# How does a RD (rear derailleur) handle variable tooth gaps?

I am trying to better understand how a rear derailleur (RD) works. I can understand how the RD moves inward and outward but suppose someone changes several of their cogs on the cassette. Let's just concentrate on the 3 largest cogs for this question. Let's suppose they were originally 36, 32, and 29. Next suppose the person wanted to change those to 38, 32, and 28. So we just increased the gaps by 2 teeth and 1 tooth respectively. So my question is now how does the RD "know" that the adjacent cogs to the 32 (which remained unchanged) now has wider tooth gaps and how does it handle it? A more general question is how does the RD "know" how much bigger or smaller the neighboring cog is? What I am getting at is the RD doesn't have any programming in it to tell it the gaps so how does it handle it?

Somewhat of a related variant of this question is does the RD always expect the neighboring cogs to have the same pattern of less or more teeth going inboard or outboard, thus giving us the familiar cone shaped (sloped) cassettes or freewheels? What would happen if someone made a custom cassette and had a break in that pattern such as 36, 32, 29, 30... What would do RD do when it encountered a disruption in the normal pattern when it got to the 30 tooth cog? How would the RD "know" which cogs were getting larger and which are getting smaller or do they do something to the cogs themselves to assist with this? What if someone wanted to repeat a cog such as 38, 34, 29, 29... would that work?

• For a smooth change the cogs have ramps for the chain to climb to the bigger cog. Certain combinations might have the ramps incorrectly aligned resulting in a more hesitant gear change, that's the worst that could happen. – Carel Jan 27 '16 at 11:38
• What are the downvotes for? What part of this question is unclear or not useful? It seems people on here are doing it just for spite. – David Jan 27 '16 at 11:39
• Ok so then if the cogs have ramps to shift to even larger cogs, does that then imply that the direction of the cogs (larger or smaller) must be maintained otherwise the RD will get "confused", perhaps seeing a smaller cog when it is expecting a larger one? – David Jan 27 '16 at 11:43
• Down votes. Have you done any research? sheldonbrown.com/derailer-adjustment.html The first comment was ramps - did you do any research on ramps before just asking more questions? Have you ever seen a cassette with a break in pattern? How is how would a RD work with an imaginary product useful? You propose radical changes to bicycle design as in your opinion the industry has failed to innovate during your time away from bicycles. Yet you have done very little to understand existing design. Why don't you start with learning how main stream bicycles work with existing components? – paparazzo Jan 27 '16 at 12:37
• Cuz if something is not on the market that makes sense to me, then asking about it helps me to understand the current technology and why my idea is not present. Regarding the ramps, someone told me dual ramps is an idea that nobody else is doing so even if I researched it thoroughly I may have found nothing on it. – David Feb 6 '16 at 18:55

In reality the derailleur doesn't care what the tooth count difference is between gears. The shifter simply moves the derailleur towards or away from the wheel. This causes the chain to be misaligned with the current gear and move to the selected gear.

• How could it "not care"? It has to get the chain to go to the next larger cog when downshifting so if the difference is only 1 tooth vs say 10 teeth how does it handle that? That is a huge difference and occurs in real gearsets on bikes. RD's have limitations on how many total teeth of slack they can handle and also a largest cog limitation but I wonder if they also have a "largest tooth gap" limitation. For example, if someone made a special 9 speed bike (3x3) with large gaps in both the front and rear sprockets (generic term used here for chainrings and cogs). – David Jan 27 '16 at 11:47
• @David - What is this 'tooth gap' you're talking about? Do you mean the spacing between the teeth on each cog? Do you honestly think that differs at all between gear sizes? It always has to be the same, or the chain wouldn't fit. It's why there are more teeth on a larger cog... – JHCL Jan 27 '16 at 15:36
• largest tooth gap meaning from adjacent cogs is there any limit like no more than a 10 tooth gap between granny cog and next cog. No I am not talking about the chain pitch which is constant. – David Jan 27 '16 at 16:08
• Okay I see. Answer to follow. – JHCL Jan 27 '16 at 16:57
• I guess the tooth gap change will be related to the geometry of the parallelogram in the RD. Which is a design choice so might vary with different models. You could also research how Shimano megarange achieves such a large gap to the largest sprocket. – Bent Spoke Cycle Repair Feb 7 '16 at 17:29

Think of the problem from the other perspective. The designer of the cassette knows how much the derailleur will move the chain laterally - nearly constant each shift. That's all the derailleur really does.

The designer can therefore design the profile of the cassette teeth and the position of each cog 'clockwise' relative to the next one to enable good shifting, change some cogs and the shifting might not be as smooth.

Then the derailleur springs simply take up the slack in the chain. If the cog gets bigger, the chain pulls on the springs more. If it gets smaller, the springs take up the slack. Simple as that, the derailleur doesn't need to "know" anything.

Edit:

The parallelogram design of the derailleur will mean that the jockey wheels are moving downwards as the chain moves inboard. This is proportional which is not a problem for conventional cassettes. If the cassette had a large sprocket inserted like you suggest, the gap between the jockey wheel and the sprocket would be reduced significantly- potentially a problem if sprocket chosen is too large. The b-screw could be adjusted to allow a certain compromise but only to an extent.