An exercise in pure altruism you understand; not at all a thinly-veiled effort to tap the hive mind for my own dubious ends 
Overview
As will come as no surprise to anyone who's attempted to mix and match indexed gear shifting components in the past, many different standards exist dictating how these parts work - usually tied to brand, number of speeds, application (road / MTB) and how many pints the design team had had for lunch before signing off their latest effort.
Whether gear control components will play nicely together is determined prodominently by two factors:
Shifter cable pull: How much linear displacement is applied to the cable per single indexed actuation of the shifter
Derailleur movement ratio: How much lateral displacement occurs at the derailleur cage per unit of cable pull
These two factors work together to give the correct amount of derailleur travel for a clean shift between cassette sprockets or crankset chainrings.
For example, the left / front shifter on my 2x11 105 R7000 pulls around 4.7mm of cable per actuation, which coupled with a derailleur ratio of about 1.7:1 gives around 8mm of lateral displacement at the derailleur cage to move the chain from one ring to the other.
Rear Shifting
In the case of Shimano it's pretty common knowledge that for groupsets of between 5 and 9 speeds rear derailleur movement ratios remained the same across both road and MTB groupsets (at around 1.7:1), with shifter cable pulls shrinking incrementally as the number of speeds grew to reflect the narrowing sprocket pitch.
So, within this range of components road and MTB bits could be mixed as long as the shifter and cassette are intended for the same number of speeds (giving the correct cable pull and number of indexed shifts) while (tooth capacity and other factors notwithstanding) derailleurs would work with any shifter and cassette.
In terms of campatability things went south with the advent of 10 & 11 speed. Typically derailleur ratios fell in order to remove the need for shifter cable pulls to keep shrinking to accommodate the reduced pitch of a greater number of sprockets; making tolerancing less critical than it would have become on ever-shrinking pulls.
For road components this happened at the introduction of the first 11sp groupset - Dura-Ace 9000 in 2012 - and set the standard for all other 11sp road groupsets with a derailleur ratio of around 1.4:1. Conversely 10sp road retained the same 1.7:1 derailleur ratio until the introduction of Tiagra 4700 - Shimano's final 10sp road groupset - in 2016. Until this point all other 10sp road groupsets shared the same, shorter cable pulls as each other and the same higher derailleur movement ratio (1.7:1) as all 5-9sp groupsets that had gone before; while Tiagra 4700 uses the same 1.4:1 derailleur ratio as all 11sp groupsets, and has a unique cable pull to suit.
Similar changes were made with the introduction of 10sp MTB groupsets, however the increase in cable pull and reduction in derailleur ratio was more severe (apparently to further minimise the effect of manufacturing tolerances in the face of elevated levels of contamination and wear). As such Shimano 10sp MTB RDs have a ratio of around 1.2:1, and 11sp around 1.1:1... remaining (arguably) incompatable with each other and certainly incompatable with road components.
A deeper, more quantitative explanation on both derailleur movement ratios and shifter cable pull can be found on the ever-excellent Bike Gremlin site
Front Shifting
As with the rear setup we have a similar number of potentially significant variables - brand, no. speeds and application - as well as possible differences between derailleurs for double and triple cranksets.
Unfortunately unlike rear shifting which is pretty well-documented online, it seems that very little information available regarding front derailleurs and shifters beyond unquantified assertions about which parts may or may not work together. It's possible that due to the fact that chainring pitches are both wider and less affected by packaging constraints than the sprocket pitch on cassettes, that FD pull ratios / shifter cable pulls have changed less than / differently to those at the back of the bike.
As such I've started attempting to quantfy shifter cable pull and derailleur movement ratios for front derailleur setups, although I'm restricted to my own gear so currently data is scant.
I hope I'm not the only one who'd find this information useful and that the good people of CC might spend a little time measuring their own gear so that we can collectively get a fuller picture of compatability at the front end of the bike.
Measuring Method
Measuring cable pull and derailleur movement can be achieved to a reasonable standard with the following method. All references to shifting assume the crank is being rotated at the same time to allow the chain to move across rings and complete the shift. Alternatively the process could be carried out with with no chain fitted if convenient.
- With the bike ideally held in a stand, select the approximately middle gear on the cassette with the rear shifter.
- Mark a postion on the FD gear cable close / within measuring distance of a cable stop - I use a piece of masking tape wrapped around with sticky-faces stuck together to create a little flag maybe 10-20mm from the opposite face of the stop on the head tube.
- Use the front shifter to select the most outboard derailleur postition on the largest ring (i.e. keep shifting up until it won't go any further). On systems with a small incremental "trim" function at each end of the shift (most 2x groupsets, not sure about 3x), apply one small downshift to drop the derailleur into its inner / lower trim position for the big ring. This eliminates the influence of the derailleur limit screws at the end of its travel for a more accurate representation of what's going on.
- Measure and record the distance between the end of the cable stop and your cable position flag - I use a vernier but a decent steel rule would also likely be fine given the realistic tolerances were working with.
- Measure and record the position of the front derailleur relative to the bike's seatpost tube - I find it easiest to use a vernier across the NDS of the tube and outside of the derailleur cage on the DS.
- On doubles gently down-shift once (to reach the high-trim point on the small ring) and repeat measurements at the cable flag and derailleur cage. On triples down-shift again to get values for the shift between the middle and small ring.
- Repeat the process above in reverse by shifting up, again taking measurements at each point.
- Repeat whole upshift / downshift measurement process as much as you can tolerate to get a good representative sample.
- Average all downshift measured cable pull values to get a mean.
- Calculate the downshift derailleur displacement values by subtracting the tube to cage measurement from before and after the shift.
- Average all downshift derailleur displacement values to get a mean.
- For the downshift values, divide the mean derailleur displacment by the mean cable pull to get the mean derailleur value.
- Repeat the above three steps for the values obtained from up-shifting.
- Average the FD ratios for both up-and-downshifts to get a mean, and repeat for the cable pulls.
While doubtless open to some error the method above is pretty straightforward and accessible, while hopefully being sufficiently accurate to be able to identify differences between systems.
Results
So far results are limited to the three relevant bikes I have access to, but I'm hoping might be bolsted by some contributions
All values are averaged from both upshifts and downshifts.
Shifter: Shimano 105 BL-R7020 (2x11sp road) - Cable pull per shift - 4.7mm
Derailleur: Shimano 105 FD-R7000 (2x11sp road) - movement ratio 1.7:1
Shifter: Shimano Tiagra SL-4700 (2x10sp road) - Cable pull per shift - 4.5mm
Derailleur: Shimano GRX FD-RX400 (2x10sp road/gravel) - movement ratio 1.4:1
Shifter: Microshift 9s bar end job: n/a (non-indexed friction shifter) - Cable pull per shift - N/A
Derailleur: Shimano Alivio FD-T4000 (3x9sp MTB) - movement ratio 1.0:1
Discussion
Not a whole lot to discuss currently, however I was suprised to see the apparent difference between the 10sp and 11sp road setups as I'd have expected the derailleur ratios to be the same. Further data for each would be good to corroborate or discredit these findings.
Further, the 9sp derailleur ratio is significantly different again; however it's not possible to know whether this is because it's an MTB groupset, 9sp, a triple or a combination of these without further data.
From looking at compatability charts and images of components it appears the changes in rear derailleurs and shifters that occurred between 10sp Tiagra 4600 and 4700 were reflected in the 2x front derailleur; as the 4700 item gets a longer cable arm while the rest looks the same, so presumably the ratio is correspondingly lower. However, this doesn't appear to be the case for the 3x FD which seems to have remained the same despite the derailleur ratios and cable pulls changing for the rest of the system.
Likewise this 2x situation appears to be echoed in the 105 linenup - with the 11sp FD-5800 getting a longer arm than the previous 10sp 5700. Unfortunately 3x FDs can't be compared across these two generations as Shimano killed the road triple before it got to an 11sp variant.
Earlier 9sp 3x FDs look geometrically similar to later 10sp variants (possibly with a smallish change in cable and cage arm lengths at some point which likely left the ratios unchanged) however it's difficult to draw proper conclusions about these without having the ability to test them.
Likewise it's difficult to compare MTB and later road FDs from drawings in the same way as the above since their design is very different to earlier road FDs with the effective arm lengths less visibly obvious...
Anyway that's pretty much all I know so far; hopefully some might find this of interest. It would really make my day if anyone could contribute some relaible data from 9sp road and 10sp MTB triples
Overview
As will come as no surprise to anyone who's attempted to mix and match indexed gear shifting components in the past, many different standards exist dictating how these parts work - usually tied to brand, number of speeds, application (road / MTB) and how many pints the design team had had for lunch before signing off their latest effort.
Whether gear control components will play nicely together is determined prodominently by two factors:
Shifter cable pull: How much linear displacement is applied to the cable per single indexed actuation of the shifter
Derailleur movement ratio: How much lateral displacement occurs at the derailleur cage per unit of cable pull
These two factors work together to give the correct amount of derailleur travel for a clean shift between cassette sprockets or crankset chainrings.
For example, the left / front shifter on my 2x11 105 R7000 pulls around 4.7mm of cable per actuation, which coupled with a derailleur ratio of about 1.7:1 gives around 8mm of lateral displacement at the derailleur cage to move the chain from one ring to the other.
Rear Shifting
In the case of Shimano it's pretty common knowledge that for groupsets of between 5 and 9 speeds rear derailleur movement ratios remained the same across both road and MTB groupsets (at around 1.7:1), with shifter cable pulls shrinking incrementally as the number of speeds grew to reflect the narrowing sprocket pitch.
So, within this range of components road and MTB bits could be mixed as long as the shifter and cassette are intended for the same number of speeds (giving the correct cable pull and number of indexed shifts) while (tooth capacity and other factors notwithstanding) derailleurs would work with any shifter and cassette.
In terms of campatability things went south with the advent of 10 & 11 speed. Typically derailleur ratios fell in order to remove the need for shifter cable pulls to keep shrinking to accommodate the reduced pitch of a greater number of sprockets; making tolerancing less critical than it would have become on ever-shrinking pulls.
For road components this happened at the introduction of the first 11sp groupset - Dura-Ace 9000 in 2012 - and set the standard for all other 11sp road groupsets with a derailleur ratio of around 1.4:1. Conversely 10sp road retained the same 1.7:1 derailleur ratio until the introduction of Tiagra 4700 - Shimano's final 10sp road groupset - in 2016. Until this point all other 10sp road groupsets shared the same, shorter cable pulls as each other and the same higher derailleur movement ratio (1.7:1) as all 5-9sp groupsets that had gone before; while Tiagra 4700 uses the same 1.4:1 derailleur ratio as all 11sp groupsets, and has a unique cable pull to suit.
Similar changes were made with the introduction of 10sp MTB groupsets, however the increase in cable pull and reduction in derailleur ratio was more severe (apparently to further minimise the effect of manufacturing tolerances in the face of elevated levels of contamination and wear). As such Shimano 10sp MTB RDs have a ratio of around 1.2:1, and 11sp around 1.1:1... remaining (arguably) incompatable with each other and certainly incompatable with road components.
A deeper, more quantitative explanation on both derailleur movement ratios and shifter cable pull can be found on the ever-excellent Bike Gremlin site
Front Shifting
As with the rear setup we have a similar number of potentially significant variables - brand, no. speeds and application - as well as possible differences between derailleurs for double and triple cranksets.
Unfortunately unlike rear shifting which is pretty well-documented online, it seems that very little information available regarding front derailleurs and shifters beyond unquantified assertions about which parts may or may not work together. It's possible that due to the fact that chainring pitches are both wider and less affected by packaging constraints than the sprocket pitch on cassettes, that FD pull ratios / shifter cable pulls have changed less than / differently to those at the back of the bike.
As such I've started attempting to quantfy shifter cable pull and derailleur movement ratios for front derailleur setups, although I'm restricted to my own gear so currently data is scant.
I hope I'm not the only one who'd find this information useful and that the good people of CC might spend a little time measuring their own gear so that we can collectively get a fuller picture of compatability at the front end of the bike.
Measuring Method
Measuring cable pull and derailleur movement can be achieved to a reasonable standard with the following method. All references to shifting assume the crank is being rotated at the same time to allow the chain to move across rings and complete the shift. Alternatively the process could be carried out with with no chain fitted if convenient.
- With the bike ideally held in a stand, select the approximately middle gear on the cassette with the rear shifter.
- Mark a postion on the FD gear cable close / within measuring distance of a cable stop - I use a piece of masking tape wrapped around with sticky-faces stuck together to create a little flag maybe 10-20mm from the opposite face of the stop on the head tube.
- Use the front shifter to select the most outboard derailleur postition on the largest ring (i.e. keep shifting up until it won't go any further). On systems with a small incremental "trim" function at each end of the shift (most 2x groupsets, not sure about 3x), apply one small downshift to drop the derailleur into its inner / lower trim position for the big ring. This eliminates the influence of the derailleur limit screws at the end of its travel for a more accurate representation of what's going on.
- Measure and record the distance between the end of the cable stop and your cable position flag - I use a vernier but a decent steel rule would also likely be fine given the realistic tolerances were working with.
- Measure and record the position of the front derailleur relative to the bike's seatpost tube - I find it easiest to use a vernier across the NDS of the tube and outside of the derailleur cage on the DS.
- On doubles gently down-shift once (to reach the high-trim point on the small ring) and repeat measurements at the cable flag and derailleur cage. On triples down-shift again to get values for the shift between the middle and small ring.
- Repeat the process above in reverse by shifting up, again taking measurements at each point.
- Repeat whole upshift / downshift measurement process as much as you can tolerate to get a good representative sample.
- Average all downshift measured cable pull values to get a mean.
- Calculate the downshift derailleur displacement values by subtracting the tube to cage measurement from before and after the shift.
- Average all downshift derailleur displacement values to get a mean.
- For the downshift values, divide the mean derailleur displacment by the mean cable pull to get the mean derailleur value.
- Repeat the above three steps for the values obtained from up-shifting.
- Average the FD ratios for both up-and-downshifts to get a mean, and repeat for the cable pulls.
While doubtless open to some error the method above is pretty straightforward and accessible, while hopefully being sufficiently accurate to be able to identify differences between systems.
Results
So far results are limited to the three relevant bikes I have access to, but I'm hoping might be bolsted by some contributions
All values are averaged from both upshifts and downshifts.
Shifter: Shimano 105 BL-R7020 (2x11sp road) - Cable pull per shift - 4.7mm
Derailleur: Shimano 105 FD-R7000 (2x11sp road) - movement ratio 1.7:1
Shifter: Shimano Tiagra SL-4700 (2x10sp road) - Cable pull per shift - 4.5mm
Derailleur: Shimano GRX FD-RX400 (2x10sp road/gravel) - movement ratio 1.4:1
Shifter: Microshift 9s bar end job: n/a (non-indexed friction shifter) - Cable pull per shift - N/A
Derailleur: Shimano Alivio FD-T4000 (3x9sp MTB) - movement ratio 1.0:1
Discussion
Not a whole lot to discuss currently, however I was suprised to see the apparent difference between the 10sp and 11sp road setups as I'd have expected the derailleur ratios to be the same. Further data for each would be good to corroborate or discredit these findings.
Further, the 9sp derailleur ratio is significantly different again; however it's not possible to know whether this is because it's an MTB groupset, 9sp, a triple or a combination of these without further data.
From looking at compatability charts and images of components it appears the changes in rear derailleurs and shifters that occurred between 10sp Tiagra 4600 and 4700 were reflected in the 2x front derailleur; as the 4700 item gets a longer cable arm while the rest looks the same, so presumably the ratio is correspondingly lower. However, this doesn't appear to be the case for the 3x FD which seems to have remained the same despite the derailleur ratios and cable pulls changing for the rest of the system.
Likewise this 2x situation appears to be echoed in the 105 linenup - with the 11sp FD-5800 getting a longer arm than the previous 10sp 5700. Unfortunately 3x FDs can't be compared across these two generations as Shimano killed the road triple before it got to an 11sp variant.
Earlier 9sp 3x FDs look geometrically similar to later 10sp variants (possibly with a smallish change in cable and cage arm lengths at some point which likely left the ratios unchanged) however it's difficult to draw proper conclusions about these without having the ability to test them.
Likewise it's difficult to compare MTB and later road FDs from drawings in the same way as the above since their design is very different to earlier road FDs with the effective arm lengths less visibly obvious...
Anyway that's pretty much all I know so far; hopefully some might find this of interest. It would really make my day if anyone could contribute some relaible data from 9sp road and 10sp MTB triples
Last edited:
