Re: narrower bars, they felt good for one 40 mile ride so far. I'm off the bike for the next week in preparation for the Fickle Finger of Fate and its attendant PSA test, so it'll be a while before I make a final decision about cutting back the cable housings.

[MENTION=363941]hubcyclist[/MENTION] - if you're going to change the small ring, go with a 38, the smallest that'll fit on a 130 BCD crank. My eight speed bikes both have 13-26 cassettes...but I don't race them.

what's magical about 4.5 w/kg re: compacts?

fwiw, i am moderately strong, generally prefer to train on a compact (it just opens up a ton of terrain, esp when i want to keep power low) and race on a standard. horses for courses.

Because it is always, I mean always, about the ewang.

June 27, 2005 is the first use I can find of eWang on this forum. What made riders go before then?I also would like to know the significance of power to chaining. I've always liked bigger rings as it puts less tension on the chain.

I use 42cm marked bars which measure 41.x cm, something like 41.2 cm or so. I bought another same brand/model 40 cm, it measures virtually the same. Previously I had 43 cm bars outside to outside that measured about 41 cm c-c.

My shoulder width says to use 42s or even a 43. I've ridden 44s when I was doing long miles thinking I was a road racer (hahahaha…). I can't find it but there's a picture of me in here somewhere in a sprint in 2008 or 2009 and the comments went along the lines of "why do you have such a narrow bar?".

I could never figure out why people are so concerned with minimizing chain wear. They cost about the same as a tire and last about the same number of miles. Monitor the wear and replace as necessary. Maybe my eWang is insufficient?

There was a story about how Shimano had to increase durability of chains when they first came out with the hardened pin chains (for index shifting). Some pros like Phil Anderson were going through a chain in two weeks or less.

However I don't think it's about wear. It's about (probably) the perceived extra friction. In a pinch it may not matter, like I found myself in the big-big Sunday by accident (group ride, not a race, and I was already dropped). I quickly changed to a similar gear using the small ring.

It's also about risk. Someone on that ride lost their derailleur when it went into the spokes. Big-big means your derailleur is under all sorts of tension - derailleur maxed out to the inside against the in-out sprint, cage pulled taut maxing out the pulley cage spring, and derailleur pulling chain hard from the outside. You lose all your margins of error - slightly loose dropout, slight misadjustment, slight spoke looseness, etc. The small-middle equivalent offers much less risk. If you back pedal in the small-middle probably nothing will happen. In the big-big you'll realistically derail your chain.

I understand that big-big isn't something to use on a regular basis. I once tore up my DA rear dérailleur when my bike fell over to the right and I didn't check the hanger, then shifted the RD into the spokes. I see more chance for problems shifting to the small ring under pressure in a race than going to big-big...but then I'm pretty careful to check that hangar since that disaster. ;) On a long hill, I'll shift to the small ring early on.

44 to 40 is a huge change. I'm also planning a swap down to skinnier bars, though it's a bit of a long story. Currently have 40 cm 3T Rotundo on the road bike. Had 40 cm 3T Ergonova on the cross bike. It turns out that the Ergonova runs about 2.5 cm narrower at the hoods than the nominal width because the drops are flared out. It's really too narrow on a cross bike. I was able to get a steal on some 42s for the cross bike, so that freed up the Ergonovas, which were originally on the road bike and which I'm planning to put back on, albeit with some more drop and a longer stem. I originally went to the Rotundos because the Ergonovas were too short and high. I will potentially lose some sprint leverage, but shovelhd and TKP can tell you that I'm pretty skinny and I'm not putting a lot of power into my sprint through the handlebars. If I don't like the change, the Rotundos will still be in my parts bin.

But it gets better. I tried the 42s on the cross bike, and they feel ridiculously wide. Like driving a tractor. I am absolutely not a fan. So I picked up a 42 cm set of Deda bars (measured outside-to-outside, so equivalent to normal 40s) to swap in instead. That should be the butter zone.

What's crazy is when I first got into riding bikes and even early in my racing days, I bought into "wider is better" mantra of Rivendell et al, and the idea that narrower bars made it harder to breathe or something. I had 42 cm bars on my road bike, and 44 cm bars on a touring bike. They were just hideously uncomfortable. Huge mistake.

I tried a compact for a year. I could never get comfortable on it. So I went back to standard. 53/39 seems fine for most of the riding I do. The only time I really struggle is when the grade gets up to the mid- to high-teens. once I cross 15 percent, it's a struggle to keep moving. at 20 percent, I am inching up the hill.

Last year, I bought a long cage RD on and an 11-32 cassette for when I take occasional trips to the mountains. It gives me the same gearing as a 34-27. I also could steal the compact off my wife' bike if needed.

I'm concerned about it as wear = friction = power lost to heat and whatever is wearing the chain.

I mentioned chain tension. I prefer a 52TX14T over a 45TX12T (same gear) as I believe less energy is lost in the chain/frame system. A tighter chain has more friction (and more compression on the chain stay - another thing). There is more force on the link parts. A chain run at an angle adds force between the plates, also adding friction. The wear is a byproduct of the lost watts.

Smaller rings and smaller cogs bend each link more - very little, but more. Bigger rings bend more links less. I pretend lube is between all plates and rollers and there is no static friction to overcome from starting to bend at the link. The less tension on the chain, the more likely there is lube between rollers and links.

you risk equation doesn't factor in the risk of dropped chains when shifting to the little ring, which i notice way more than people shifting into the spokes.

Correct me if I'm wrong here; but if you are saying that friction = power lost, doesn't that mean that 45x12 is better than 52x14, because the chain makes contact with fewer teeth. Fewer contact points = less friction = less power losses?

It's not the contact points, it's how much the chain has to distort. At the extremes, a chain running in a circle would have minimal friction, and a chain running between two small cogs would have lots of friction. Given the same chainline, larger chainring/cog = less friction.

That's not how it works. It's been found repeatedly over the years that larger sprocket combinations yield slightly less resistance. It's not a lot - maybe a watt or two, best case - but there is a difference. From what I understand, it is because of the slightly reduced radius of the bend each link has to make as the chain wraps around the chainring in particular. This is also the reason that there have been some companies selling extra large aftermarket cages and jockey wheels for rear derailleurs, but the benefits are much more questionable for the part of the chain that is not under high tension.

The 45T has both a higher tension chain and less teeth to deal with it. The force on the 45T tooth is about 33% higher than on the 52T ring (did a quick Excel calc.). A 52T has a radius 15% bigger than 45T. The chain is under 15% less tension on the 52T. For easy math I assumed the chain contacts just 26T on the 52T ring and 22.5T on the 45T ring (it is slightly different based on rear derailleur pulley position and top of rear cog).

Experience shows smaller gears and rings wear faster.

But the friction I'm concerned about is in the chain. That between the rollers and pins and between the plates. I want less tension and a straight shot from rear cog to front ring whenever I can get it.

How much does this all matter? Hard to tell but chain optimization through different lubes seems to save 4-5W in testing. I guess its in that realm.

I mentioned the bars earlier and put a tape on them. Hood saddle 38cm c-c, plugs 44cm c-c. It flairs the brake levers out a bit. I moved the hoods about 1cm higher than this picture and they are flaired out even more, but that has not presented any problems. The flat tops make forearm resting much nicer. The round neck where the stem grabs is narrow, so putting a Garmin mount around the round part of the bar required a modified (plastic) mount clamp. For normal riding we use the provided stem Garmin mount and that also holds on the Di2 junction box.

http://bikeforums.net/attachment.php...hmentid=452392

Ah, ok. Thanks! That makes sense, I guess that's why Tony Martin does his TT on those huge huge chainrings.

Partly, but the other reason is to have a straighter chainline. Which is also slightly more efficient.

Some TT setups are pretty cool. One kid his kid with only 4-5 cogs in the middle. In the past I just ground my last two see post #1518. Current setup will allow me to put a 15 anywhere. In this case I will not use spacers as those don't add enough friction to keep cogs from stripping off at the hub under high torque. So we will look into running 14-16-17-15-18-19... 10 speeds with 11 speed spacing. Problem is we really want the 14T in the middle so were talking about going to a 56T front and grinding the teeth off a 13T and putting it in 1st position.

EDIT: Found this for a 14T middle http://www.ebay.com/itm/201345102045
So "Big Gear" ends up anyplace we want.

Some of this stuff seems waaaaay more involved than I'd ever want to be with the tech side of things, but then again, my goals are much more modest (like, remember to bring the deeper wheel to a TT).

What kind of improvements/marginal gains do you think are you getting from these tweaks? Are the gains theoretical or have you actually seen directly measurable improvements?

Good point. In fact I never shift into the small ring in my races to avoid even the chance of dropping a chain. Hills don't matter because I don't do ones big enough to require a small ring.

SRAM is marketing 1x11 for crit racing now. It's kind of not a bad idea.

I think any one thing may not be measurable. Some do nothing. All introduce risk and stuff that when not tested - may blow up. Some is calculated risk - like the silk tires I had made by Fraincois (FMB) with no sidewall coating and record tread to save 30g/wheel. I knew they wouldn't last after one rain - and it rained their first use. Anyways, it is fun and we both enjoy it.

Results of it all? Guesses. The optimized chain was tested reported 5W savings. Other stuff is obvious, other are guesses. TT helmet choice is again a guess.

My guess is 1 step on the podium - 2-3 sec per half hour. Had Peter Sagan had these tweaks in this year's ToC I think he would have held yellow.

Last year Daniel got 1st in the 15-16 TT at VOS (all the best were there) by 1.6 sec. I do not think he would have been on the top step.
Daniel rides a 13#and grams Di2 Venge in race dress.
He will typically beat a kid in the TT/climb that has similar or more power.
(Brandon has way more power and it doesn't work against him.)

As I said - it is fun, but I am seeing more of this stuff done at the International / World class level now.

Last couple decades I only built one ring on TT bikes for both wife and son. That is what I did on my road bike since about 1990. I had the frame brazed for just the cable stops for the rear.