While Sram would like you to use their own off-road changers and MTB levers this will not provide compatibility with Campag's Ergo road levers. So, yet again, the problem arises of finding a suitable Campag or Shimano rear changer which is compatible with 11-36 teeth. The big sprocket is the problem. Using any "normal" 11 speed road changer is likely to cause a collision between the cassette teeth and the top jockey pulley regardless of cage length. Push the changer away from the cassette with an extended hanger or foolishly extended B-screw and the chain is now too far from the smaller cogs to change crisply.
I'm presently running an 11-32T cassette with an Athena RD for a 29" bottom gear. For some reason this provided a much nicer change than the previous Ultegra. Campagnolo claim a maximum sprocket capacity of 29T but then introduced 30T cassettes. Just to show how little each department knew about their own products. Or, rather, what they were actually willing to admit based on their current catalogue. When in fact very many cyclists mix and match equipment to suit their own specific needs or budgets. The hype department is probably continuing their long term office feud with the R&D dept. over some suspected slight back in 1931.
The present trend towards much lower gears has missed Campagnolo by several hundred miles, or rather, a few thousand kilometers. Probably because they pretend to only provide "road racing" equipment. Normally they would supply their kit [give it away] to professional team "he men" who can spin 55x11 all day long [without drugs, of course] up the steepest mountain passes without actually drawing breath. Except, perhaps, to exhale their [expensively hand-rolled] cigar smoke. Well, millionaires have to spend their fortune on something apart for their WAGS and mansions. Even if it's never on the actual bikes they ride between their frequent falling offs.
Sram and Shimano have long supplied both the road and MTB crowds, of course. This has given them far greater insight into wide range gearing in both theory and practice. The present limit being cassettes with 10-42T sprockets intended for a single [front] chainwheel. [Though not exclusively.] Ideal, one might have thought, for the serious cyclo-tourist on mountainous routes? Except that the MTB changer pull ratios remain incompatible with road [dropped] handlebar "brifter" type levers. Keen tourists usually prefer dropped bars for their wider range of potential hand positions than the MTB's [nominally] straight 'bars. Which can quickly cause agonizing wrist pain unless equipped with right angle "bar ends."
The 11 speed sprocket spacing is close enough to avoid indexing problems as has been confirmed by the expert mechanics with assorted machines to test in practice. So swapping 11 speed wheels between Campag, Sram or Shimano equipped racing machines is as simple as setting the rear changer travel, stop screws. For the first time in many years they [the manufacturers and money printers] have been forced, by the limits in maximum cassette width and wheel dish, to finally match sprocket spacing with their 11 speed offerings.
One poster, on a bike forum, suggested the Shimano 105 11 speed RD has considerable extra sprocket capacity. This is not confirmed by the Shimano hype/techspeek literature though. However desirable it might be to get an order off for an 11-36t cassette it would be a tad premature unless I can source a suitable rear derailleur. I seriously doubt the Athena could cope. Why, oh why, do the major accessory manufacturers insist on playing tyrannical dictators to their own customers? Haven't they heard of the Arab spring? Their sales could rocket if they introduced far greater cross compatibility between their own MTB and Road components and far more friendly, lower gear options.
(28x33)/36 would give me a 26" bottom gear compared to my present 29". Assuming a 90rpm pedal cadence, while climbing, that reduces my bottom gear speed from 7.3 to 6.5mph. This may not sound much but is quite a big deal on a long or steep climb if one wants to maintain enough revs to avoid struggling.
BikeCalc.com - Bicycle Gear Speed Chart
I am stuck with a minimum 33t inner chainring on the Spa Cycles XD-2/Stronglight/Sugino/TA double chainset. I have no desire to change to a fuggly MTB chainset or go back to a triple now. Having tried a number of front triple changers I have lost my appetite for clumsy changes under load. A double gives much smoother steps between gears provided a wide ratio cassette is available at the rear.
The triple always feels like too much of a downward step. Which usually causes over-revving until one's speed has dropped to match. The double chainset, in comparison, gives progressive reductions in speed provided one is already on the small [inner] chainring. So one's speed can be maintained. The step down from 43 to 33 is also much more comfortable than dropping from the largest to smallest ring on a triple. With no middle ring to overshoot in either direction as a hill rears up or drops away! Rear gear changing is, and always has been, much more pleasant than at the front. I have no idea why, except for the relative lack of low gears, why I insisted on sticking with triples for so long.
Nobody [normal] ever begs for a higher gear! While the vast majority of all cyclists need and want lower gears at least some of the time. Giving them the option would actually help to increase the number of cyclists and therefore potential customers. This must surely appeal to the cycle accessory manufacturer's sense of obscene greed for blatant profiteering? The average gear ratios offered by typical new road bikes is a slap in the face to anyone unfit and getting into the sport/pastime if they live anywhere near the slightest incline. It's no wonder MTBs are so popular!
After failing to be [at all] pleased with the Jtech Shiftmate, double pulley, pull-ratio changer, I had a better idea. A simple lever pivoted near the bottom bracket would anchor the cable from the downtube run. The rear changer cable would also be anchored to the same lever but at a different radius. This incredibly simple arrangement would change the pull-click ratio of the cable itself. Whatever difference in pull was required, either up or down in ratio, the lever would provide by changing the radius of each cable clamp. Suitable pre-drilled holes might be preferable to sliding clamps.
This rather rough drawing shows the basic method of changing pull-click ratios but is neither [remotely] to scale nor reflects the likely finished design in any way. The bellcrank pivot point and relative lengths of the lever arms are open to considerable variation.
The cable clamps would need to be backlash free but [probably] able to rotate to avoid distortion by local bending. Which might eventually cause fraying through work-hardening the cable strands at the clamp. Perhaps anchor rotation is not really that important? Only an actual trial will confirm this. The advantage of the ratio changing lever is its absolute simplicity. Hidden behind the chainwheel there is no reason for it to be visible. Any gear change lever could become instantly compatible with any rear changer simply by adjusting the cable anchor point to a specific radius. Though the actual number of gears could never exceed that provided by the indexed lever itself.
If the lever were made long enough then small changes in radius, through lever tilt, would be minimized. The arc would make the pull-per-click decidedly non-linear with a short lever at the limits of bellcrank travel. Depending on the lever support clamp position and the shape of the lever itself, it could act like a seesaw above and below the bottom bracket. Or entirely above or entirely below the bottom bracket to taste. Or even pivoted on the chainstay via a suitable clamp. The cable dressing arrangements would remain identical with a normal rear gear changer and its usual rear loop. The inner gear cable would merely be parted at the bellcrank on a machine with normal, bare cable runs. Each half would be fixed by its own cable clamp to the bellcrank.
An inverted V-shaped lever [or bellcrank] makes good sense. Since the change in cable direction, at the lever, wants to match the downtube angle relative to the chainstay angle. Some rear cable anchor points are above and some fixed below the chainstay. This would tend to make the ratio changing lever machine-specific to avoid the cable rubbing on the chainstay. Longer lever "arms" would make radius choice both ore accurate and easier to achieve with less cable distortion. Fortunately, my foolish invention is hardly likely to register on the Chinese patent copy industry's radar. Nobody else would be daft enough to try! But how else am I to overcome the deliberate failure of the cycle accessory monopoly of two in having low enough gears to climb all those snow-capped, Danish mountains? Watch this space. ;-)