The Chainlink

Does anyone know *exactly* the gearing ratios in a Divvy?

I'm trying to spec a single speed for a friend for whom Divvy's 3rd gear is a good reference point for just a bit too easy, whereas my fixie's a bit too hard. So I know the upper bound is 44:17(2.58) but I don't want to guess Divvy's.

Also I've also often heard (and told) the story that Divvy's first gear is so low that to experienced cyclists it feels like the chain's fallen off, so I'm wondering exactly how low that is.

Anyone with Divvy maintenance experience got the specs?  Either the three ratios, or the teeth counts would be nice.  I'd also welcome weight and any other details.  A few cursory web searches found little.

Views: 2401

Reply to This

Replies to This Discussion

Agreed.  I hope they're aware, and through discussions like this reiteration is made, that their lowest gear is useless.

Putting aside the tangential value to supplement planning, the ridicule of the righteous gods of bikecraft, and the pursuit of more detailed metrics, the simple ratios sought here are non-obvious and worth knowing.

I'll try to count and measure and weigh one this weekend.  Followup with more research and contact them directly.  Anyone wanna help?

I think he's looking for "gear inches," but whatever. For those of you keeping score at home, I did send an email to Divvy customer service on 4/24 @ 1:52 pm asking this question. I have not received a response. I'll post if and when I ever do.

THANX KEVIN!  I'd originally just wanted to know teeth on front / teeth on rear to start; all that's ever mattered on my fixie.  But now I wanna know all the specs, even it's aerodynamic efficiency quotient.  Anyone got a wind tunnel?

And good to hear about the other cities' hubs.

Yes, but is that supposed to address my comment? Did you even read my comment, pause for 10 seconds and think maybe you did misspeak? Or are you too confident for that possibility to cross your mind?  You act like you know all about gear ratios, tell people to get advice from a pro, a pro like you I suppose, do all that in a condescending tone, yet what you write is inaccurate and very likely confusing for someone thinking or learning about these things for the first time. And after I take you seriously and read your long post, and hint that maybe you said something wrong.. what do you do?  dump more wisdom, or should i say bullshit? And now what, perhaps you know a lot, you got lots to share, but why would I ever listen to what you have to say?  Why would anybody?

So frustrating, and disappointing..

 
Hey! Bike Shop Guy said:

The Sheldon Brown calculator will give you out put from all those numbers in gear inches, development or gain ratio as well as cadence at specific RPM.  Gear inches is the number I like to use.

ilter said:


Another interesting question is whether you simply misspoke or if you are truly confusing gear inches and Sheldon Brown's gain ratio :)


Hey! Bike Shop Guy said:

[snip]

Of course the more interesting question here is why, if you don't know any of this are you trying to set up somebody's bike?  Have them go to a shop and get advice from a professional before you end up guiding them into a bad choice because you apparently don't really know what the heck you are doing.  DIY is great and all but pleas people know how to do it before you start.

Strictly speaking this doesn't answer your question, but you can calculate the information you're looking for by taking two measurements using a tape measure and a piece of chalk.

You really don't care what the ratios are on the Divvy. You just want to set a lower bound for the ratio on your friend's bike. And what I'm about to describe may be exactly what some of the others have discussed, just without any formal terminology.

You can determine the lower limit gear ratio from a) the ratio on your bike, and b) a comparison of the operation of your bike and the Divvy.

So rent a Divvy. Unless you're a lot more coordinated than average, have your friend hold it upright by the bars, and giving some resistance to forward motion. Get it in the gear you're measuring, then cranking a pedal with your hand, slowly move forward until the crank arm is precisely aligned with the seat tube. Put a chalk mark on the ground in the center of the rear tire contact patch. (If you want to get extremely fussy about precision, tape a laser pointer to the bike aimed down, and mark where the beam hits.) Now crank slowly forward, making sure the bike never coasts ahead of the crank movement, until you've made one complete crank revolution and the crank is aligned exactly how it was when  you made the first mark. Make a second mark, then measure the distance between them.

Repeat with your bike.

Divide your bike's ratio by the distance measurement for your bike, then multiply by the Divvy measurement. This is the low gear ratio limit, for a bike with the same wheels/tires and same crank arm length as yours. This works regardless of Divvy specs - it could have 16" wheels or penny farthing wheels. That's irrelevant to this measurement. (If your friend is getting a bike with wheels, tire diameter or crank length different from yours, you'll have to adjust the number accordingly.)

The number you get will likely not correspond to a real ratio, but since this number is a lower limit rather than the desired number that doesn't matter. For example if your measurements gave you a result of 2.25, then that would correspond to 44:19.6.... It's already too low so round to 44:19 or 2.31.

The rest, as they say, is left as an exercise for the reader.

jeq

Thanx!  Yes, the point was to at least set a lower bound based on a too easy bike she'd tried (Divvy), and with an assumption of most geometry on the planned bike being very close to mine (a KHS UrbanSoul).  Yours is a great deductive approach.

We eventually got her this (http://publicbikes.com/p/PUBLIC-V1-2014) at BFF because it met most specs, was cheap, she liked a brief test ride, and we assumed the gearing was sane.  Soon after receipt though she began to complain it was too "spinny" and we found by simple counting that as shipped it was geared as 39/20=1.95.  A few dozen dollars later The Bike Lane put a 16 on it (=2.44) and she's much happier.

But I still hope to someday have full Divvy specs.

I finally got around to doing the measurements that have been suggested several times in this thread. I laid down a tape measure on the sidewalk, sat on the Divvy bike, and slowly performed one crank revolution (while holding the brake to prevent coasting, and keeping a hand on a wall for balance). I did this for each gear, and here are the distances traveled:

1st gear: 98 inches

2nd gear: 134 inches

3rd gear: 173 inches

The 3rd-to-1st ratio (173:98) equals 1.77, which is equal to the ratio of some (old?) Shimano Nexus 3 hubs (according to Sheldon Brown's Internal Gear Calculator), and a bit off from the 1.86 ratio on other Shimano Nexus 3 hubs. The 3rd-to-2nd ratio unfortunately doesn't match either hub precisely. 173:134 equals 1.29, which is lower than the 1.33 or 1.36 3rd-to-2nd ratio of either Shimano hub. Until I have reason to believe otherwise, I'll assume this is just a measurement error on my part of the 2nd gear distance, even though it's at least 4 inches too high, and I repeated it at least twice on different bikes on different days and got similar results (I admit I used no laser pointers or chalk, and merely leaned over the front wheel and used the front of the Divvy rack as a sighting tool to line up the bike with my tape measure).

To confirm my measurement by using a different method, I then counted the teeth on the front and rear chainrings. The front is more difficult than the rear, as I could only find one tiny window to view it through the chainguard as I turned the cranks. I came up with 30 teeth for the front, and 23 for the rear. I also measured the wheel+tire diameter to be approximately 27 inches, and thus I calculated that the circumference is approximately 85 inches.

Unfortunately this didn't confirm my first measurements. It seems that 2nd gear on a Shimano 3-speed is the direct-drive gear, so the external gears and wheel size should be the only factors. My own calculations (confirmed by Sheldon's calculator) show that (30 teeth / 23 teeth) * 85 inches = 111 inches. This is nowhere near the 134 inches I measured in 2nd gear, nor is it close to any gear. The numbers work out better if I assume the front chainring is actually 35 teeth, so while it would surprise me if I was off that much, it's the simplest explanation for this conflict. Otherwise the hub would have to be fitted with some additional multiplier I'm unaware of.

=============================================

Anyway, enough about the mechanical and mathematical issues. In terms of the real-world usage of the gearing, here are the calculated speeds in each gear at 100rpm (using the numbers from my original methodology):

1st: 9.3 mph

2nd: 12.7 mph

3rd: 16.4 mph

I've heard lots of whining that the gearing is too low. I disagree, and think the data supports me. What percentage of people ride their bikes in the city (much less the heavy, wind-jamming beast that is Divvy) faster than 16.4 mph? Very few, in my observation. So if you can't reach 16 mph on a Divvy, it's not the gearing that's the problem, it's you, and the solution is simple: pedal faster. It'll make you a more-efficient cyclist anyway. On top of that, since the 3rd-to-1st ratio is constant and fairly limited on any 3-speed hub, shifting the gearing higher would reduce the usefulness of 1st gear, which I use every single time I start from a stop (in fact, the gear I downshift to at every stop on my own bike is even *lower* than Divvy's 1st gear). It's super-helpful for getting you a jump on cars and clearing the intersection quickly.

If you want to calculate speeds at a different cadence, the formula is:

0.00095 * inches-per-crank-revolution * rpm = mph

So for example, in 3rd gear at 60 rpm, it would be:

0.00095 * 173 inches * 60 rpm = 9.9 mph

For what it's worth, this site claims that London's Boris bikes have 38 teeth on the front and 23 at the rear, with a Shimano Nexus 3-speed hub.  I don't know for certain, but I wouldn't be at all surprised if Divvy bikes and Boris bikes weren't identical models, both built by Devinci Cycles in Quebec.  The same article claims Boris bikes weigh 23 kilograms, or roughly 51 pounds.  Having lifted hundreds of Divvies, I don't find that at all surprising: they're beasts.
 
skyrefuge said:

To confirm my measurement by using a different method, I then counted the teeth on the front and rear chainrings. The front is more difficult than the rear, as I could only find one tiny window to view it through the chainguard as I turned the cranks. I came up with 30 teeth for the front, and 23 for the rear.

Ah, thanks. Yes, looking at photos of both bikes, they certainly look identical, including the front chainring size. Using a high-res photo of a Divvy, and using a chain pitch of 0.5in. as a known reference, I was able to "measure" the approximate chainring diameter, and it does correlate with the diameter of a 38-tooth chainring. The chainring has 5 ribs, and 38/5 ~= 8 teeth-per-segment, so maybe I only counted 4 ribs worth of teeth rather than 5.

The next problem apparently was that my measurement for wheel+tire was somehow too large. Measurement-by-photo shows the diameter is more like 25in. rather than 27in. The smaller dimension corresponds to a standard "26-inch" mountain bike wheel with a 1.5in tire (and that seems to be the standard size of Kenda Kwest tires, which is the name printed on the sidewall in the photo).

Putting all that together, and punching those numbers into Sheldon Brown's Internal Gear Calculator again, using "Shimano 3-speed (old) 333, F, FA, G, SG-3S21, G-3S23" as the internal hub, calculating "Meters Development" and converting to inches, it spits out:

1st: 98 in.

2nd: 130 in.

3rd: 173 in.

...which matches my first methodology quite closely. So I'm now much more confident that those numbers are the correct ones. Thanks for the help!


Thunder Snow said:

For what it's worth, this site claims that London's Boris bikes have 38 teeth on the front and 23 at the rear, with a Shimano Nexus 3-speed hub.  I don't know for certain, but I wouldn't be at all surprised if Divvy bikes and Boris bikes weren't identical models, both built by Devinci Cycles in Quebec.

This is AWESOME!!!
Thanx SO MUCH!!

You've done a great service to bikenerdkind.

RSS

© 2008-2016   The Chainlink Community, L.L.C.   Powered by

Disclaimer  |  Report an Issue  |  Terms of Service