Bike Forums

Bike Forums (https://www.bikeforums.net/forum.php)
-   Bicycle Mechanics (https://www.bikeforums.net/bicycle-mechanics/)
-   -   Gearing and shifting patterns (https://www.bikeforums.net/bicycle-mechanics/1162536-gearing-shifting-patterns.html)

Hatsuwr 12-19-18 10:04 AM

Gearing and shifting patterns
 
As part of my never-ending quest to overthink everything, I took a quick look at my shifting pattern. This has probably been done before, but I haven't seen it. Goal was to shift from the smallest gear ratio to largest with the most even steps and in a way that was simple and avoided high-loss combinations.

I have 30/39/50 teeth on the front, 11/13/15/17/20/23/26/30/34 teeth on the back.

I measured the approximate lateral separation of the rings, in mm. Higher separations were avoided. Also gave a bit more preference toward higher tooth count combinations, since I believe those result in less friction loss.

Tried to avoid small/no changes in gear ratio (delta ratio), as well as large ones.

Weighted all of these things to assign an overall value to the combination. Value is from 0-10, with higher being better. It really doesn't mean anything by itself, but might be a sign to look closer at a particular combination.

https://cimg9.ibsrv.net/gimg/bikefor...940ec727ee.png

Here is what I came up with. Nothing terribly surprising here, but it should optimize everything a bit and isn't hard to remember. Will probably paint a couple small dots on my shifting indicators to help.

One interesting thing is that there are only 13 combinations here. The eliminated ones really wouldn't add much functionality. This supports my plan to eventually replace that 3 ring front.

https://cimg6.ibsrv.net/gimg/bikefor...a842e66738.png

Got a couple questions too.

Anyone have a better way to measure lateral separation of the gears? I just used a straightedge on the largest front ring and measured from there. Fairly accurate, but could be better.

Also, why don't we see something like 4 front and 4 rear gears? It shouldn't be too hard to write up something that, when given min/max ratio, min/max gear size, and number of gears of front/back would assess all possible tooth counts to find the one with the most even spacing. I suppose it would make things a bit less intuitive - needing to change both the front and back gear for more of the jumps. Would be interesting to see that combined with computerized gear selection. That might be a project for another month though.

fietsbob 12-19-18 11:10 AM

next place The cog T count on a vertical side, chainring on top of a grid .. plug in the numbers

Draw a zigzag line between them as the Ratio values are closer to each other .. and there is your sequence

visual.. tape it to your bike..

BITD we had half step gearing... middle & large chain ring close together in size, so gear sequence alternated between them..
'granny' took up at the bottom end..

Brompton 6 speed uses 2 cogs half stepping .. a wider ratio 3 speed hub.. now..


_________________________________

In the 50's I put together a 3 speed hub, with 3 cogs and a triple chainring crankset

now the crank can be another internal gear .. rear internal gear hub's engineers sort out the sequence for you..

Like that dual range German 7 speed hub.... for 14 speeds all in a row..





....





.....

AnkleWork 12-19-18 11:26 AM

OP: Your methodology will not find the optimum shift patterns, which depend on conditions, exertion and some other stuff not in your spreadsheet. If you ride a lot more you will find them by experience and intuition.

Retro Grouch 12-19-18 11:38 AM

Back in the day when men were men 10-speed meant 2 on the front and 5 on the back we had to make use of every combination so we did a lot of that kind of stuff. Some guys would, indeed figure out the optimum gear progression and tape it to their handlebars.

As the number of rear cogs has gradually progressed, we have many, many more ratios to choose from. So many that most riders don't use them all. I think of a 3 X 10 drive train as having 3 gear ranges: 1 for steep uphills, one for use most of the time and 1 reserved for downhills and tail wind days. The rear derailleur serves as a trim within those 3 ranges.

I have neither experience nor interest with electronic shifting, but I can see where that may actually bring back the old days. Electronic shifting can "remember" all of the potential ratios to pick the next one in sequence and shift both derailleurs to optimize the chain line.

Hatsuwr 12-19-18 11:44 AM


Originally Posted by AnkleWork (Post 20711303)
OP: Your methodology will not find the optimum shift patterns, which depend on conditions, exertion and some other stuff not in your spreadsheet. If you ride a lot more you will find them by experience and intuition.

Well, I wasn't trying to define optimum. "Goal was to shift from the smallest gear ratio to largest with the most even steps and in a way that was simple and avoided high-loss combinations."

Don't think conditions will affect much either way, and unless you want to add more complication to the progression, this should help with exertion if anything.

Still, when choosing between 50/34 and 30/20 which have only a .03 difference in ratio, the decision is pretty clear to me. I'd even go so far as to say optimal, regardless of conditions, exertion, etc. Maybe if you broke one of your arms, although with that choice I'd still probably just lean over and shift carefully

Phamilton 12-19-18 12:20 PM


Originally Posted by Retro Grouch (Post 20711326)
Back in the day when men were men 10-speed meant 2 on the front and 5 on the back we had to make use of every combination so we did a lot of that kind of stuff. Some guys would, indeed figure out the optimum gear progression and tape it to their handlebars.

As the number of rear cogs has gradually progressed, we have many, many more ratios to choose from. So many that most riders don't use them all. I think of a 3 X 10 drive train as having 3 gear ranges: 1 for steep uphills, one for use most of the time and 1 reserved for downhills and tail wind days. The rear derailleur serves as a trim within those 3 ranges.

I have neither experience nor interest with electronic shifting, but I can see where that may actually bring back the old days. Electronic shifting can "remember" all of the potential ratios to pick the next one in sequence and shift both derailleurs to optimize the chain line.

Just a month ago I switched from commuting on 2 x 6 to 3 x 7. Just like you said, cruise in middle ring, climb in the small ring, descend in big ring (or tailwind cruise). No sequential shifting required, and redundancies are irrelevant. Adjust for wind and fine tune cadence (trim) with RD. If the wide range 9 spd 11-34 cassette leaves one with steps that are bigger than desired, tighten them up with a 12-27 or similar. If I was running a 30/39/50 triple, a 10 spd 12-30 would give nice tight steps between all the gears in all 3 ranges only sacrificing the very top and bottom end, and not by much.

Edit: if it's really really windy, just don't use the big ring at all. It took some practiced restraint on my part to get out of the habit. "It's there, so I'm gonna use it".

AnkleWork 12-19-18 12:36 PM


Originally Posted by Hatsuwr (Post 20711334)
Well, I wasn't trying to define optimum. "Goal was to shift from the smallest gear ratio to largest with the most even steps and in a way that was simple and avoided high-loss combinations."

Don't think conditions will affect much either way, and unless you want to add more complication to the progression, this should help with exertion if anything.

Still, when choosing between 50/34 and 30/20 which have only a .03 difference in ratio, the decision is pretty clear to me. I'd even go so far as to say optimal, regardless of conditions, exertion, etc. Maybe if you broke one of your arms, although with that choice I'd still probably just lean over and shift carefully

So you're pulling a steep hill and you're getting gassed and you're already on the wrong side of your power curve -- what do you do? Consult a spreadsheet and think about it?
Most people who ride more than a little will just grab some lower gears as quickly as possible to minimize further slowing. Exactly which lower gear is not as important as shifting quickly. Can't seem to find those factors in your chart. [And most people who ride more than a little will know other scenarios.]

Friendly suggestion: Ride more and you'll be much happier.

Hatsuwr 12-19-18 12:39 PM


Originally Posted by Phamilton (Post 20711395)
Just a month ago I switched from commuting on 2 x 6 to 3 x 7. Just like you said, cruise in middle ring, climb in the small ring, descend in big ring (or tailwind cruise). No sequential shifting required, and redundancies are irrelevant. Adjust for wind and fine tune cadence (trim) with RD. If the wide range 9 spd 11-34 cassette leaves one with steps that are bigger than desired, tighten them up with a 12-27 or similar. If I was running a 30/39/50 triple, a 10 spd 12-30 would give nice tight steps between all the gears in all 3 ranges only sacrificing the very top and bottom end, and not by much.

Edit: if it's really really windy, just don't use the big ring at all. It took some practiced restraint on my part to get out of the habit. "It's there, so I'm gonna use it".

That's essentially what my results come down to, although with that added point that the middle ring use is going to be (roughly) most effective when using the 15/17/20/23 rear gears, in my specific situation. Also don't use the large/small ring on the wrong side of that range. Personally, I'd go up or down one for short bursts if I'm not planning on using the small or large ring anyway. Should be more efficient and have less wear by not doing that long term though.


Originally Posted by AnkleWork (Post 20711421)
So you're pulling a steep hill and you're getting gassed and you're already on the wrong side of your power curve -- what do you do? Consult a spreadsheet and think about it?
Most people who ride more than a little will just grab some lower gears as quickly as possible to minimize further slowing. Exactly which lower gear is not as important as shifting quickly. Can't seem to find those factors in your chart. [And most people who ride more than a little will know other scenarios.]

Friendly suggestion: Ride more and you'll be much happier.

Nope, that's why I specified a couple times that the pattern should be simple. And it is I think - two dots on the rear derailleur indicator would suffice, if you needed a reference at all. Are you maybe looking at the larger chart? That's just a list of all combinations.

Also, I'm quite happy, but thank you for the concern!

fietsbob 12-19-18 01:01 PM

OK but not everyone uses a out of the box road triple, which your50,39,30 is ,

but I note, computer programmers want to show off here how they spend their non riding time..

as a presentation of that .. nice work.. adding colors and etc..

there are others, that take input from users of the tool ..


I like the one John Allen maintains for the late Sheldon Brown gears ..




...

Phamilton 12-19-18 01:08 PM


Originally Posted by Hatsuwr (Post 20711427)
That's essentially what my results come down to, although with that added point that the middle ring use is going to be (roughly) most effective when using the 15/17/20/23 rear gears, in my specific situation. Also don't use the large/small ring on the wrong side of that range. Personally, I'd go up or down one for short bursts if I'm not planning on using the small or large ring anyway. Should be more efficient and have less wear by not doing that long term though.

I think you're on the right track. I don't worry as much about the wear because I'm commuting and the stuff gets thrashed anyway. The logic of the little ring climb, middle ring cruise, big ring descend keeps you out of those weird combinations anyway. FWIW, my drivetrain is 24/32/38, 11-13-15-18-21-24-28. I basically only use the 13,15,18, and 21t cogs at all, and the little ring barely gets used. I'd like to switch to a 22/32/42 crank as the jumps on the 24/32/38 are a little "soft".

Phamilton 12-19-18 01:10 PM


Originally Posted by fietsbob (Post 20711465)
...
I like the one John Allen maintains for the late Sheldon Brown gears ..

...

It's my "go-to" for crunching gear numbers. I'm sure glad they don't charge for using it. There are better ones online and even apps, but that's the first one I ever saw.

Hatsuwr 12-19-18 01:12 PM


Originally Posted by fietsbob (Post 20711465)
OK but not everyone uses a out of the box road triple, which your50,39,30 is ,

but I note, computer programmers want to show off here how they spend their non riding time..

as a presentation of that .. nice work.. adding colors and etc..

there are others, that take input from users of the tool ..

I like the one John Allen maintains for the late Sheldon Brown gears ..

I did browse his stuff - lots of good foundational info!

And ya, I don't expect the final results to be applicable to anyone else. You'd have to match up all the gear sizes, as well as their relative lateral positions.

It seems like you might be getting the idea that I'm trying to push a rigid mathematical framework of riding on others. Just want to assure you that this isn't the case! Firstly because I'm just sharing my results - throw them in the trash if you want, doesn't bother me! But I'm also not blind to the experiential aspect of riding. I do think there are a lot of less-than-ideal habits out there worth examining, and giving a fair chance to their alternatives. At the end of the day though, do what makes you happy. Efficiency might be the point of this post, but I'd rather have fun going slow than be miserable and fast!

kingston 12-19-18 02:16 PM

It looks like you figured out that you should use the low end of the cassette with the little ring, the high end with the big ring and the middle with the middle. What am I missing?

Hatsuwr 12-19-18 02:28 PM


Originally Posted by kingston (Post 20711578)
It looks like you figured out that you should use the low end of the cassette with the little ring, the high end with the big ring and the middle with the middle. What am I missing?

I think I wrote it out a couple times already, but just to restate - nothing terribly surprising here. A different set of gear sizes might have had something more interesting with ratio intervals, and different lateral measurements might have contraindicated some otherwise good combinations. Not here though - fairly sure Shimano thought all this out already.

Besides the analysis technique, the primary useful result is to point out the specific points at which I should ideally transition from the small/medium/large rings, and noting that a single step up/down in the cassette should accompany those transitions (with my particular setup.)

nfmisso 12-19-18 02:40 PM


Originally Posted by Hatsuwr (Post 20711427)
That's essentially what my results come down to, although with that added point that the middle ring use is going to be (roughly) most effective when using the 15/17/20/23 rear gears, in my specific situation. Also don't use the large/small ring on the wrong side of that range. Personally, I'd go up or down one for short bursts if I'm not planning on using the small or large ring anyway. Should be more efficient and have less wear by not doing that long term though........

For my commuter - I have well defined routes (grade/wind/etc); I found that a 1 x 9 with 44T chainring, 12-36 HG400 cassette with 38-622 tire covered my needs perfectly. My commuting bike only goes to/from work and occasionally around the neighbor hood - so no need to carry things (triple chainring & FD for example) that I'll never need. The 44-36 is sufficient for climbing the steepest grade I encounter from a dead stop (under 101 by SJC airport). 44-12 is sufficient for 15mph tailwind riding.on a crowded bike path.

There is no need for every bike to be able to handle every possible terrain; modify your bike to suit how and where you ride.

Retro Grouch 12-19-18 04:32 PM


Originally Posted by Phamilton (Post 20711478)
It's my "go-to" for crunching gear numbers. I'm sure glad they don't charge for using it. There are better ones online and even apps, but that's the first one I ever saw.

I find I can crank out the numbers using a pocket calculator almost as fast as looking them up on a chart.

Phamilton 12-19-18 05:06 PM


Originally Posted by Retro Grouch (Post 20711784)
I find I can crank out the numbers using a pocket calculator almost as fast as looking them up on a chart.

Slide rule.

kingston 12-19-18 05:20 PM

I like the Mike Sherman gear calculator the best.

TimothyH 12-19-18 05:25 PM


Originally Posted by Hatsuwr (Post 20711136)
high-loss combinations.

What does this mean?


-Tim-

Hatsuwr 12-19-18 05:29 PM


Originally Posted by Phamilton (Post 20711815)


Slide rule.

Gotta break out the abacus if you want people to take you seriously.


Originally Posted by TimothyH (Post 20711836)
What does this mean?

Avoiding high lateral separation of gear combinations and also slightly favoring larger gears.

davidad 12-24-18 05:08 PM

Bicycle Gear Inch & Shifting Pattern Calculator

Ironfish653 12-30-18 03:02 PM


Originally Posted by Hatsuwr (Post 20711844)
Gotta break out the abacus if you want people to take you seriousl

Avoiding high lateral separation of gear combinations and also slightly favoring larger gears.

Calculating gear ratio progressions isn’t all that big of a deal; and what you’ve found is pretty standard stuff for most bikes’ drivetrains as delivered.
Essentially, that there’s a lot of overlap in the middle, vintage half-step setups notwithstanding.

Using the gear-inch calculator also accounts for wheel and tire sizes, which is useful when comparing different bikes in your stable.

I think what’s confusing people is that you’re looking really hard at the ‘lateral measurement’ between the cogs as some sort of indicator. I mean, the cogs are always going to be evenly spaced, and the point on the cassette where the ratios overlap has only to do with the number of teeth on the cogs, not where they are in the order.

MikeWMass 12-30-18 03:27 PM


Originally Posted by Also, [i
why don't we see something like 4 front and 4 rear gears?[/i] It shouldn't be too hard to write up something that, when given min/max ratio, min/max gear size, and number of gears of front/back would assess all possible tooth counts to find the one with the most even spacing. I suppose it would make things a bit less intuitive - needing to change both the front and back gear for more of the jumps. Would be interesting to see that combined with computerized gear selection. That might be a project for another month though.

Don't forger that shifting the front is fundamentally different from shifting the back. Front is moving the tensioned part of the chain, rear the slack part. Before ramped and pegged cogs, we soft-pedaled for all shifts. Not necessary (but still nice) for rear shifts now; but front doesn't like to be shifted under power, and is more likely to drop the chain, or at least skip. This is (I believe) why 1x is becoming popular in mountain biking.

In the old half-step gearing mentioned above, both were shifted if you needed a small change in gearing. With down-tube shifters you could shift both with one hand. It is actually easier with todays indexed brake-shifters, but why bother with 9-11 (or is it 12) cogs on the back?


All times are GMT -6. The time now is 05:08 AM.


Copyright © 2026 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.