ThermionicScott 

A "gravel bike" is a bike you ride on gravel.




(Ray Hosler photos.)

Kapusta

Critical? Somehow I had no issues mountain biking for 20 years without a clutch. But they are "critical" for cobblestones?

Badger6

Mountain biking (XC is all I ever did, so I've got a limited view) and proper Belgian cobblestones are incomparable, and it is clear after being here now for a year that the clutched RD was a good call. Overkill in Germany, but not here. Come on over, I'll be happy to go ride with you.

jonathanf2

I just finished building out my gravel-inspired road bike. It won't handle the really chunky stuff (I'll reserve that for my GRX equipped alloy bike), but it should handle the poor roads of Los Angeles and mountain paved/hard packed fire roads. I initially had 38c GK slicks, but I find 32c GK semi-slicks to offer better grip on loose dirt, while having more responsive acceleration. It has Ritchey Comp flared drops and a 130mm stem. It's running a compact 50/34t crankset and 11 speed 11-32t cassette. Gearing-wise it's similar enough in low gear at 34:32t to 30:28t on my GRX bike which I normally use on climbs. The only difference is that I won't have the bail out low gears, but I gain faster top speed with 50:11 in high gear which I wanted mainly for fast group rides. Later on I might add a bigger cassette, but 32t seems to work well for me.

I guess this is what I meant by gravel inspired road bike! I mean it could still be a light gravel bike or a road bike with grippy tires?

thehammerdog

bianchi megatube Ti. 32mm

road bike with canti's 35mm is max tire size

Fentuz

the cassettes are much bigger now. Back in the 90’s they were about 30/32T now most 1x are running 42 or more. They require longer cage, chain length which increase the mechanical advantage of the whip hence the clutch.
if you look @ current 2x running up to 34T on short or medium cage, they are still spring loaded only.

Germany_chris

Smaller chainrings and a tighter cassette means no need for a clutch

Kapusta

I believe some of this is a little off, and maybe not entirely relevant to the OPs situation, anyway

First: the larger cassettes of 1x does NOT mean they have more chain to control. The total length of the chain is not the relevant factor, here (and with the smaller front rings, I don't know that the chains are longer, anyway). The needed chain wrap capacity is what counts, because that is what determines how long of a RD cage you need, as well as how long the chain will get between the chainring and the lower pulley. And this is not really any bigger for 1x than for typical 2x or old 3x setups. 1x MTB setups can vary from 11-42 (31twrap) to 10-51 (41t). looking at a 3x8 or 3x9 system from back in the days long for clutches, a 22/32/44 x 11-32 setup had a 43t wrap. A more modern 2x10 mtb setup that was common when clutches first came out would have been something like 26/39 x 11-36 (38t wrap). What few 2x system you do see on newer MTBs DO use a clutch, and are typically require even higher wrap capacities.

Second, the clutchless 2x systems you are referring to with 11-34 cassettes are going to be on Road bikes, not MTBs. THAT is why there has not been a push to get clutches on them. It is definitely NOT due to needing a shorter cage. A 34/50 x 11-34 2x road setup is still a 39t wrap.

In other words, it is the intended use, NOT the chain wrap, that drives the need for the clutch.

One caveat to this: 1x does specifically benefit from a clutch in a way that 2x and 3x do not: chain retention. A clutch helps prevent trowing a chain on a 1x setup that does not have a chain guide. Though modern 1x setups also have NW rings which already helps a lot in this regard.

But the above is not that relevant unless the OP is running 1X, which I do not believe he is.

The only reason for the OP to be needing a clutch is if he is encountering terrain mare akin to mountain biking than road biking, which according to him, he is. So he is looking at one for the right reasons. I still question if he actually NEEDS it, as even mountain bikes with FDs never actually NEEDED it, but it is certainly nice to have to keep the chain from slapping around and can eliminate the need for a chain-stay protector. I am not saying he should not get one or that there is not some benefit to having it, I just don't think he actually NEEDS it for chain retention. New gravel group-sets are coming out with clutches, and it is a nice option, but again I would say they are not really NEEDED. On my gravel bike I personally have not even considered replacing my RD just to gain a clutch. Heck I never even did for my MTB (but I have one now since I went 1x11).

jonathanf2

I'm using a 105 5800 rd on this bike build and on the first few outings, when hitting the bumpy roads the chain slap hasn't been too bad. I put on a neoprene chain-stay protector for added protection, but after riding on rough paved fire roads I think the 105 road derailleur should be fine. I think the clutch on my GRX equipped bike is definitely needed since I do ride that bike on single track and bumpy dirt fire roads. I do have a bad habit of wanting to jump off curbs and bunny hills on occasion, but I'll try to refrain on this particular bike.

Fentuz

so you wrote a very long reply without having read properly what I said… if you have a longer cage, you get greater torque generated by the chain whip than the same whip on a short cage… so as you mentioned , chain wrapping is important and linked to the tension of the chain. And clutch-less long cage will not ensure the adequate tension on bumpy terrain…
on short and medium cage they are not needed and are not on 2x gravel (2x9 is sora by default) or 3x9 mtb (altus/acera).

long cage is required for 40+T… hence the clutch

note that shimano introduces clutch in mtb derailleur from deore range (max cassette 42)
altus, acera, alivio don t and are up to 36.

Kapusta

I misread you? Were you not saying: Larger cassette cogs on modern MTBS mean you need longer cages.

If so, this is incorrect.

And from what I can tell, the reason you think this is based on a misunderstanding about how chain wrap capacity (which determines cage length) is calculated. Here is how you determine how much chain wrap capacity you need: (Big cog - Little cog) + (Big ring - Little ring)

While a larger range cassette requires more chain wrap, so does having multiple rings up front. And fewer rings mean less wrap capacity needed.

This is why my 1x setup 11-46 cassette uses a medium cage length, while a 3x9 with 22/32/44 x 11-32 uses a long cage. That's right, much bigger cassette, less chain wrap needed. Do the math and it makes perfect sense:
For a 3x9 with 22/32/44 rings and 11-32 cassette: (32t - 11t) + (44t - 22t) = 43t
for a 1x11 with a 32t ring and 11-46 cassette: (46t - 11t) + (32t - 32t) = 35t

So when you say that modern mtbs with wider range cassettes need longer cages, you are mistaken. The move to 1x offset the need for additional chain wrap capacity.

The reason low end mtb RDs (like acera and altus) don’t have clutches is because they are low end RDs. Higher end 2x mtb setups DO have clutches (or at least they did while 2x was still a thing).

And as a history lesson, shimano introduced clutches back when the largest cogs they made were 36t.

Fentuz

Your chain capacity argument is OK however, you fail to look at maximum sprocket limitation... short cage tend to be limited with 28t , medium 36t and large is 42 to 52 depending of the model... Hence Sram rival and force long cage being used...
sram medium tend to be limited to 36T max sprocket and 37T
sram medium tend to be limited to 42T max sprocket and 43/45T
so if considering capacity only, on 1x, you should use medium cage with a sram 11-42 cassette (31 capacity) but yet Sram clearly advise to use a long cage et manufacturer are using long cage when above 40T max...

Again, it is not my understanding failing, it is manufacturer specification and advise. And yes, for a given force applied on the bottom pulley, a long cage (mine is ~100mm between axis on my rival long)) will generate more torque to the rotation axles than a short (60mm on my sora short) or medium cage (80mm on my sora medium). So, you need a "better" reaction load for a long cage than a short and medium cage to ensure the same level of tension. Increasing the spring rate is not ideal as the extra load will generate more frictions. So the next best thing is a dampening effect in this instance a clutch...

so ready, you are most likely going against manufacturer advice using a 11-46T with a medium cage... which is probably going to work because manufacturers apply at least 25 to 50% safety factors by "detuning". and again, it may be fine (I do it on my kids bike) but because it works for you, it does not mean it is correct.

Kapusta

You are making arguments that have almost nothing to do with backing up your original point. Do you even know what your point was anymore?

Let's back up. I had stated
To which you responded
....and this is simply not true. They DO NOT require longer cages than they used to for typical setups. Typical MTB setups in the 90s ran med and long cages, just like they do now, because the chain wrap needed has not changed much. If anything is is typically LESS now (as I pointed out in an earlier post).

Why you are now dragging Road bike components into this to make a point about MTB RDs, I don't really know. But they don't back up your point anyway.

Take the SRAM models you just mentioned. Many SRAM and Shimano RDs come in varying cage lengths. Are there any models you are aware of for which the long cage has a clutch, but the medium does not? No, because that is not the driver for clutches. The reason your Soras don't have clutches is because the are lower end road components, NOT because they are short/med cage. RIval you mentioned is clutched for BOTH long and med cage.

So, you want to know what manufacturers think? Go look through Road and MTB RDs. Again, do you see any models where clutch vs-non-clutch changes with cage length? No, Clutch vs no-clutch is based on application and price point (thus model), NOT on cage length.

Maybe a longer cage does benefit slightly more from a clutch than a medium cage, but that is irrelevant to the original point here, as there have been long caged clutchless RDs all along. And furthermore, no manufacturer bases clutch decisions on cage length.

Yes, in SOME cases (not all) max cog size goes up with max wrap capacity. But you are confusing (rather loose) correlation with causation, here, and it completely breaks down when you start comparing new MTB RDs and drivetrains to old ones (which is what you were doing in your original response to me).

Really? Here is my RD in Long and Medium Cages. Exact same max cog size (which is why I call the max cog to wrap capacity correlation "loose"). And BOTH have clutches. And contrary to what you assumed, I am using this as directed.

Anyway, this entire sidebar debate is utterly irrelevant to the OP, and I am not interested in continuing it much longer. I am trying to help you understand a couple misconception you seem to be under (that modern MTBs need longer cages than old ones, or that manufacturers only put clutches on long cage RDs), but I've already put more time into this than I should have.

Fentuz

My mid-2000 x9 long cage derailleur that came with 11-34T (rated 36 max with 45 capacity) is 80mm long... which would be a medium cage nowadays.
then, yes, the chain length will be different... the difference of chain stay between a road bike and gravel bike is 4 to 8" . Since the original question is about making a road bike from a gravel, 4 to 8" in chiantray is relevant to the chain length and will affect chain whip on bumps.

at this point, you are clearly a licenced professional mechanical engineer and will not accept that great larger modern sprocket requires longer cage (longer than 20 year ago) which will create more torque due to mechanical advantage...

clearly you know better... and better than manufacturer

Kapusta

Yes, that would be considered a Med cage now, but a medium cage is what most 1x systems have actually been using. Like all three of my 1x11 setups. The label is not the important part here, the actual length is.

Here are some numbers for you from a bunch of my RDs: (I'll grant that my measurements could be off my a mm or two as these are all on bikes)

First lets look at two old RDs I have from back in the day when 34 was the largest cogs Shimano offered that were used on 3x9 setups. These were the standard at the time.
LX: 85mm
XT: 91mm

Now lets look at two newer XT and SLX, RDs, both run 1x11 with 46t cogs (which, it you read Shimano's specs are completely per manufacturer's recommendations)
SLX: 92mm
XT: 90mm

And lastly, an NX Rd running a 42 t large cog (again, within recommended specs)
NX: 75mm

MUCH larger cogs, no real difference in cage length (the NX is actually a bit shorter, but it can be hard to compare SRAM vs Shimano). These are the real world numbers of what is and was commonly used.

Further, this makes perfect sense, because the needed wrap capacity did not increase.

I will grant one thing: that very longest RD cage options have gotten longer than in the past, in order to accommodate wide range cassettes that are also used with multiple rings, so you see RDs with 47t wrap capacities. But in practice, these are rarely used, because MTBs have gone almost entirely 1x. Only the very widest range 1x12 systems (like 10-51 or 11-52 cassettes) actually need a longer cage than MTBs used to run. And these are a fairly recent development. However, the reason for the long cage is NOT the size of the cog. It is, like I have been repeating at nausea, the WRAP CAPACITY. A 10-51 is 41t of chain to wrap.

I will repeat an important point: Shimano came out with the clutch when the biggest cassette they made was 11-36. The development of the clutche had nothing to do with huge cassette cogs.

And finally... what on earth does your point even have to do with my comment to the OP? He is not using a wide range MTB cassette or RD. Your whole argument is actually irrelevant.