length that will still allow you to shift into large/large. There will be a slight variance between bikes depending on the length of the bike's chainstay. The small chainring isn't relevant. Large-large isn't a combo you should be in, in particular, but you probably want a chain long enough that you can shift into it without breaking anything (in practice in my experience the shift just doesn't work if your chain is too short.) Others describe issues involving derailleurs shearing off, maybe, I have never seen it. If your chain is sized according to this method you won't be able to go much bigger at the back. It is of course entirely possible that your chain is longer already than it needs to be for the back in which case you could.

If you have a chain sized for 53/23 as the largest pair, you can certainly change the cassette to a 27 and go up to that 27 in the small ring. Just not in the large one. Incidentally I have found I can swap between 53/39 11-23 and 50/34 12-27 with the same length chain as the reduction at front compensates for the increase at the back.

In my own experience I have a short cage with 50/34 12-27 (and indeed 11-28) on a road bike but the same combo gave too much slack and risk of chain derailment on my cross bike so when I inevitably had to change the RD I went for a long cage there instead to take up the slack.

henboy

I put on the knew chain to day went for the method of large chainring to large sprocket. Just added the 2 links to take off and it worked a treat (hopefully). Have to say chains are a bit tricky , need to invest in a good chain splitter .

Hopefully thats it with chains for awhile . Thanks folks

doozerie

blorg wrote:

The large chainring/large sprocket method of determining chain length basically gives you the minimum chain length that will still allow you to shift into large/large. There will be a slight variance between bikes depending on the length of the bike's chainstay. The small chainring isn't relevant.

You're right, my reference to potential issues on the smaller chainring is a different issue which relates to a different commonly-used method (below) of measuring chain length - I referred to one measurement method and wrote of the potential issues of another in a garbled mess.

I disagree that the small chainring isn't relevant, but it's relevance does depend on the method used to measure the chain:

* If you measure using the large chainring to largest sprocket method, then because this doesn't take account of the smaller chainring or smallest sprocket there is a risk that you'll have "too much" chain slack in small-small. How much chain slack is "too much" is debatable though. And yes, you should avoid small-small.

* If you measure using the large chainring to smallest sprocket method, wrapping the chain through the jockey wheels and sizing when the jockey wheels are vertically aligned, then you risk having a chain which is too short for large-large. Again, yes you should avoid large-large (ironically, both extremes of chain crossing probably occur most often in races, where circumstances don't always allow riders much time to focus on chain crossing, and probably occur under the worst circumstances where a lot of force is being applied to the pedals).

In reality, the usual methods for measuring chain length seem to have a lot of flexibility built in so that perhaps only in extreme cases will they yield a chain that is either too long or too short. I used the second method above for years, for both road bikes and MTB's and only on a small number of occasions had a problem that I can recall. In recent years I have been using the first method above and have had no problems at all (though I've been using relatively small cassette ranges in that time). Of the two I suspect that the first method gives a shorter chain, which is my preference, but I've never actually compared the results of both so they might be closer than I think.

None of this is particularly relevant to the original question on this thread of course, it's really just me musing on component design and fit. I'm often surprised at the arbitrary nature of some of the stuff relating to bikes - despite the huge amount of money that bike and component manufacturers plough into research, devising elaborate methods and designs to shave seconds or fractions of second off performance, things like how long to make your chain seem to fall into the category of "yerra, whatever" (and things like fitting headset races and square taper cranks fall into the category of "just apply more brute force"). The current 11-speed Campag Record chain is one example of what seems to me to be an odd mix of extreme engineering and a very casual approach to sizing - each pin is hollow to save weight, the chain is described by Campag as being so finely engineered that they will only stand over a fitting of it that uses their new (and very expensive) 11-speed chain tool, you are not supposed to re-join the chain more than twice as otherwise Campag wash their hands of any failure that arises, etc. They've clearly gone to great lengths to save weight on the chain, to the extent that it apparently can't be split and savely rejoined more than twice, yet their method of how to size the chain is the second one above (well, they throw in a hint of accuracy/science as instead of sizing according to when the jockey wheels are vertically aligned by eye, you size according to the measured distance of chain to chain around the lower jockey wheel).

Perhaps the above methods are as accurate as it is possible to be when determining optimum chain length, but I do wonder whether the mechanics on pro teams actually use these methods or whether they use some other method that is more specific to the range of gears on the individual bikes they are working on. I, for one, won't sleep until such important questions are answered...

blorg

doozerie wrote: »

Perhaps the above methods are as accurate as it is possible to be when determining optimum chain length, but I do wonder whether the mechanics on pro teams actually use these methods or whether they use some other method that is more specific to the range of gears on the individual bikes they are working on. I, for one, won't sleep until such important questions are answered...

doozerie

That equation is still based on the large front and large rear so it doesn't account for the jump between the chainrings or the spread of the sprockets any more than the other methods. It does look more scientific though, particularly when shouted, which is all important of course.

seamus

While we're on the subject of compacts, I'd like to give myself more range at the high level (i.e more speed) without sacrificing too much.
I'm currently running a 105 groupset 34/50 at the front and 12-27 at the back. I've seen tonnes of stuff about what you can and cannot run, but it's all a bit insane.

Ideally I think a running 36/52 (or 54) at the front and 11-28 (or 30) at the back would give me better gearing without causing much hassle on heavy climbs. Any ideas? I've heard that short-cage mechs cause sprocket size to be limited.

blorg

seamus wrote: »

While we're on the subject of compacts, I'd like to give myself more range at the high level (i.e more speed) without sacrificing too much.

You would be better off just working on your cadence; and descending on aero tuck and cornering. If you think about it, pro riders do not use gearing fundamentally different than amateurs (largest gear generally used outside TTs is 53-11, and they do actually often use 53-12.) They are however much faster. How do they do this on the same gearing? Cadence (and of course the power to back it up.)

Apart from that if you really feel you need it 1 tooth on the back makes far more of a difference than what you do on the front. New cassettes are generally also cheaper. The only issue is that while 12-27 I think is a decent climbing cassette with sensible gaps the new Ultegra 11-28 is a bit of a mess and very gappy at the climbing end. AFAIK there is no 11-27 available (other than new Dura-Ace at crazy money) so you are looking at 11-25 and losing two teeth for climbing. *EDIT: SRAM actually has a 11-26 which might be worth a try. SRAM cassettes are Shimano compatible.

28t is the limit for a Shimano road RD, doesn't matter whether it is short or long cage, they both have 28 as the limit. Long can just handle more chain- it is theoretically only needed for a triple but can be useful for example on a cross bike with a wide gear range due to the increased likelihood of the chain coming off. Technically of course you don't need a wide gear range for cross but that is what I had on the bike.

Again, I would suggest what you have is pretty optimal (I use 50/34 12-27 myself on most of my bikes other than my race bike) and there is no particular need to change it.

doozerie

@seamus, One physical constraint that you should check on if you are planning on changing the chainrings is whether you can get a 52-tooth chainring to fit your compact crankset. As far as I know, 50 is the largest size for a compact though I haven't checked that to confirm. If so, then you'd have to change your entire crankset to a standard double to get a 52. You might also have to change your front mech too.

A 54 chainring would be huge, you'd need a lot of power to push that.

TinyExplosions

1 tooth at the back is roughly (very roughly) equivalent to 3 at the front

seamus

Cool thanks for the info. Probably no point in changing what I have then for a while, if at all.

blorg

You can get 52 tooth chainrings to fit a 110 (compact) BCD. They are a slight speciality item but they do exist; FSA and SRAM make them. You don't have to change your front derailleur just move it up a bit; the same goes if changing the crankset to a standard (130 BCD.) What you do need to bear in mind though is that road front derailleurs generally have a capacity of 16T so 50/34 is already at the limit. As such if changing the outer ring to 52T the minimum you would need to up the inner to at least 36T.

Triple front derailleurs have a larger capacity to deal with the granny ring but even there I don't think they would handle over 16T in one shift. It is the upshift that is the problem. I vaguely recall a pro team setting up bikes for the Vuelta on a stage that was finishing up L'Angliru with an excessive gap; they had them set up so that they would shift down into the small ring but not back up again.

As doozerie says 54T is nuts although amazingly even that exists in a 110 BCD. A given BCD only restricts how small a ring you can put on, not how large. The potential issue you get with very large rings on a compact spider is chainring flex.

Again though, what you have now is probably optimal.

Lumen

blorg wrote: »

I vaguely recall a pro team setting up bikes for the Vuelta on a stage that was finishing up L'Angliru with an excessive gap; they had them set up so that they would shift down into the small ring but not back up again.

:eek: