Bike Gremlin

The way it's going now, I expect my (grand)son will live in a world where driving is forbidden as dangerous and all done by a computer controlled vehicles.

Cycling might also be labelled as not enough eco and also too dangerous (even now they force people to ride with helmets in many countries), so also forbidden.

Sy Reene

hmm.. I actually increased my pedal spindle lengths by a fraction and found my knees like it better.

Schwinn me

When I first got into the mechanics of bicycles at around 8 or 9 years old, the Schwinn 5spd stingray was the coolest thing around. I loved the looks of all those gears, the function and purpose was a secondary consideration but still very cool. When I got older I thought 15 speed triple bikes were cool. Didn't need all those gears in the relatively flat area where I lived but that didn't matter, I wanted them. Now I see 9 speed and more rear cogs, I think that's cool with the skinny chain and all and I'm in my 50's. I think a lot of us are still the young boys we once were and just love what looks cool. Complicated looking drive trains are just plain cool, if that floats your boat you should have it. Who cares if you need it or not.

noglider 

And then there is the simple beauty of a track racing bike.

Schwinn me

Completely agree although I would say that this is an acquired taste that comes with maturity. The opposite of elaborate, things boiled down to the bare essentials becomes very cool.

Jeff Wills

True, but my knees want to argue the point.

oldbobcat

On the front, half-step upshifts are easy since the chain doesn't have far to climb. On downshifts, the chain would undershift, landing between the rings. It wouldn't jam, of course, but the chain landed short of where the teeth could catch the chain.

I never thought much about the cause until now, so here is my theory. On a downshift the chain moves diagonally, down and sideways to the inside. The short drop to the inner ring means the chain is still moving sideways when it hits the inner ring. If the ring were smaller, the chain would have more time to complete its sideways motion before dropping down over the teeth of the inner ring.

Maybe a stiffer spring might have fixed this. I never bothered to research this more carefully. I just bought a 42t ring and started using single-step freewheels.

oldbobcat

When experimenting with spindle lengths you must keep the chainline in mind. Ideally, the centerline of the cassette or freewheel should be the same as the centerline of the chainrings. So in the chainline sense, a longer BB axle would necessitate a longer hub axle.

In real life, moving the chainrings out a millimeter or two might not make a noticeable difference.

HTupolev

Interesting. I've actually seen this behavior occasionally on the 53->44 jump on my 1979 Fuji America, but I've previously assumed that it was because the "8 speed" chain was too narrow for the vintage 6-speed crankset, and was going to try something like a KMC Z7 chain that excluded 8-speed in its description, to see if that solved the issue. Perhaps it wouldn't.

Retro Grouch 

Good point! One of the advantages being put forth for self-driving cars is that they will be more safe. A self driving car can never be distracted by it's cell phone, never gets drunk and never falls asleep.

Bike Gremlin

Until we get T1000, or 1984 (the latter one's already here).

Jeff Wills

This happened a couple years ago, but I wonder about the infinite scenarios that the "real world" will throw at self-driving cars:

https://www.washingtonpost.com/news/...=.6fb093edbbaf

oldbobcat

Pros have been running 53-44 and 53-46 for flat cobbled classics for decades, apparently without problem, and my first Campagnolo-equipped bike, a Raleigh International, came with a 52-45 chainset.

The cog width of freewheels, 5- to 8-speed, is the same, so inner chain width is identical. 7- and 8-speed chains are thinner on the outside, using thinner plates and a bend and/ or a bevel on the inner plates. I was running 5-speed with half-step. 7-speed hadn't been invented yet.

You might try loosening the low limit screw a little more than prudence would suggest That's something I've done for recalcitrant triples all the time.

HTupolev

So that we're on the same page, this is what I'm seeing happen when I shift from the 53 to the 44:



And the chains I'm using are modern "6-8 speed", so possibly a narrower outer width that what the bike originally would have come with. I'm thinking that if the outer was wider, it might force the chain to ride farther away from the big ring, possibly forcing it to slide entirely onto the 44T on its own.

It's a 53-44-36 triple, so overshifting on the 53-44 jump is already very doable, and it's how I resolve the issue when it comes up.

It's more of a minor inconvenience than a major problem, but I've still been curious about it.

dabac

That's why I said "comfy" - not "well pleased" or "fully satisfied".


I run 3x7 and 26" wheels on my main commuter. But never use the granny gear outside the snowy season.
Even then on my regular commute, there are only minutes out of each hour where I might need it.


So I probably have about 10 ratios that are different enough to matter.
And that probably covers 90+ % of my riding.


Once - if ever - 7-speed cassettes becomes rare enough to shift the price point advantage, going to an 1x setup will not be a big deal for my commuting experience.
Shorter drivetrain life, more service, that's all.


But since I'm not hugely inconvenienced by front shifting, I prefer my rugged, hard-wearing 3x7 over 1x setups for longer life and less maintenance requirements.

bud16415

3x9 half step + granny (45-42-24) X (12,14,16,18,21,24,28,32,36) for me

Same as a 15 cog cassette off the half steps only and another 4 gears off the granny for when you need them but I don’t count them into the riding gears.

The two easiest shifts on the bike are the 42-45 & 45-42.

Why would I want the chain to move over 14 or 15 cogs when I can do the same thing with 7 or 8 with a half step.

9 for me is the perfect number but a 7 or 8 with half step would work nicely also. No super thin chain or tiny teeth and a solid rear wheel design.

Bike Gremlin

Prefer 7 over 8+ for the wheel, but 7 speed freehubs are getting harder and harder to find (new).

As for gearing, yours makes perfect sense. Though, if at all possible, I'd prefer 13t, or even a 14t starting cassette, with similar chainring layout. 42T being very nice for flats IMO.

Reynolds 

Yesterday I rode 80kms in 3hrs on my road bike, on flat terrain. Only used 42/17, 16 & 15 solo, and about 15mins in 52/16 & 17 taking turns with 2 other riders. I think those 5 gears could cover 90% of my road riding.

davidad

My tourer is set up with half-step plus a grandpa. The front is 44-40-20 and the back is 11,13,16,19 23,28,34. This gives me 18 distinct usable gears. Most with a 10% spacing.

Jeff Wills

Huh. I think Frank Berto advocated that setup.

bud16415

I agree and with starting with a 9 speed rather than a 7 or 8 speed you could say my cassette starts with a 14t. The 12t is there for the 45t ring and a nice straight chain line with 100GI. That’s the gear I like for going downhill and keeping the legs warm and moving. Likewise, the 36t is there to go with the 24t granny for the climb anything seated 18GI.

So what I have in effect is a 7 speed centered in a 9 speed with no cross chaining off the perfect for me 42t center ring. I figured all my gearing around the center ring / center cog as being the center of my useful gears. That way all normal riding is done with the best chain lines. The half step is a simple shift when I need just a smidgen more or less. There are all kinds of ways to get the same range and the same gears (like mountain triples) but they split the range in two rather than weaving them together.

As to the OP question I would always take the double with the half step over new technology trying to cram more gears in the back. For me it is a trade off the super wide cassette is simplistic and easy to understand, but I would rather not have to skip over 15 gears to get from one end to the other and have everything finer and more delicate.

I have the above gearing on my tour bike that is also my all around bike but I wouldn’t mind having the same setup on a lighter bike as well.

bud16415

Your 44-40 half step allows you to go all the way down to a 20t on your grandpa gear with a 20 tooth drop on that shift. Where I have the 42-45 half step and was limited at 24 tooth with an 18 tooth drop. I bumped my big cog to a 36 and small cog to 12 where you have 11-34. So we are all most the same except I’m a 9 speed and you have a 7 speed.

In effect you have a lower low and a higher high and for touring that 16 GI compared to my 18 GI would be nice. I tried 15 GI once and spun out trying to stay upright. For me 18GI was about as low as I could stay upright

It shows you how subtle changes in a half step work and if my bike was strictly loaded touring I could see the exact same gearing you have being perfect for me as well. I have a little bit smaller steps is all and we have very close center / center gear yours being 40,19 (56 GI) and mine 42,21 (53,GI)

Kimmo

@HTupolev too!
Wait, no!

Instead of messing around with plate-sized cassettes, we can do something pretty cool with standard cassettes. See, there's a fundamental problem with derailer drivetrains: because the percentage jump increases as cogs get smaller, and there's room for many cogs but few rings, if you plot the ratios, the line curves the wrong way - you get wider ratios as you go faster.

I'm sorry, but that's just crap. This right here quite probably represents the biggest improvement to the derailer drivetrain yet remaining while we stick with 1/2" pitch. The 38/50/52 rings were carefully chosen; anything else in the neighborhood made a more jagged line with less even steps. The red is a standard 39/53, the blue and green is my setup, blue using the extra half-step ratios. Grey is unused combinations.



Here's rpm vs km/h:



I actually gave this a crack (link in my tag), but the main snag was that I really needed a FD with a cage designed for this, and there isn't one. Also, without a special crank, the load path for the middle ring was so horrible the flex affected shifting... not to mention I sort of lost steam after machining some pulleys for my servos, although by that point I had my sequential shifting code all sorted, all nicely based on interrupts. The mechanical engineering OTOH is extremely hard to do nicely with limited means

ThermionicScott 

Hey, good to see you @Kimmo!

Kimmo