# Can a double rear derailleur work?

I was wondering instead of a long cage rear dearailleur (RD), what would happen if 2 rear short cage RDs were used instead? I realize the 2nd one might cause some mounting issues but assuming there was a proper bracket to hold it in place, could they somehow be synchronized to pad out or take up similar amounts of chain but about half each so they don't need to angle as much as a single? For example, if the same cable passed to both of them but there were calibrated to shift one rear cog each laterally, but individually would only take up or pad out enough chain for half a shift, so that combined they would make a complete shift but with half the travel/deflection or whatever the proper term is when it moves. If they were both identical RDs (with same spring tension), would they automatically share the chain takeup/padout equally (half each) or would the system be biased more towards one than the other?

I am not suggesting this is a good thing to try. I am just asking if it is possible and what problems may occur as a result and if it would actually work.

• I have serious doubts whether you can get it to work. Might be good in theory, but terrible in practice. Easier to simply modify the existing derailer to have a longer cage (and stiffer spring). Or, if ground clearance is the issue, simply use a 3-pully system like the Suntour (which was one of the best ever made but couldn't get much traction in the Spandex-and-shaved-legs crowd). May 7, 2016 at 12:15
• What does that mean good in theory but terrible in practice? If it doesn't work in reality how can that be good in theory? May 7, 2016 at 12:29
• I mean that you might have an elegant scheme for doing this, with all the math worked out, but it will keep tying itself in knots when you try it. May 7, 2016 at 12:32
• (And it's unclear what problem you're trying to solve with this.) May 7, 2016 at 12:33
• How do you know for sure it will not work if you haven't yet tried it? What would cause it to fail? Assume both RDs are identical. They share a cable. They move in tandem. Where is the problem? May 7, 2016 at 12:33

The reason why the RD has that bend in it is because it is what tension the chain in every single cog in the cassete. Also if it does somehow work. The length of your chain will increase, because it will have 2 cages to go through.

• A small "price to pay" if it works. I wonder what it would do to the largest cog limit (if it would change) and the "spread" (# of teeth it can change from largest to smallest cog). May 7, 2016 at 12:01
• The price is small, but what is the benefit?
– ojs
May 7, 2016 at 12:12
• Improved ground clearance, bragging rights, more range of cogs (perhaps something like 40 to 10 or even 50 to 10), cool looking, innovative... May 7, 2016 at 12:38
• Also, perhaps the 2nd RD could be tweaked to perform its function better and would be slightly different than the main RD. Perhaps some manufacturer could design a bike where the pair of RDs work "perfectly" (or at least very well) together and would have individual adjustments to sync them. It seems like this could be made to work and if it doubles the tooth range of cogs, it would be a marketable item. There may also be some other benefits I am not yet aware of. May 7, 2016 at 12:48
• Well if you are looking for bragging rights then you've got one if it works. But I really don't see the sense of this, taking into account your problem of ground clerance, the easiest way to solve that is to add more links to your chain. Range of cogs? do you think that there already a 50 tooth ring for the back? The range will be awful too. 10-50 even in a 12x system. That's an awful lot of jumps in between each cog. So I stand by what I said, a long cage derailure is a better way to solve your problem. May 7, 2016 at 13:42

I don't think this works for several reasons. First, the top cog on the first RD would have a much higher tension on it than designed as the chain comes up from the tension cog on the first RD and then back down towards the tension cog on second RD.
The other issue is the indexing. Proper indexing requires that the top cog is close to the cassette. For example, think of what happens when the B screw is adjusted too far away from the cassette. You get lousy shifting. Slow, chain jumps etc. So that said, having 1/2 the index taken up by each RD is pointless. The only one that matters is the one closest to the cassette. If you are looking for something that can take up more chain in a tighter area, it might better to go for a single indexing RD and the second RD would really just be a couple of cogs that take up chain slack.

• Why not just have a single RD then and a 2nd "half" HD which is only the part where the 2 roller guide sprockets are. So there would be 2 spring mechanisms that each would take up half of the chain slack and would move parallel to each other (at the same angle). The rear RD would function as normal but the front RD has only 1 purpose which is to take up or pad out half of the require chain slack for each gear. I think this could be made to work. May 7, 2016 at 14:04
• Yes, exactly my point in the last sentence of my answer.... May 7, 2016 at 23:38