ThermionicScott 

The title of this thread can be read both ways: that it's a peculiar question, and that it's a question about some C&V gear choices that I've found puzzling.

Perhaps I'm spoiled by modern freewheels and cassettes with fairly even jumps between gears (excepting special-purpose ones like MegaRange and their clones), but I've noticed a few examples in the literature where even if a rider probably could have chosen more even jumps from the available cogs, they didn't.

Jean Dejeans' 1948 PBP bike used a 16-17-18-22-24 freewheel per https://www.flickr.com/photos/strong...7618749510933/
Gear chart: Bicycle Gear Calculator

And Charly Gaul (1958ish) chose a 15-17-19-23-26 freewheel per https://web.archive.org/web/20140706...gauf-mont.html
Gear chart: Bicycle Gear Calculator

In both cases, the gears are relatively close on top, and on the bottom, leaving quite a hole in the middle.
Was there a general gearing/riding theory back in the day that would call for this, and is now history?

Thanks.

P.S. On a sad side note, I noticed that the domain for Fast Eddy's Flandria Cafe must have expired and been purchased by some random travel/food blogger. The writing makes me think it may even be computer-generated content.

bikemig 

Charly Gaul had a bog simple cross over set up with 6 very usable gears from 45.3 gear inches to 87.4. He had a 49/44 up front. This set up makes a lot of sense since it involves very little thought (a big plus when doing PBP!). You run three gears on the big chain ring, then flip down to the small and get 3 more. The technical limits back in the day were that you only had 5 on the back and those many RDs couldn't handle much chain wrap or too big a cog.

My guess is that he avoided half step gearing which would have provided more usable gears because he prioritized ease of shifting over maximizing usable gears. Half step gearing means a lot of double shifts.

Looks like Jean DeJean also used cross over gearing with a "wide" compact up front. I like it. I'm doing something similar on a vintage build with 48/32, 12-30 in the rear.

That gap is likely the price you pay for having 5 in the back. I think they had cruising gears and climbing gears and lived with a donut hole to get the gearing they needed at the top and bottom.

noglider 

I think you're assuming that decisions were made rationally. It could very well be that some people didn't realize that percentage changes matter more than number of teeth changes.

corrado33

For a long time I've wanted a 6 speed freewheel that had a fairly large gap right in the middle, with 3 "easy" gears and 3 "hard" gears. The easiest of the hard gear would be used on flat ground when pedaling into a headwind, and the hardest for when you're pedaling very hard on flat ground with no wind. Maybe even 2 really easy gears and 4 hard gears.

The easy gears would be for climbing or towing.

Like a megarange except with 3 easy gears instead of 1. (Or 2 easy gears... depending)

Something like a 14-15-16-18-24-28. With a front crankset selected so that I would almost never fall in the hole in the middle.

noglider 

I've tried unevenly spaced gears and found I didn't like it. I suspect others will find the same, once they try it. Basically, I want my next shift to be 10% from the gear I'm currently in. If I want a 20% change, I'll shift twice.

bikemig 

I dunno. These are very experienced bike riders who had their pick of gear to do a very demanding ride. They had their pick of gear. I think the operating assumption is that their gearing and equipment choices were rational.

Sir_Name 

My understanding is that small jumps in tooth count at higher speeds require more effort than the the same jump at lower speeds. With a limited number of sprockets to spread wanted gear range across it makes sense to space out the lower ratios and more tightly pack the higher ratios for a given terrain.

noglider 

If you're talking about the fact that power is proportional to the cube of speed, because of wind resistance, then it makes a little sense. The trouble is, cogs have to have a tooth every half inch, so we can't get the granularity we want. All gearing systems present some kind of compromise or another, and usually more than one.

79pmooney

We are talking about the gears of pro racers. I suspect they chose gears that they felt would serve them best in the conditions they raced. Those conditions are set by not just the road gradients but also big time by the way the peloton rode them. I'll postulate they knew the big action of their races happened on major climbs and late in races at high speed on the flat. Having a working choice of gears at key moments might very well be make or break. Not having the "right" gears for the many miles of slower miles would have been less than ideal but quite possibly worth the sacrifice to have the key gears when needed. (If their competitors had the same philosophy, there might well have been a general "detente" for most of the race as they all rode poor gearing, much like, in the years before derailleurs, the peloton held informal group "timeouts" to stop and flip wheels around to go to mountain gears. When one competitor faked a flat, stopped and turned his wheel around before the timeout, caught up and kept on going to gain 2 minutes and ultimately the Tour de France, it was widely felt that he cheated.)

Ben

noglider 

Well put, @79pmooney. I think this means that for the rest of us, doing less specialized and more generalized types of riding, standard gear configurations make more sense for us.

bwilli88 

If you changed Charly Gaul's 23t to a 22t you loose the hole in the middle. If you use that with the 48x32 it is a great spread gear setup.

John E

Schwinn Paramounts had a similar gap that I never liked:
52-49 / 14-16-18-23-26.
Nothing a 6-speed freewheel can't solve. I run 14-16-18-20-23-26 on the Bianchi, with either a 50-47 half-step for flat rides or a 50-42 1.5-step for hills.

Trakhak

@bikemig's post (no. 2 above) explained the reason for the gear spread perfectly; speculations about whether or not the riders using those gears were ignorant are unnecessary.

My first 10-speed bike was a Raleigh Blue Streak---got it for my birthday in 1963, I think. It had Cyclo-Benelux derailleurs that were very limited in their chain wrap capacity. They worked best using the outer chainring with the middle and outer two sprockets and the inner chainring with the middle and inner two sprockets.

Given that kind of compulsory shifting pattern, Charly Gaul's bike's gear development described in the first post clearly makes sense, with two two-tooth jumps followed by two three-tooth jumps.

And Jean Dejeans's PBP bike's one-tooth jumps in the smallest sprockets followed by an abrupt jump to the largest two sprockets also make sense if you know that the Paris-Brest-Paris race was by far the longest race on the European calendar and, with the exception of a few sharp rises, was nearly dead flat.

(How long was the race? In 1948, according to Wikipedia, the winner finished in 41 hours, 36 minutes, 42 seconds.)

Grinding along in a high gear was what they did in those days. Remember, most of the racers were coming from a tradition of racing through the Alps and Pyrenees with single-speed bikes.

ThermionicScott 

Some good thoughts in here, thanks folks.

jimmuller 

That's what I'm thinkin'. When something seems out of the ordinary it may mean the builder didn't understand, but more likely it means the builder did understand to a level deeper than generic. The even spacing on other bikes, mass-produced or not, was meant to handle a range of unknown situations, but a racer might know exactly what he needs and what he doesn't. F'rexample, what if you are racing in the mountains where you are either going up or going down, but rarely run flat stretches for any great distance?

Warning: Potentially obvious statement coming.
A point about the significance of one tooth at high or low speeds: It is more than just that power goes up as the cube of speed and wind drag goes up as the square of speed. It is that one tooth is a bigger fraction of the total when the tooth count is lower. One tooth on a 14 is twice one tooth on a 28. That's why the gaps on nominally evenly-spaced FWs always get bigger toward the big end. Like I said, it's obvious.

repechage

I like Gaul's '58 set, the 26 is the bailout in my view, I trained on 15-23 for years, always liked the odd numbered cog set up when one only had 5 in back.

John E

Thumbs up on that. I use a 6-speed version of the same odd-number progression, 13-23, with a 1.5-step setup on the 1959 Capo. A similar sequence is a the core of the 8-speed cassette my mountain bike: 12-13-15-17-19-21-24-28. A 2-tooth development in back lets you shift up through the gears very efficiently as you accelerate, and with a 1.5-step or a half-step you can then bring in the front shifter to fine-tune at cruising speed.