Russ Roth

Junior gearing would be my guess, up till last year Jrs were restricted in their road gearing. Often a 14t cassette was a solution to make it work. I built my kid's bike with a 40t 1x and 11-34. That 40/11 squeaked the rollout by a hair and better than the more common which was I think 50/14. But not all gearing is the same and different brands can produce different roll out which is why distance and not gearing is the determiner for passing. Her current track bike technically has the wrong gearing, the chainring is 1 tooth "too big" but has now passed rollout in 3 state races, the NE regionals and one Nats race.

Duragrouch

Good explanation, but I would debate the bolded above; Chain driving toothed cogs is about as discrete, dare I say digital, as it gets. These are not frictional belts. Perhaps as you get to very small cogs, if not good teeth profiles, so less smooth tooth transitions, you can get very slight rotational speed variations, but in terms of displacement per rotation, it should still be same. Discuss.

Anything affecting tire rolling circumference, different story, that can change rollout for same gearing.

79pmooney

I went triple when I quit racing and ordered my Mooney non-racing bike. My racing gears plus a 28 tooth inner ring. For the next few years, freewheels went from 5 to 7 and I simply added the new cogs as the next lower. Then I aged some more and I dropped the inner, eventually down to 24. Followed this pattern on my other geared bikes too. When I got my custom 16 years ago, I went 9-speed with as small as 12 teeth. I aged and dropped the outer and middle rings to 50-38. (Campy cassette so I've never had bigger than 28 teeth. I go 28, 25 or 23 depending on my conditioning and the hills planned.) My two 7-speed bikes have 13-26 FWs and 52-42-30 (Campy on an old racing bike and 52-42-24 on a Raleigh Competition. (I need to get smaller rings.) The race bike is just that. I suck it up and grunt.

Not all that different from ridelikeaturtle's friend's bike. (I ride Portland so having both rather large and low gears is good. And I still hate bigger than 2 tooth gaps. Still love the one tooth gaps.)

Russ Roth

When you're dealing with Jr gearing, now only in track, a minute difference in the rear cog can mean the difference between skating under the line or over it. But even chainrings can have a small difference. Remember, the rear turns a little over 3x for the full rotation of the front that determines rollout. Tires matter as well, which is why many parents in track try to get their kids set up in the spring and not touch the bike again. It is possible to take 2 different 14t cogs and hold them against each other to find out the valleys don't match, whether different brands, different quality levels, or even just which was made when the tools are more or less worn. When you're dealing with number like 6.45m or 6.93m, your junior is already strong enough the gear is a hold back, and that 1/2" difference matters you test out different cogs and rings. For road, where it no longer matters, the difference of a 25c tire, something that was gaining traction in 06 as a racing size, the swap from a common 50t to a 49t might have been needed to avoid disqualification. Last year, in the first road race my kid did, practically half the kids were in danger of disqualification, but since the Jr's were racing with adults the officials decided to give the kids who went over rollout a time handicap to make it fair for the other Jr's. Month later the road rule went away and all the kids ditched the 14t for 11s. Had to deal with a parent at states who was dealing with it last year, his kids had the same bikes, same gearing and one failed, he'd bought 2 different 16t cogs due to stock issues, you had to hold them against each other to see the difference which was very small but at 16 valleys times 3 rotations the 1/4" difference means one fails and the other skates.

tomato coupe

As stated by Duragrouch, the chain essentially "digitizes" the drivetrain, so any two 16 tooth cogs will give you the same rollout. Any difference in measured rollout comes from the tire circumference.

Russ Roth

And both of you would be wrong. The chain doesn't "digitize" anything. Toss on 4 different 16t cogs from different brands and see what the results are, high end brands will give you more consistent results, cheaper brands less consistent. For a deeper dive look into bmx fractional gearing which over emphasizes the differences, but when you have kids who are trying to get that 1/2" longer gear inch a 16 1/4 cog can do the job. There's no 1/4 tooth, they've just cut the valleys shallower to add the difference. Example, https://identifybmx.com/products/5-bolt-fractional-gear BMXers get deeper into it than any other sport and the chain isn't digitizing anything. The same issue exists accidentally when trying to keep a kid's bike within an allowed limit. When they say 19' 10 1/4" they really mean it. Absolutely, tires make a difference, but so does the rest of the parts, it might not even be a quality difference but simply that different manufacturers have different standards on where the valley sits, but that's what really determines the real diameter of the cog.

tomato coupe

Every 16 tooth cog will rotate 360 degrees with the passage of 16 chain links, regardless of the depth of the valley between the teeth. Likewise, it will rotate exactly 1,000,000 times with the passage of 16,000,000 chain links. It is not possible to get anything other than 1,000,000 rotations with the passage of 16,000,000 chain links unless the chain skips.

Russ Roth

That would be correct but has no bearing on the actual distance traveled. If you have a cog with 16t and a diameter of 4" (made up number, didn't measure) and a 16t cog with a diameter of 4.15" the there will be a rotation of 16 teeth but the chain will still have traveled a different distance to travel that same 16 teeth. As an extreme example, if you were to cut the valleys shallow enough to match the bottoms of a 17t cog, besides being hard to avoid skipping, the chain would still travel a distance equivalent to a 17t cog, which means the bike would travel the distance of a 17t cog. And we're not always talking about a tiny bit. Take a Dura Ace cog as the average and hold one up that isn't cut as deep and another that's cut more shallow and the difference between the two can be a several mm in diameter and that matters.

tomato coupe

Sorry, but your understanding of how chain drives work is incorrect.

urbanknight

ok I goolged fractional gearing to understand what the heck they're talking about. Based on the explanation and demonstration videos, you should be able to create a similar effect by simply moving the rear wheel to pull the chain tighter or looser.

The fact remains that if you turn a crank with a 45t chainring one rotation, a wheel with a 15t cog will turn 3 rotations. Even if it really did change the force required, the rollout would be physically the same. It's also obvious that if you make the valleys deeper or shallower, the chain won't mesh as well.

Yan 

Imagine a simple scenario where the chainring and cog have the same number of teeth. Let's say a 24t chainring driving a 24t cog. It's a 1:1 ratio, so for every one time your leg goes around, the rear wheel also revolves exactly one time.

You're saying that if you install a poorly manufactured cog, let's say for example a cog that is still the same 24t, but slightly smaller in diameter than the correct size, then under your theory, the wheel will over time slowly overtake your leg and somehow make additional revolutions?

That's wrong. The cog might have poor tolerance and be slightly different diameter, but as long as the chain is not skipping over the teeth, then the 24t chainring and 24t cog are hard locked together. If the chainring goes around one time, then the cog will also go around one time. Nothing can break this 1:1 hard lock. The only way this hard lock can be broken is if the chain is skipping. At that point your bike is simply malfunctioning and you need replacements.

And what happens if you intentionally make a scaled down 24t cog that is the same size as a 23t cog, yet by some miracle of tolerance the chain still manages to barely cling on without skipping? It doesn't matter. Your wheel will still turn exactly one revolution every time your leg goes around.

Yan 

Back in the day I used to tour on a 48-38-24 triple crank with a 11-34 cassette. The 48 front / 11 rear high gear is very fast, while the 24 front / 34 rear low gear could get me up pretty much any slope.

On your bike the smallest chainring is 30t and the largest cog is 26t. That's not very low at all, not even reaching 1:1 ratio, so it would not be suitable for climbing with a loaded touring bike.

​​​​​Personally I think the previous owner simply had a drivetrain that was not very well thought out.

I stopped using triples a long time ago. These days the fastest gear on my touring bike is 36 front / 10 rear (equivalent to 40 / 11).

Kimmo

10t jumps were the rule before big rings were elaborated on to include the vast improvement of shift gates and lift pins, which brought the unfortunate restriction of being specific to a particular-sized smaller ring; the improvement to shifting being largely negated without the correct jump for the gates and pins. This innovation likely paved the way for larger jumps becoming more feasible.

Kimmo

Or probably the same for new vs worn. But as folks are trying to convince you, there's no way it can't average out the same; the only way the ratio of crank turns to wheel turns can vary from the ratio of ring teeth to cog teeth is if the chain can jump over teeth.

The only possible variable which can affect rollout (over a sufficient distance to allow discrepancies introduced by roller/tooth slop and chain wear to average out) is the tyre. The effective circumference of the tyre is going to vary by nominal size (obviously), brand, pressure, temperature, bike and rider weight, tread wear and any damage to the carcass, and possibly variation in diameter of the bead seat on the rim if clincher, or glue buildup on tubulars, which might also vary slightly in size from rim to rim.

​​​​​​Think about it. Every turn of the crank is x number of teeth, every turn of the wheel is y number of teeth, regardless of the fit of chain on cog. At the end of the day, the only way it can't add up is if the chain jumps. If the distance used to measure rollout is too short, the measurement will show noise from poor drivetrain tolerances.

Kimmo

As I said earlier, it seems rather specific to be an accident; I'd suggest it only looks strange because you're not privy to the previous owner's evaluation criteria.

As an extension of this thought, I'd also suggest it's far better to err on the side of humility rather than hubris; that's a much more productive and enlightening road. For instance, perhaps the PO was tripping balls with this gear selection and was not achieving their desired effect, but you realise something interesting and useful while trying to understand it, after assuming it was somehow sensible.

Duragrouch

OK, in opposition to my own stated logic, thinking real hard about this, I may see a possibility of non-exact-digital tooth-for-tooth rotation. Or I could be wrong. Follow me here... Let's start with the chain, and you have a chain that is greatly "stretched", meaning a lot of wear at the pins and inner plates, so that the distance over 8 links is slightly more than it should be; It's wrapped about 180 degrees around the rear (16 tooth) cog; The link on top is fully engaged with the tooth it is pulling on; Each subsequent trailing link is incrementally a bit looser, has the tiniest gap between the roller and the tooth in front of it; So as the chain advances under drive torque, by the time the 8th link passes off the last tooth on top, if staying fully under drive torque (tension in chain, resistance to rolling by the wheel), by the 8th link, the cog is very slightly lagging in rotation, and with each subsequent link, you get cumulative error. Hmm... I'm not certain about the above, but thinking about it. And I can see the same source of variation being due to varation in the cog. So I think it depends on the relationship between the two, and the key being, how much rotation of the cog for each linear passage of the chain, under drive load? If each chain roller passing over the chain is perfectly centered in the valley between teeth, I think you get no variation in rotation, because the angular movement is perfectly synched. But, if that is not the case, the top link pushing on the tooth in front of it, and chain or cog relationship such that each following link having an increasing gap in front of it until its tooth reaches the top, under constant drive load, then I could see having less rotation per given chain passage, assuming that is constant, driven by crank rotation. For the above to happen, there must be looseness, i.e., variation in pitch, between chain and cog. I could be wrong. I'm just trying to offer the chance that I was wrong previously.

Discuss.

ridelikeaturtle

Has ChatGPT taken over this thread?

Duragrouch

In a good way, or bad way? Debating for the sake of debating, bad. Thinking with the ability of a supercomputer of all possibilities, not bad.

tomato coupe

You weren't.

Duragrouch

You might be right, that I was originally right. What I wrote above may make sense in the short term, like 8 or 10 links passage, but over the whole chain length, it has to average out, unless, there is such a difference in chain and cog pitch, that somewhere, you gain or lose a tooth, otherwise, the ratio of crank to cogs remains constant. And that would only occur if the stretch in chain over 180 degrees around the cog, is at least half a pitch, so a roller is up on the tip of a tooth or further back, and I think that is not going to happen. The only question then is, for a short travel of chain, can there be enough variation in chain displacement vs cog rotation, to make a difference in the rollout test? No, you're right, it just occured to me. As each link comes off the cog at the top, the link following immediately takes up any slack and everthing is in synch. Just the cog wears more because the chain is only pulling on one tooth at a time. So yes, constant ratio.

"I thought I was mistaken, but I was mistaken."

bbattle

I was wondering what is the capability of the rear derailleur on this bike. Does it have a short arm or long "touring" arm to handle the chain wrap. This wasn't a stock setup, that's for sure.

rm -rf

Back to the odd gearing on that bike:

My Campagnolo Athena 11 speed crankset is 30-39-52. I rarely use the 52, since the 39-11 reaches 25 mph around 90 rpm. A 49 could be useful.

That cassette 14-26 cassette was a "junior racing cassette". There were restrictions on the highest gear ratio for younger racers.
I have a 11-speed "junior" 14-28, which is 14-15-16-17-18-19-20-21 and 23,25,28. I liked it with my 34-50 Di2 bike, for some fast group rides I used to do, with very close shifts from 18 to 24 mph, right where I was trying to hang on the group.

The downside of that 14-21 straight block is that the shifts are too close together in a 30, 34, or 39 chainring, often needing to shift two cogs. (That worked okay with my Di2's easy, fast shifting.)
And this 14-26 doesn't have a really low gear for steep climbs. My 30-39-52 usually has 11-speed 11-34. That 30 front - 34 rear is great!

rm -rf

The odd 14-26 with the 30-39-49 at typical flat road cadences.
Extremely close shifts, but this needs lots of chainring shifts in the 10 to 20 mph range. Annoying.


~~~~
Compare to the 30-39-49 with a 13-28. (Campagnolo liked to use the 13 as the smallest cog in 9 or 10 speed cassettes.)

The 39 has a better, wider range of speeds. And the low gears are better too.
Compared to a 53, the 49 lowers the big chainring shifts about one rear shift. That's pretty good.

njkayaker

What is wrong with you?

If a “big comeback” doesn’t mean more people are buying something, what do you mean?

There is nothing that indicates a wide-spread increase in the popularity (“making a big comeback”) of triples.

You didn’t ask about it being a “common thing” and running the numbers through a gear calculator might provide insight into the “logic”. Again, what is wrong with you?

Expecting that a 49 chain ring “which look to have been replaced over the years” on an old bike (one data point) might be “common” (or that people would guess that was what you were asking) is pretty “out there” too.
​​​​​

You asked for “thoughts” and then have ridiculous complaints about them.

PromptCritical 

What did you use for a FD when you made the switch to a triple?