Gearing (was Getting Over My Head)

[Ted Durant]

Hi All -

Leah's "Getting Over My Head" thread seems to have evolved into a gearing thread. I retired from a career in data mining and statistical modeling, so you can just imagine how much time I've wasted on "optimizing" bicycle gearing. 

Here's the TL;DR version: it's impossible to optimize gearing, so stop sweating it.

The slightly longer version is that we are locked into integer tooth counts; when the optimal cog is 14.5 teeth, that ain't an option! So, we are forced into compromising or, better yet, satisficing. 

There are essentially three key points we are trying to fix with gears: the lowest low, the highest high, and the steps in between. Given those, we then try create a system that reliably and easily shifts among the gears. For me, a 2x system using components available today provides the best combination of low-low, high-high, steps in between, simplicity, consistency, and reliability. But that depends highly on the chain, the chainrings, and the front derailer playing nicely together. It also works for me because I use it as 2 gearing ranges, one for flats and downhills, the other for long, steeper uphills.

The lowest low and the highest high are pretty straightforward, and plenty of ink has been spilled on how to choose those, so there's no point in elaborating on that. It's the "steps in between" part that drives us wild. In theory, we think, we'd like to have perfectly even steps between the high and low. In practice, though, 1) that's simply not possible with a cog-and-chain drivetrain, and 2) it might not even be that desirable. A lot of riders, myself included, find that we prefer smaller steps between gears in the range in which we normally ride, and larger steps out in the extremes. 

What I definitely don't like is having a big difference between 3 adjacent cogs in the middle of my cruising range. For example, a 1-tooth difference one way and a 2-tooth difference the other. Unfortunately, this is a common occurrence in large cog count cassettes with tiny small cogs - they go from a 1-tooth difference to a 2-tooth difference near the middle of the cluster. That's twice the amount of reduction/increase in effort. So, when I'm looking at cassettes, I'm looking for ones where that 1-to-2 transition occurs as close to the small cog as I can get it. A major factor here is the movement to smaller smallest cogs, which has gone from 14 to 10 in my time. Starting from 11 (or, God forbid, 10!) you use a lot of cogs to get to the point where 2-tooth steps start to make sense.

On the other hand, those small smallest cogs mean we can use small outer chainrings, and that's something of a boon if your front derailer can handle it, because it means we can also use smaller inner rings on a 2x to get sufficiently low gearing. The difference between chainrings is worth examining a bit. Typical road double front derailers have a 16-tooth max difference specification, which derives from the standard "compact double" 50-34. That's a 39% difference, which is a pretty big jump, roughly 3.3x the average jump on the cassettes often paired with those chainrings. So, shifting up front is the equivalent of around 3 1/3 cogs in back. On my Waterford I use a 42-tooth large ring, and a 26-tooth small ring is 16t smaller, but that's a whopping 48% difference, which is 5x the average jump on my cassette. Piaw mentioned going with smaller tooth differences up front, and there's a good example of why. When I built my Breadwinner I went with 44x32 up front, a 32% difference that is 3.1x the average difference on the cassette. I find that to be a much less disruptive change than on my Waterford. The front chainring difference as a multiple of the average in back turned out to be a significant factor for me. Sure, it means I have more overlap in gears, but that's less important to me than the change in cadence caused by shifting.

One last consideration as I'm designing a drivetrain is that I want my preferred cruising gear near the center of the rear cogs, maybe a little closer to the small end, so that I can be on the large ring for most of my riding around home. That's around a 5.0 gain ratio for me now, which is the 42x17 on my Waterford. And, as noted, I want a consistent difference above and below that gear, which means, for me, a 2-tooth change on each side of it. 

Ted Durant

Milwaukee, WI USA

[George Schick]

Thanks a bunch, Ted.  It's very helpful.  What you say here is pretty much the way I've set things up on my bikes and I've enjoyed riding them with gearing like this.

[Patrick Moore]

Following Ted's good example by taking this to his new thread:

Back in the 7 and perhaps 8 speed days cassettes were bolted together and it was easy to disassemble them. Then they started using rivets which were easy to drill out. The only real problem in disassembling cassettes for individual cogs is cassettes built on spiders, tho' I suppose you can find a spidered cog subset of your choice and add it to loose smaller cogs; I stopped at the rivets. But I used to have a (really) 20 lb box of 7, 8, 9, and 10 sp HG cogs and some Uniglides from which I built 7, 8, 9, and even 10 sp cassettes that worked fine; then I ordered Shimano-type Miche cogs in more or less the right series; these shifted better; in fact,  with 11 sp chain on 10 of them, perfectly, and I sold that box of cogs and stocked up on Miches.

I bought my Miches from Cycle Clinic in England during COVID when you couldn't find them in the US; perhaps now again available domestically; quick looks says Modern Bike has 53 different ones. Perhaps QBP is now carrying them again?

Parick Moore, who also filed modern 3-prong SA 3/32 cogs to fit ancient 2-spline 1/8" SA drivers.

On Sat, May 20, 2023 at 11:49 AM George Schick <bhi...@gmail.com> wrote:

Now that we've pretty thoroughly hijacked this blog thread:  Patrick, if you've been building your own cassettes for 20 years, where have you been able to purchase individual cogs (and spacers) in order to assemble your own desired combination?  Everything I've noticed when removing cassettes from their splined "body" for cleaning or substituting a different combo's was that several of the central cogs are riveted together (which is what I think Piaw was getting at when he said "...after cassettes were introduced you couldn't pick individual sprockets any more...")?

On Saturday, May 20, 2023 at 11:58:59 AM UTC-5 Patrick Moore wrote:

I've been home-building cassettes for 20 years at least after I got sufficiently frustrated with stock combinations; in friction (and even indexed) they have shifted fine. I agree about 11 sp chains; at least, I'm using an 11 on my (custom, built with Miche cogs) 10 sp cassette and I've never had better shifting. Have read many places that 10 sp chains ar longer lasting than 9, 11 than 10. Perhaps will try making an 11 sp cassette with my 10 sp Miche cogs by substituting 11 sp spacers for the 10 sp ones and will try a 12 sp chain.

[Patrick Moore]

Just remembered: Even after selling the big box o' cogs I re-acquired another big box of them, not quite 20 lb, still sitting under my workbench. These are from 8 but mostly 9-10 and some 11 discard cassettes; no more 7s or Uniglides. If anyone needs a particular loose HG cog, I'll be happy to look and, if it's not a strategic one I want to hoard I'll be happy to share. Note: I can search only by # teeth, not era (# cogs in cassette) or series (position of cog in cassette). The present box also includes about 2.5 lbs each of fixd cogs, ss Shimano/clone cogs, and SA cogs

--

-----------------------------------------------------------------------

Patrick Moore

Alburquerque, Nuevo Mexico, Etats Unis d'Amerique, Orbis Terrarum

[John Hawrylak, Woodstown NJ]

I agree with Ted on the problem of getting Ideal Cog Teeth & first cogs being too small (11T & less) , but I believe a 46-36-26 triple with a wide range Shimano HG-400 12-36 9 speed cassette (Deore RD-M591-SGS), 130mm OLD Tiagra 4000 rear hub, and 650Bx38's gives me:

good gearing in my cruising range (75 to 55gi) with 3 cogs to use in the Outer and Middle rings

an even 16% to 12% gearing change between all the cogs.  

good chain line (<= 0.040% chain stretch) over a 6 cog range on each chain ring (front CL's of 52, 44, 38mm & rear CL of 44mm)

good high/low range (101 to 19 gi)

good FD performance with old school style FDs (using a 9 speed Sora  FD-R3000, but also used a 1980's Shimano FD-MT60) 

The learning for me was to be willing to shift between the Outer and Middle rings much more often than I thought before, e.g. shift to Middle 3 cog when using the Outer 4 cog and the terrain indicates a need for lower gearing.   Before I tended to stay on the Outer ring too long.  Now, the outer and middle rings are used much more evenly.  The 10T ring difference makes shifting the rings easy.  The 26T Inner is still for big hills, but I feel I get to it more efficiently than before.

The triple limits you to a 44T Outer (FD cage hitting the C/S with a 42T). but the 46T/12T combo gives me the 100 gi high I want with 650Bx38, so I see the chain ring size limit as theoretical limit I don't encounter as long I have a large stock of HG400-9 12-36 cassettes.

As for Leah's problem with shifting to lower gears, I think Sheldon said it best (paraphasing),  "when you come to hills, shift to your lower gears before you need too".  Easier said than done.

John Hawrylak

Woodstown NJ

 

[Garth]

I was watching the uphill penultimate TT road racing stage of the Giro D' Italia this weekend where the eventual winner Primoz Roglic chose to run a 1x for his TT bike. The announcers were "praising" him for the

"innovative" choice of using a mtb gearing setup for the TT. Well lo and behold, at a low speed(which is fast for mere mortals, hah hah) he was about 2/3 up the course and he ran over what appeared to be a small break in the road crosswise and his chain derailed off the ring. He put it back on cool as a cucumber and still trounced everyone, but so much for 1x's not failing like any other system, because they do.

I also ran across a video of a "pro" mechanic touting stuff he likes. I was with him on some things, then he mention that cables were so passe',  and that the way to go is electronic shifting and 1x. A lifetime of buying failing batteries, waning replacements, software updates and subscription fees .... yummm ... doesn't that sound appealing ?

In regards to cassette cogs, the Miche 8-11sp are available from European websites, let your fingers do the walkin'. Also, Sunrace 7-9 speed lower end cassettes(no spiders) are all held together with removable pins. Find a backside photo of the potential cassette and you can confirm it.

[Patrick Moore]

Garth: Do you by any chance know what cog combination or what a typical cog combination would be for such a climb by a pro? I'll be he didn't use a 52 t inner cog and that he had a pretty close spread in the center of the cassette, but would be happy to be proved right or wrong by evidence.

As to 1Xs for pro racing: with 11 or 12 cogs, do pros need to shift between rings at all, even using 2Xs, except at major terrain transitions?

I got my collection of 10 sp Miche cogs at Cycle Clinic in the UK.

--
You received this message because you are subscribed to the Google Groups "RBW Owners Bunch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rbw-owners-bun...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rbw-owners-bunch/94e4b0af-bdfb-4307-999c-c5ad38971f9dn%40googlegroups.com.

[Drew Saunders]

This article: https://road.cc/content/tech-news/primoz-roglic-uses-gravel-gearing-decisive-giro-tt-301511

Says he used the "XPLR-1271" from SRAM, which only offers a 10-44 (10,11,13,15,17,19,21,24,28,32,38,44) cassette, with even chainrings between 38 through 46, most likely a 44. I don't know if there are better articles on what he ended up using but with a 44t ring, that's a 4.4:1 through 1:1 range. 

[Patrick Moore]

Thanks, Drew. I watched the Twitter video; pretty steep even with a 1:1!

I looked for a clip of Coppi or Bartali climbing similar cols in their 46 X 19 or whatever low gears were back then (someone told me but I can't find the email now) but couldn't find one in the first 5 minutes, so gave up.

To view this discussion on the web visit https://groups.google.com/d/msgid/rbw-owners-bunch/b9af1a2e-f105-41b5-8ee4-945280ab0583n%40googlegroups.com.

[Ken Yokanovich]

I am a big fan of the Shimano HG50 14-25 9 speed cassettes:  (14-15-16-17-18-19-21-23-25T) paired with a 42T ring for a 1x drivetrain or 44/30 2x drivetrain

On Saturday, May 20, 2023 at 12:55:27 PM UTC-5 Ted Durant wrote:

[Patrick Moore]

That 9 speed is very close to the home-brew 10 I'm using; simply add an outer 13 and 10 speed cogs, tho mine goes 19-20-22-25; the difference is minor. I'll have to remember that particular cassette.

[Garth]

Patrick, Drew and everyone, Eurosport said it was a 44t chainring with a "pie plate" cassette" during the live broadcast, you can see it here : https://www.eurosport.com/cycling/giro-d-italia/2023/primoz-roglic-mechanical-watch-the-moment-giro-ditalia-dreams-crumbled-until-he-stormed-back-to-win_sto9626269/story.shtml

The announcers Rob Hatch and Sean Kelley saying "oh no not again" is in reference to his past history of disastrous things happening to him on in stage races. But he was a champion pro ski jumper before he switched to pro cycling, so no matter what happens it's always "next jump". No time to worry over what was. He's multi Grand Tour winner anyways. That he didn't grow up a bike racer he has no fear of ditching "pro cycling tradition" because it's "next jump", all about  embracing The Present.

Sean Kelley saying "we've had the discussion of the one ring, but let's not get into that now" . He was a pro racer in the Lemond days and has a very dry wit about him, he can be very funny !

So if Drew said SRAM only has the 10-44 then that's what is was. If you ask most racers what their preferred ring is to ride all the time it is inevitably the big ring. A 44t for an uphill TT qualifies as a big(enough) ring. I don't know exactly what it is about riding big rings and big cogs but it just feels more efficient than a small ring. Pros will often ride up mountains in the big ring and largest cog, which can be anywhere from a 53/34 to a 50/28 and the like. I know on my road bike my favorite uphill combo is a 46/28 for as long as I can take it. I have a 32 max but I guess I'm too cautious to ride that !

[lconley]

Larger diameter cogs and chainrings have less friction because the chain does not bend/wrap as much. Not sure how the angle between the front and rear impacts the overall friction. But lets say you have two single speed bikes - one is a 52-26 and the other is a 26-13 - same gear ratio. The 52-26 has less friction, but the 26-13 is lighter.

Laing

[Ted Durant]

On Wednesday, May 31, 2023 at 12:57:57 PM UTC-5 lconley wrote:

Larger diameter cogs and chainrings have less friction because the chain does not bend/wrap as much. Not sure how the angle between the front and rear impacts the overall friction. But lets say you have two single speed bikes - one is a 52-26 and the other is a 26-13 - same gear ratio. The 52-26 has less friction, but the 26-13 is lighter.

 

That's a good set of secondary considerations to my original thoughts on gearing. Along with making it easier to "optimize" steps between gears, larger cogs have lower drivetrain losses and, because there is less force exerted on the teeth, they wear longer. There is certainly a weight trade-off there. Modern, narrow spacing chains seem to handle cross gearing pretty well, but I'm sure that marginal gains fans can tell you what the power loss is at extreme angles. That seems to be one of the things favored by 1x fans - that your chain line is (if setup correctly) in the center of the rear cluster and never at the extreme.

I say secondary, but for some riders those factors might have a higher priority than they do for me.

[Nick Payne]

On Wednesday, 31 May 2023 at 11:23:36 am UTC+10 Garth wrote:

I was watching the uphill penultimate TT road racing stage of the Giro D' Italia this weekend where the eventual winner Primoz Roglic chose to run a 1x for his TT bike. The announcers were "praising" him for the "innovative" choice of using a mtb gearing setup for the TT. Well lo and behold, at a low speed(which is fast for mere mortals, hah hah) he was about 2/3 up the course and he ran over what appeared to be a small break in the road crosswise and his chain derailed off the ring. He put it back on cool as a cucumber and still trounced everyone, but so much for 1x's not failing like any other system, because they do.

I'd say he was either not using a clutch derailleur, or had the clutch disengaged to minimise drivetrain losses. My wife has been using 1x drivetrains for years with clutch derailleurs, on all sorts of road surfaces, and has never had the chain unship. She uses SRAM GX Eagle 1x12 on one bike, and Eagle AXS on another.

Nick Payne

[Garth]

Well here's about the best article I found on the many possibilities of why his chain flew off the big ring going over such a relatively small bump. The bottom line is that nobody knows. "Stuff happens".

My best guess is a wonky chainline part, as his model of bike is not spec'd as being suitable for 1x. These bikes require one very specific "left side only" asymmetrical type of BB, and it's not spec'd for 1x use. Yes, high end bikes have gotten very very specific of what can and can't be used on them.

https://en.brujulabike.com/roglic-single-chainring-problem/

In the article SRAM "orbit" dampening system is mentioned. From SRAMs website it's not something the rider can turn off or have any say in it's function.

Orbit

Reducing chain bounce is a crucial need for today’s riders that frequently venture off smoothly paved roads. SRAM’s Orbit™ technology, used in SRAM RED and Force eTap AXS rear derailleurs, limits chain bounce by controlling the chain’s downward movement when riding over bumps.

Orbit is a new, patented design that consists of a silicone fluid damper with a one-way valve that controls the rear derailleur cage’s counter-clockwise rotation. This prevents the lower derailleur pulley from moving forward when bumps in the road cause g-force spikes that pull the chain downward. The damped cage rotation keeps the drivetrain quiet and helps prevent the chain from derailing off the chainring.

The beauty of the lightweight Orbit fluid damper is that it is speed sensitive—meaning, the harsher the bump, the more resistance it applies to keep the chain under control. So, when the derailleur cage rotates counter-clockwise slowly—as it does during a downshift to an easier rear—the damper offers so little resistance, it’s as if the damper isn’t there. This allows you to get the benefits of a mechanical friction-style clutch when you need it, without the excessive chain tension these clutch types provide when you don’t. It’s the best of both worlds. And it’s a maintenance-free design so you spend less time wrenching and more time riding.