mtnbke

Chain drive isn't best for everything. Chain drive absolutely can not handle really high revolutions. There is a reason that the timing belt on almost every car is, in fact, a belt. Only on a very few special applications is a timing chain used.

However, for a simple application belt and shaft drive just does not, and can not compare to chain drive in terms of driveline efficiency.

You're kidding right? One hundred and twenty-five years of mechanical engineering goes out the window because a company that markets 'carbon drive' hooked up an electric motor to a Powertap? C'mon...

Any intelligent test would control for power input with a measured and calibrated source of power being applied to the drivetrain.

You've got to be kidding. We have cellular phones, microwave ovens, wireless internet, television, hybrid gas/electric cars, the Segway, personal laser printers, hand held lasers used only as pointing devices, we can brake the sound barrier in an airplane, we can go to the depths of the ocean, but somehow you think in the past 125+ years people forgot to actually explore how to make the velocopede better?

The bicycle is the world's most common form of transportation. Racing bicycles is an absolute obsession in Japan and in Europe (and used to be in the US aka Major Taylor).

The reason that the simple chain wheel, chain, and derailleur is still in use is not because its the best method of having selectable gears (there are a thousand better ways), but rather because its the most efficient. When its human powered, every watt counts.

There are motorcycle manufacturers that make bikes with chain drive, shaft drive, and belt drive. Their research and design budgets are astronomical. Yet they have never, ever been able to improve on the efficiency of chain drive with belt drive or shaft drive. However, they still manufacture different drive bikes for different applications, though they would never claim that they are more efficient, or equally efficient.

djsincla

I can't say the same. I ridden various bikes and I have not seen the flex you talk about in current models of all the manufacturers. Without getting hooked up on the Santana brochure which I'm not a fan of myself, the geometry and tubing used across the industry is pretty much the same. My biggest beef with the Treks and Cannondales is they will save money where they can - Bottom Brackets, wheelsets, and cranks.

Belt drive/Shaft Drive - A "solution" looking for a problem.

mtnbke

I'm a cyclist, not a mechanical engineer. I know this not because I've ridden one and my 'seat of the pants' meter told me, but because belt drive is less efficient than chain drive as a documented fact. I trust the mechanical engineers know what they are talking about...

Engineers have tested the efficiency of a bicycle drive train to be roughly 98.6% and that was before 10 and 11 speed narrower chains...

I can't believe that in a discussion of driveline efficiency that these three reason are actually being espoused. First of all any variance in efficiency is going to be along the lines of 1-3%. I'd be astounded if the Gates system was even 5% less efficient. In a bicycle drive study conducted by Johns Hopkins engineering students the least efficient drive mechanism was only 18% less efficient than a chain drive bicycle.

Regarding your 'spin it backwards' test. Its patently absurd that the belt drive would spin as backwards 'at least as freely'. I'm sure to the naked eye it seems like it is spinning 'as much.' However, science isn't conducted using anecdotal evidence. The limiting factor in spinning the pedals anyway is going to be friction from the bottom brackets and the pawls in the rear hub. Trying to detect the subtle difference between a 98.6% efficient drivetrain and say a 95% efficient drivetrain when while the bearings and pawls dominate (in terms of momentum loss) as compared to drivetrain friction is just absurd. Trying to detect a difference anecdotally is intellectually dishonest.

Regarding your fastest time. I'm sure you did reach your fastest time. Congratulations. However, I don't think you intended to convey that you did so in a controlled environment where the only variable was the driveline.

Regarding keeping up with Cat 2 roadies. Of course you can keep up with roadies. That's the whole argument, that tandems are so efficient (at least non-wriggly ones) that it makes no sense to forego efficiency to save 10oz of weight from the bike. You can put any two reasonably in/out of shape folks on a tandem and any competitive cyclist would have trouble keeping pace. The tandem is that much more efficient. Why throw efficiency out the window, its what makes tandeming fun.

The power loss from using a chain drive is somewhere around 1.2% on a single. I have no idea what it is on a tandem, but assuredly it is less than using belt drive. Not a considerable margin. We're probably talking percentage points, I'd be astounded if it was even 5%, but that's a HUGE amount of wattage to have to make up over a ride. Any loss of efficiency is just an unacceptable trade off, for pounds, let alone a couple of ounces.

Let me know when you spot the first belt drive bike on a Grand Tour...

Hey, Santana now claims that they didn't used to spec Softride beams but on a couple of custom bikes. Saying it doesn't make it true.

If Santana says their steel tandems are stiff and efficient and Gates says their belt drive is 'as efficient' as a timing chain, well then what could possibly be more efficient than a steel Santana with a Gates belt drive?

Did I throw 10% out there? If I did I'm an idiot.

No drive system ever designed has ever rivaled that of the chain driven derailleur changing bicycle, in terms of efficiency. That hasn't changed with the Gates belt drive. I understand that Gates wants to sell more belts, but it ain't more efficient. Its probably not a lot less efficient, but why leave the drag brake on, so to speak? For a measly 10oz? No thanks.

C'mon. We're riding hand laid custom carbon fiber bikes decked out with titanium components. Others are riding Scandium and Aluminum alloy frames on composite wheels. Yet the drive train is almost completely unchanged in a hundred and twenty five years. There is a reason, its the most efficient.

If sales literature were accurate I'd own Santanas. Of course Gates isn't going to state that their drive system is less efficient. They want you to buy it.

djsincla

Do you have a front eccentric bottom bracket or a chain tensioner? If you have an eccentric, your timing chain is way too loose and you need to adjust it.

mtnbke

Now that is a brilliant perspective.

I've never developed a smooth cadence with any stoker I've ever had.

There are many reasons why belt drive would be better on some bikes. They eliminate chain tatoo as they don't need to be lubed. I'm willing to bet that commuter bikes are mostly belt drive ten years from now. Grease is never in style. Belt drive would be preferable for a roun' the world unsupported tour. Who wants to try and have a chain airmailed to Timbuktu communicating via short wave radio (same argument for Rohloff, so maybe belt drive to Rohloff?). You wouldn't have to replace rings or chains, as belts last forever and don't wear out the other driveline components. Who want to replace Ti cassettes that are tired because of a worn steel chain?

I'm not saying that belt drive doesn't have a place. I was just saying that a tandem is fun because it is so efficient, and I can' see why someone would want to reduce that fun factor (by any margin).

mtnbke

I haven't always been 375lbs. I used to play basketball competitively, at around 205lbs.

I wanted a new mountain bike around '92 and was looking at a Bontrager Race frame (before the Trek buyout). It was a legendary steel frame, but it was just wriggly at the bottom bracket. I couldn't believe that the frame that was reknown for being strong was so flexy.

Not all cyclists can put out the same wattage. Not everyone uses the same crank length, or rides at the same cadence. However, 26" wheel is much stronger than a 700c wheel. By far. I've never ridden a teeny tiny little bike, nor can I get on a small framed tandem to compare. However, I can definitely tell you that most tandem frames aren't that stiff.

I can say that even with clown cranks, and even when I was a stick trying to bulk up (to hold my own in the paint) from 205 that a flexy bike was noticeable. I actually can't put out nearly the wattage now that I've gained weight. Bottom bracket deflection is actually less noticeable.

However, a wriggly frame is a wriggley frame. Would anyone still really argue that the 'best' bikes are steel? That the 'best' tandems are steel? C'mon. They are cheap to manufacture, nothing more.

I don't even think Cannondale makes the 'best' tandem frames. I would think that a Titanium frame could be constructed to be stiffer and lighter, and more forgiving than aluminum. I bet Calfee could make an amazingly stiff bike out of carbon if the couple wanted that.

Its not so much that Cannondale makes great tandems, but that Santana doesn't. At least not at their low end. I just find it disingenuous that they claim to be exactly what they are not, and that the shrewd tandem customer can find a much more efficient bike (even perhaps with the Gates belt drive ) than what is available from Santana. If Santana focused on their tandems being as stiff as they do on claiming they are stiff, this conversation would be moot.

Read the Santana sales magazine sometime, then actually go and test ride a Santana steel tandem. The claim versus the ride is just pure audacity...

An oversized aluminum frame isn't going to flex, period. There will be no bottom bracket deflection. It will be very efficient, which is what makes tandeming fun. If it did flex, the aluminum would fatigue, and a catastrophic failure would occur (think Airplanes).

mtnbke

Cannondale's tubesets for the bikes I have ('92 & '96 tandems, '89 single road bike, '92 mountain bike) have to be some of the 'least' engineered around. I swear these tubes aren't even butted...

My '92 came with some funky Grip Shift roadie shifters. Those were removed from the bike before our second test ride. It also came with an 8 speed freewheel (not cassette). However, to be honest, I think Santana's component specs are worse than Cannondales. I like what Co-motion specs.

TandemGeek

Last time I checked, reading each posting is optional not compulsory.

You'll need to be more specific as I don't have a clue here based on what's in this thread. embankmentlb took umbrage when mtnbke described his experience with Santana and apparently didn't have an appreciation for mtnbke's needs for a very robust frame or appreciate that Cannondale has always offered a very robust, well-designed and well-made tandem frame. The tubesets aren't sexy, but neither are the tubes on our Ventana: straight gauge stuff to be sure, but well-suited for the intended design objective.

Bottom Line: I really don't give a rats-patootie what anyone rides so long as they like it and it meets their needs and pretty much said as much in my postings to this thread and most others that get into brand debates.

Not hardly. Although we sold our Santana back in '98, Bill and I have remained in contact ever since. We don't correspond as often as we used to given that he's apparently sworn off posting to discussion forums -- the latter was often the catalyst for an off-list phone call or Email exchange. Phone calls tend to be quite long, but are always informative.

In regard to this thread, the one thing I haven't spoken to Bill about and probably should is his take on the belt drive. I suspect his position would be "if I thought it was a good idea Santana would have already offered a belt drive", but I'll have to ask or see if there's some type of a mention in the next Santana catalog that conveys their view on the belt drive.

merlinextraligh

Apples/Oranges. Simply because belts have traditionally been less efficient in other applications does not mean this belt in this application is less efficient.

What is absurd it that you would categorically reject the possibility that the Gates belt drive has no meaningful loss of efficiency when 1) you have no experience with it, and 2) the available data points specific to this sytem, not belts in general, would indicate a slight improvement in efficiency.



You're defeating your own argument here. It's a known fact that BB friction takes up a tiny amount of the power necessary to propel a bike. Thus if any loss of efficiency from the belt is insiginficant compared to BB friction, then it's also insiginificant in terms of the effort necessary to propel the bike. And if the Belt is less efficient, but the decrease in efficiency is not signifcant in propelling the bike, then the 10 ounce weight savings is going to make you faster climbing with no meaningful loss of speed on the flats.

And by the way you can question someone's reasoning or methodology without calling them a lier

Obviously not a controlled test. However, if we lost 10% of the captain's power by putting the belt on, it seems highly improbable that we could be faster.

To sprint from 15 mph into the low 30's at our size takes 5 sec power above 2000 watts . Probably 2/3rds of that comes from me, the captain. Thus if we lost 10% of the captain's power, my 5 second power would have to rise 130 watts just to stay at the same speed. I know for a fact my 5 second power did not jump 130 watts the week I put the belt drive on.

My point here is that for us to ride with the truely fast guys, we're at our limit. If the belt drive made us measurably slower, we'd have been beyond that limit and dropped.


As soon as they have the Tour de France Grand Velo I will. However, you did miss my point that the Gates system has been used successfully in the Co-Motion Classic Tandem Stage Race.


Bottom line, there are a number of reasons to believe the system has no meaningful loss of efficiency.

Other than extrapolations from other belt drive applications, there is little reason to believe that the sytem is less efficient.

I'm open to the possibilty that time may show the system to be sufficiently less efficient that the efficiency loss trumps the weight advantage. I've placed my bet however that that will not be the case.

You however, are totally closed to the possibilty that the system works well, with the benefit of zero experience, or direct data.

joe@vwvortex

How high are you talking? Audi's V8 in their RS4 which has an over 8,000 rpm redline uses chains - in fact both Audi V8 and VW's VR6 both use chains exclusively. In fact their new 2.0T went back to chain drive for timing..... pretty rare . Belts are simply cheaper and easier maintain over time. While there are other advantages of using a belt for timing and their might be some relationship to chains not being able to handle higher rpms - for a street car - the fact that a engine has a belt of chains isn't based upon it's ability to handle higher rpms.

zzzwillzzz

most (if not all) race replica/sport bikes motorcycles have chain drive for timing. the 600 cc bikes have a redline around 16,000 rpm. so speed isn't the reason to use a belt

since the belt comes in only one size to fit the 28.5" bb to bb distance, would a little larger 'chainrings' work on bike with a 27.5" bb to bb distance or is that distance not standard enough between other manufacturers? or would the larger rings have interference problems with the chainstay?

chichi

Wow, This is the most hotly discussed item since "Tandem Parking - How do YOU do it?".

TandemGeek

There's very little standardization. The following is just a partial listing and it could be a bit dated as I forget when I first put it together noting that I threw in some of the new Cannondale specs so it really is cats and dogs:

>30" = Erickson / Bushnell / Rodriguez / Bohemian
29.38" = Meridian (last year of production, R.I.P.)
29.1" = '99 - '08 and '09 Cannondales in larger size(s)
28.6" = '09 M/L/X-S Cannondales
28.5" - 28.8" = Trek
28.5" = Co-Motion & Meridian (R.I.P.)
28.4" = daVinci & Calfee
28.1" = X/M Cannondale & XL/L Seven
28.0" = Current Santana
27.75" - 28.25" = Bilenky
27.75" = Older Santana
27.5" = Burley (R.I.P.)
27.1" = M/S, L/S, X/S Cannondale & KHS
26.5" - 27.75" = */S-*/M Seven"

In theory, I suspect you could cover several frames with a few different sized pulleys if you play in the margins with the eccentric adjustment. Someone would actually need to measure each frame with the eccentric all the forward and backward to figure out what 'range' could be covered by a small family of pulleys with the standard length belt. Pulley size could definitely create a clearance issue on on some tandems due to the design of their stays, etc...

oldacura

As CGinOhio said, for a belt manufacturer to make belt molds in small incremental sizes at this size would be very expensive. If one tandem were to need a 450 tooth belt and another needed 440 teeth and a third needed 460 teeth, the belt manufacturer would have to make 3 large (expensive) molds. Similarly, if one bike maker wanted to use 100 tooth pulleys and another settled on 90 teeth, the pulley maker would have to make several different (expensive) molds for a very small market.

The reinforcing cords in timing belts are commonly polyester, fiberglass and Kevlar. Kevlar has the least stretch (but is the most expensive). I think the reduction in efficiency in belt transmission is due to the non-elastic stretching of the belt. To the extent that the stretching is elastic, the energy put into the belt is returned to the drive train. However, belts have a natural inelastic stretch that absorbs energy (like a shock absorber) that is lost. Chains have almost no inelastic stretch. However, part of the quietness of the belt is the inelastic stretch.

WebsterBikeMan

Interesting. Assuming you have both pulleys the same diameter, and that each end of the belt wraps exactly half way around its pulley, increasing a pulley's radius by an inch gives you an extra 2 pi inches in total belt length. So to go from the current Gates standard all the way down to the 26.5 only requires 1/pi inches increase in pulley radius; a little less than 1/3 of an inch; or 8 mm. Might or might not create a clearance issue.

Now that math ignores that pulleys don't come in continuous increments of radius, only in unit-tooth circumferences, which I believe to be 8 mm. That corresponds to an increment of 1.27 mm in radius. So 6 such increments takes you to 7.6, close enough to do it with an eccentric. Going the other way you might need another 6. So it would mean 12 additional molds and sizes to inventory.

CGinOhio

+1
For my part I place greater weight on more recent data obtained from testing modern belt constructions. I would like to see more data, but I have no reason to doubt the claims.

+1
Composite materials development currently is very rich in innovation. Composites, laminates, nanofibers, etc are being combined with novel performance balance. Who among us would have thought just a few years ago that a modern wide-body jet would be produced with heavy reliance on non metallic structural components.

+1
This thread reminds me of the very opinionated discussion of pros /cons of rim vs disk brakes, minus actual user experience (Merlin's nice summary the exception). It appears the belt technology may bring a nice performance balance option vs chain. Not for everyone, particularly at the current price, but I think its quite interesting.

jccaclimber

Every Chevy v8 block I've seen does rather well with timing chains. I've seen tons of those go to 100,000 miles without replacing or maintaining the timing chain without problems. I'm betting the v6 blocks have a chain as well. Belts get used because they are cheaper, and perhaps quieter. I've also seen 700hp race engines which turn 10,000 rpms, the timing chain is the last thing they are worried about breaking.

dvs cycles

Quick question merlinextralight.
How big are the pulleys compared to a comparable diameter timing ring?
I currently run a 40 and having a bigger diameter than that would effect my method for carrying it in my truck.
Thanks for the review and pics.

Chris_W

We have the Gates carbon belt drive. I just put a spare 42 tooth chainring up against one of the belt's ring, the chainring is maybe 5mm smaller diameter than the flange on the outside of the belt-ring, but if the chaingring were to have a chain wrapped around it then I'm pretty sure it would be the same diameter as the belt-ring with a belt wrapped around it (because the belt doesn't really protrude above the flange).

mtnbke

There is every reason to believe this. Chain drive is chain drive, Belt drive is belt drive, shaft drive is shaft drive. The application doesn't make it apples to oranges. Belt drive doesn't become magically more efficient than chain drive on bicycles, but somehow less efficient on motorcycles.

The efficiency of these various drive systems has been studied exhaustively by mechanical engineers for the better part of a hundred and twenty five years.

The chain drive system on a bicycle isn't just more efficient than belt drive, its the most efficient energy transfer device known.

Again, the variances here in terms of efficiency aren't significant proportionately. If the Gates belt drive system is only 3% less efficient it doesn't seem like much. However, we're talking human powered machines here with very real constraints on wattage and endurance.

I'm not interested in going back in forth. This issue has been studied by mechanical engineers ad nauseum, but somehow, all that research is for naught because of some anecdotal claims? C'mon...

There are very real reasons why the belt drive would be a good idea. However, in my opinion tandeming is all about the efficiency and speed of the ride. No cyclist would prefer worn bearings, a dirty chain, or aggresively treaded tires (on pavement), all of which impede peak efficiency.

10oz is very easy to save on a tandem. The difference between XTR or XT V-Brakes and a good set of lightweight V brakes is four ounces right there...I wouldn't trade efficiency for 10oz.

However, I don't clean my chain every ride, nor do I lube it every ride. Perhaps belt drive becomes more efficient when compared to drivetrains that are have worn cogs, rings, and chains, or that haven't been lubed, or are dirty. I don't know.

What I do know is that mechanical engineers know a heck of a lot more about this issue than you and I. I'll take their word on it, which has been definitive and consistent since the 19th century.

Nothing has every rivaled chain drive on a bicycle for mechanical efficiency of pedaling power. Not a hundred and twenty five years ago, not fifty years ago, and not now that Gates wants to sell belts to tandems.

mtnbke

Chains have an upward limit afterwhich they simply fly apart.

There are real advantages and disadvantages for using either a timing belt or a chain involving what happens at failure, cost of replacement, durability etc. Most timing belts are relatively narrow, most timing chains are a bit wider.

mtnbke

Here is a list of the carbon drive sprockets available:

https://www.carbondrivesystems.com/cd_specs.html

By using a different sprockets I would think the timing system would work on more than just what it is being marketed to. Although, it would require the purchase of additional sprockets on top of the timing kit (with sprockets that wouldn't get used, or worn from use). So it will work on 130BCD and 104BCD timing cranks currently, just with an additional outlay to buy the 104BCD sprockets.

The real advantage here, at least to me, is the notion of getting rid of the timing chain altogether. The only reason the timing chain exists is to accomodate shifting on the right side. With the use of a Rohloff, or some other similar internal hub, a single drive system could be set up, and using only normal cranks.

The belt would go from the captain's sprocket to the stoker's sprocket to the rear internally geared hub.

I'm willing to bet that a single drive belt system is significantly more efficient than the dual chain setup we currently use, perhaps even counting the efficiency losses of the geared hub...

I do think belt drive will transform the bike market toward internal hub shifting. The benefit for tandems will be stronger wheels (less dish).

mtnbke

Interestingly, Gates doesn't seem to be making the 'more efficient' claim that Co-Motion is (for the Gates system). If you poke around the Gates FAQ, 'more efficient' isn't ever claimed as a benefit:

https://www.carbondrivesystems.com/support.php

Its a miracle device regarding everything but efficiency...

Regarding efficiency:

I think anyone can figure out for themselves that a chain timing drive and the belt drive are probably not both EXACTLY equally efficient. I'm curious what Gates considers to be 'close enough'...

Cycle Drive gives some insight:

https://cycledrive.com/faq.html

Check out this synopsis of a materials science study at Johns Hopkins:

https://www.jhu.edu/news_info/news/ho...ug99/bike.html

They are throwing 98.6% efficiency out there for the bicycle drivetrain.

This website discusses human pedaling efficiency:

https://auto.howstuffworks.com/question527.htm

Let's use their numbers. If Carbon Drives numbers bear out for Gates and if the variance is ONLY 0.6%, that is the belt drive is a little less than 1% less efficient, then what that means is that when the cyclists are hammering at 25mph they are not using 100 calories per mile, but rather 100.6 per. That's not really that significant. I wouldn't use the belt system in a race, but it would be darn near impossible to detect the six-tenths of a percentage variance without using a controlled environment and precise testing equipment.

So maybe that ten ounces is worth it after all...if the Carbon Drive claims bear out.

However, I'm not giving up even that much efficiency. I'm certainly not payin' $500+ to give up ANY efficiency. For $500 I can get a lot more than 280 grams off the bike.

WebsterBikeMan

But anyone with a reasonable background in statistics can understand the concept of statistically insignificant difference. In other words if you make 100 measurements of each and the variance in the measurements indicates that the means can't reliably be teased apart, they are "equal". And then there's the question of practical significance to the difference, which, if it is in the single digit tenths of a percent (and I'm not sure the experimental evidence has been adequate to say that yet) is likely not there.

But I agree, the $500+ figure is for early adopters. You have to want it for many reasons. Perhaps chain tattoo is REALLY important (your stoker threatened to stop riding if you couldn't do something about it?), or you just can't resist the latest cool thing, or more likely it is a number of them combined.

Single belt is an interesting idea. I haven't done the math to see whether the added weight of a Rolhoff would make up for the reduced weight of the two or three chainrings and the difference betweent two chains and an extra-long belt.

masiman

Except direct drive (high wheelers).