Note: Please refer to the comments for on-going dialog regarding belt-drive. I’ve revised my original blog entry twice based on the on-going dialog with Dave Walker from Paketa that has been informative but also raised more questions in my mind than it has as of yet solved. However, I am starting to turn the corner on understanding why the single side drive drive may be a better adaptation of the Gates Carbon Drive vs. the more commonly seen crossover crankset design. So, somewhat with hat-in-hand, I’ve softened this entry and will be posting a subsequent one in the near future to relay what I learn.
I usually get pretty excited when a US-based tandem builder introduces something new and interesting. But nothing diminishes that experience to me more than aggressive marketing and hyperbole.
A certain tandem maker has used marketing spin for decades that never fails to raise my blood pressure; however, it seems to work well out in the general bicycle consumer market place and also resonates and reassures their past buyers they have selected a superior product from a superior company. On the flip-side of the coin, the tandem enthusiast community’s reaction to this marketing has been labeled ‘Bashing’. In fact, coming off the mid-term elections it dawned on me that this entire marketing approach and enthusiast backlash seems to mirror the two-party political election campaigns, where two polarized groups sling mud back and forth over the heads of the undecided “consumers” who are trying to make an informed decision based on dubious claims and slick packaging. Of course, if you eliminate the marketing from the product, the product really is excellent. But, so are the products offered by their competitors. It could be easily argued that the differences in some cases could be measured in millimeters not meters and are often times simply differences in highly subjective preference.
Anyway, this blog entry really isn’t about that other company and it’s marketing. Instead, what caught my attention today was the Paketa Bicycles web page content describing their new V2r (where the r = right side drive). Don’t get me wrong, I’m genuinely excited about the new variation Paketa has offered up on its legacy magnesium tandem frame, the V2. However, there are a few things in the copy that just don’t resonate with me and some that are challenging my understanding of the pros and cons of the Gates Carbon Drive sync belt for tandems. So, let me start with what I liked before getting into what has me scratching my head.
First off, you can find the photos and Paketa’s description of the new variation at their V2r web page by clicking HERE. In short, Paketa is now offering a single-side drive version of their V2 tandem where the sync drive that connects the tandem pilot’s cranks to the stoker’s cranks is located on the right (aka, drive) side of the tandem. This is a variation that has been in use on tandems going back to the late 1800’s, along with many other variations of tandem power transfer systems. Since then, right side drive tandems have always been around, but in recent years the push to find ways of removing weight from tandems as well as more practical reasons — such as taking advantage of the ‘free’ chain ring that came available when a Rohloff internally geared hub was installed — have made right side drive tandems more common. In fact, we’ve seen a few right-side drive off-road tandems being prepped for customer delivery by our good friend Alex Nutt at MTBTandems.com.
Paketa V2R Right Side Drive Tandem (Photo by Paketa)
Paketa’s right side driver is a bit different from others in that they have revised the design of their rear bottom bracket’s chain stay yoke to allow the installation of a 69t Gates Carbon Drive sprocket on the inboard position typically occupied by the smallest chain ring (aka, alpine or granny ring) of a triple crankset, but in this case bolted to a double crankset with the same net effect. Of course, Paketa is not the first firm to play around with their chain stays (or rear bottom bracket / yoke) to accommodate things like the 69t Gates timing sprocket. In fact, one of the newer Calfees I’ve seen that was built for the Gates system appeared to have a a new kink in the left-hand chain stay that accomplished for a crossover crankset what Paketa is now offering in a same-side drive, a tighter fit to against the bottom bracket area. I’m not sure if there’s a patent pending on it or not, anymore than there were patents taken out for other chain stay shapes that provided more heel clearance, increased tire clearance, or wider bottom brackets to allow for larger, more robust stays.
Getting back to Paketa’s right side drive and that inboard Gates sprocket position, therein lies one of the cons for a right side drive: right side drives need to use one of the main drive crankset’s chainring positions. There are several other pro and con arguments typically associated with right side vs crossover cranks, some are more valid than others. Clearly, eliminating an extra crank spider and chain ring will reduce weight and opens up the ability to use a wider range of standard cranksets on a tandem… that is, so long as they’re rated for the combined weight / power of a tandem team. Again, there are some others that I’ll analyze in a moment as I go through Paketa’s Web material with you.
The potential gearing range limitation isn’t a big deal for a tandem that will be used for riding or racing on all but mountainous circuits where the really short climbing gears are never used. Instead, a team that doesn’t need serious climbing gears can mate what is essentially now a standard double chain ring crank configuration to a nice close-ratio cassette and enjoy single-bike like shifting or even use Shimano’s Di2 components. This could be a really slick set-up for a dedicated race bike or strong recreational teams who live, ride and race in places that don’t have significant climbs. Now, bear in mind that anyone with a tandem could easily replicate this right side drive configuration today using sync chains and timing rings on a triple rear and standard double front crankset with the new, wider-range 11X36t cassettes from Shimano and SRAM. So, the only real difference here is figuring out how to get that wider belt sprocket into the narrow space between the chain ring and cranks without using some bizarre, wide, off-set bottom bracket axle and moving the front derailleur outboard with a special clamp. Seems like an awful lot of work to save an extra pound or two on a racing tandem that isn’t ideally geared for steep terrain… where a rider would actually benefit from frame weight reductions.
So, here are the things in the on-line copy that didn’t resonate with me and challenged my understanding of the pros and cons of the Gates Carbon Drive sync belt for tandems, starting with the ‘better’ gearing system, which reads:
Better gearing system
As for gearing range and shifting, the V2r again excels. The combination of 2X10 road and MTB gearing provides the same wide range as a conventional triple-chain-ring road setup. With a 39/53 double sprocket in front and 11X36 cassette in back (as shown in the photos), the gear range is identical to a road 30/39/53 triple crank with an 11X28 cassette—certainly enough for most any tandem team interested in racing or fast sport riding, for sure. And, with a double sprocket in front, the shifting precision is as good as any single bike, and any road shifter compatible with a matching MTB rear derailleur and cassette will work (SRAM™ Red™ Double Tap™ shifters, Red™ front deraiiluer, SRAM™ XX™ cassette and XX™ rear derailleur shown). Say goodbye to triple front shifters and substandard front shifting!
As I mentioned earlier, for high-performance teams who ride tight ratio cassettes like an 11-23, the right side drive V2r makes sense and would be an attractive package. In fact, I just ran the numbers on the gear inches and was pretty surprised at how closely they did match up to the 11x28t triple… OK, I was REALLY surprised. So surprised that I’ve actually re-visited this blog entry and made some changes and I must thank Terry Malouf who is one of the owners of the first Paketa V2r for responding and making me do a re-read of what I wrote. However, that said, I still believe that for more average fast recreational / sport teams who must deal with any steep stuff, those folks will really need to be really honest with themselves regarding their current gearing: could they really live with just a 30x28t. If so, this might be a good option. However, if they ever needed something shorter than a 28″ gear… noting that a 30x32t granny yields a 24.9″ gear and a 30x34t granny yields a 23.3″ gear, they could find themselves walking a few hills.
Let’s take a closer look at the Paketa 2×10 scenario that saves 1lb: “With a 39/53 double sprocket in front and 11X36 cassette in back, the gear range is identical to a road 30/39/53 triple crank with an 11X28 cassette“. Here’s how those two different cassettes stack-up:
–
This is what you give up…11-28: 11-12-13-14-15-17-19-21-24-28 And this is what you get…11-36: 11-13-15-17-19-21-24-28-32-36Again, potential buyers really need to be sure they can deal with whatever length and grade climbs they expect to encounter with nothing larger than what they could have handled with a 53/39/30t triple and an 11x28t cassette. Most tandems these days tend to come with an 11x34t cassette, which is a very different animal than an 11x28t.
For reference, we’re hardly elite racers but we do tend to fall into the demographic that does the suggested fast sport riding and rides higher-end performance tandems. Here’s our “normal” gearing for hilly but not the really steep stuff… a very nice progression:
12-27: 12-13-14-15-16-17-19-21-24-27
Pending release of the 11x32t Shimano CS-M771 cassette, the first cassette listed below (an 11x32t 9 speed) is our current ‘alpine’ gearing. It’s less than ideal with that awkward 18 to 21t jump that we have to deal with a lot on moderately long or steep climbs. Thankfully, the 11X32 CS-M771 will beak that up a bit. Again, here’s how all these different cassettes stack-up:
11-32: 11-12-14-16-18- 21 – 24 – 28-32 12-27: 12-13-14-15-16-17-19-21-24-27 11-28: 11-12-13-14-15-17-19-21-24-28 11-32: 11-12-14-16-18-20-22-25-28-32 11-36: 11-13-15-17-19-21-24-28-32-36Compact drives and doubles have not been the hot ticket for tandems, heretofore. Now, as is always the way with bicycles, with Shimano & SRAM’s introduction of 2×10 systems for off-road bikes they’ll likely become more common, just as compact drive did for the road and 29’er bikes for off-road. So, perhaps 2×10 drives on road tandems are inevitable. However, there are limits to just how flexible a 2×10 drive train can be, and therein lies the problem with the marketing spin: it’s trying to over-reach and make a credible product offering appealing to consumers under the allure of ‘lighter weight and better shifting‘ who may not be well-served by that product over the long-haul. Obviously, the V2r can always be fitted with a crossover crankset that would allow for the use of a triple, so it doesn’t lock in an owner to the right-side drive / double chain ring configuration.
What sticks in the back of my mind is, it’s been my observation that the stronger tandem teams of median income will have only one tandem they use for everything vs. a dedicated race machine, and they still keep that granny gear on the bike with their 11x23t or 11x25t cassettes for use as a ‘bail-out’ gear when they find themselves on those longish double-digit grades or run into a wall while touring or attending a rally away from their normal terrain. Yes, we’ve seen one or two teams with older Santana Team tandems from the 90’s that had DuraAce doubles or somewhat newer tandems with unused granny gears hammering up some 18-20% grades, but they’re the exception and not the rule. However, I’m not sure they’d have been willing to give up their closer ratio cassettes for what the 11×36 offers. But, to be fair, that’s just a guess since none of these teams were all that concerned about or giving up much in the way of their dominating performance as they rode their 40lb tandems up those walls vs. mere mortals with deeper pockets who struggled up using the granny rings and the largest sprocket they had on the newer 30 lb lightweights and exotics.
It’s also worth noting that another tandem builder has been dangling a compact 11x36t drive train set-up with a Di2 option out there for 2-years or so. It would be interesting to know how many of those have been sold and what folks think of them, as well as how they’re being used since that is fairly important too. For example, I ride a compact drive on my single (50/36) and it’s great, but… not sure I’d want to use it on the tandem where we live and ride. You give up a lot of useful gear range that, while you don’t use it a lot on the tandem, you appreciate that it’s there for those infrequent occasions when you want to spin-out that 53/11 gear (and it happens a lot sooner in a 50/11) or you hit a wall that you just rather not grind-out standing on the pedals at 50 rpm.
Finally, I still don’t get the spin on “Say goodbye to triple front shifters and substandard front shifting!”. It must be a Shimano STI thing, as I just don’t hear of folks who use bar-ends or Campy Ergo having front shifting problems, except where their technique is all screwed-up, i.e., trying to shift under load and way too late. Well, and then there’s always the problem where so few mechanics know how to work on tandems, which also leads to problems. But I digress.
Looking back at the rest of the copy on the V2r web page, there are now a few other things that caught my attention once the ‘Better gearing system‘ comments caused me to crank up the gain on my ‘does it pass the common sense test’. Here are the items that caught my attention:
How do you quantify this? “Why is the V2r the absolute best racing tandem on the market?”
Let’s take a look at those inherent Crossover drive disadvantages:
- Tandem-specific cranks are required with a spider or other mount for the transfer sprockets (Actually, that’s not quite true. A crossover crankset can be easily constructed using three right-side standard cranks and a single left-hand cranks. R&E cycles in Seattle has been outfitting their Rodriguez tandems with Campy cranks in this manner for years. Yes, it requires some minor re-work on the cranks, but it’s a tried-and-true practice.)
- Three spider-mount crank arms are required, which adds to the weight of the system (Unless you use daVinci or Middleburn cranks, which don’t use a spiders, noting that daVinci’s cranks have always been some of the lightest cranks on the market… lighter than many of the carbon offerings until perhaps recently. But, yes… in general a crossover crankset will be heavier than a right side drive)
- The rear bottom bracket experiences high loading due to the force of three riders’ legs (both of the captain’s plus the left stoker’s leg) being transmitted through the rear crank axle. (OK, I’ll give them that. However, before the bicycle industry started messing around with what were promised to be lighter, stronger bottom brackets and weren’t, just how many tandems running normal width square taper bottom brackets were experiencing abnormal failure rates? I’ve see Shimano UN72’s deliver over 20k miles of hard tandem use without any maintenance )
- Both torque and bending forces are significantly higher—as much as three times as high—as on a comparable single bike, resulting in either reduced component lifetime or higher weight, or both. (See prior comment )
- Modern belt transfer drives on a tendem’s left side, place much higher loads on the bearings due to the higher tension required compared to a traditional transfer chain. (This is where my understanding understanding of the pros and cons of the Gates Carbon Drive sync belt for tandems is seriously challenged. Before the ensuring dialog with Dave Walker was created by my original version of this blog entry I believed and wrote, “… this is probably true for when a tandem is sitting still and no loads are being applied to the pedals since a belt needs more static tension than a chain to ensure it does not slip once pedal loads are applied. However, those loads are miniscule compared to the loads applied by the pilot during use. Seriously, belts are hand tightened using a pin spanner. If you assume that similar size rings and sprockets are used for the chain or belt, there should be no significant difference in the loads being transferred to the bottom brackets by the tandem pilot’s power output. As of Nov 11th, I’m still trying to come up the learning curve on the belt technology vs. what I thought I’ve observed. The latter is call covered in the comments appended to this blog entry.)
Here’s another one of those great, subjective dubious marketing statements: “The Paketa V2r tandem (patent pending) eliminates all of these problems” What problems? If previous ‘innovations’ in bottom bracket designs weren’t so poorly thought out before being rolled-out to the general public, most tandem teams would still be using their original bottom brackets after 20k miles of use. Reducing weight is novel, but as over-rated as reducing rolling mass when it comes to generating performance improvements for all but the most elite athletes. Therefore, I’m hardly inclined to believe that shaving yet an additional pound off a tandem with 320 lbs of riders, never mind even 10 lbs of bike weight, is anything more than marketing hype feeding consumer vanity and hubris. Well, yeah, it’s also fun to see how light you can get a bike, but that’s simply a distraction from the reality that our physical performance is what truly limits the speed and power envelope of our tandems. We’ve already discussed that you can already use just about any cranks on a tandem, so long as the spider is rated for the total loads that a tandem with several hundred pounds of riders standing and pedaling up a 20% grade in a high-purchase gear like a 39×36. And, as far as boom tube deflection under loads, if a modern tandem’s frame isn’t robust enough to prevent unpredictable handling with a crossover crankset, why would it be immune to right-side wind-up that now runs from the rear axle to the front bottom bracket? This is actually a red herring for most tandem teams if they’re riding a tandem that was designed to deal with their weight and power. IMHO, the real place to focus attention with tandems — especially open frame / lateral-less tube models — is making sure the top tube and overall structure provides sufficient stiffness to resist torsion flex which truly can create a tandem that handles poorly under heavy efforts or rider / luggage loads.
Let’s see what else is on the 2nd page. Yes, Paketa does a fantastic job with its Magnesium frames, just based on feedback from most of the owners I’ve spoken with or who have documented their impressions. And, yes, I’m sure the new yoke is also exceptional. But, this is a stretch: The yoke design allows the transfer sprocket to fit in as close as possible to the center line of the frame, yet still provides enough clearance for up to a 28 mm tire—enough for even Clydesdale-class tandem teams to enjoy the benefits of a light weight Paketa tandem. If a single-bike rider is classified as a Clydesdale or an Athena when their weight exceeds 200lbs / 90kg, a Clydesdale-class tandem team would be one that weighs over 400lbs. A 28mm tire is marginal, at best, for a team that tips the scales at 400lbs. Yes, they can get away with it but the tires need to be pumped at or beyond max psi rating to preclude pinch flats which works against the team on anything but the most smoothly paved roads, both in terms of rolling resistance and transfer of road shock.
Thankfully, we’ve already addressed my perceived value or lack thereof of reducing tandem weight and the pitfalls associated with using wide-range cassettes so I think I’m done.
So, in closing let me be clear that I really wanted to be excited by the V2r… and I was until I read the print material which I’m still trying to understand. I’m sure the Paketa V2r is just as good of a frame as the Paketa V2, albeit somewhat limited in gear range / gearing efficiency. But, for certain teams with very specific needs and abilities, it will no doubt be a great choice.
The TandemGeek's Blog 
Well written and I agree completely. I hope you post some of this to the Tandem hobbs list.
Wayne Sulak
Thanks for noticing my (half mine, anyway) V2r tandem. Some of your points are well taken, others are a stretch with not much meaning behind them. But that’s okay. In general, I very much appreciate all the thought and attention you’ve given to this incredible machine. One thing that has me scratching my head is all the fuss about the gearing. This design loses nothing on either the top or bottom of the range, so with 18 additional gears in between, why the controversy?
My husband and I also own Paketa’s V2 (not by any means outdated by the V2r), a BikeE recumbent tandem, a 57-pound single speed tandem, a 42-pound tandem, a couple of mountain bike tandems including a Ventana ECdM, and I could go on. Suffice it to say that we are bike enthusiasts who, even though we are still both working full time, manage to put in 5,000 to 6,000 miles a year. We love all of the bikes for different reasons.
People who design and engineer next generation products do this because they are thinkers, tinkerers, and have an instinctive curiosity about what can be accomplished. The V2r is the result of this. And it is not just for racers or strong recreational cyclists. Anyone who is at the intermediate (or higher) stage in tandem cycling, and appreciates advanced technological features of a bike, would be a good candidate for this machine. I will tell you that I am a lackluster climber at best, and even tend to tense up before embarking on a steep climb. Realistically, at age 60 (and 52 for my captain), I am probably not going to see a lot of gain, but riding this machine up some of Colorado’s steepest climbs is actually fun for me. The same cannot be said for several of our other tandems. Likewise, this bike is incredible on the flats.
If you’d like, see our Garmin ride profile from a couple weeks ago on the V2r.
http://connect.garmin.com/activity/53320496?sms_ss=email&at_xt=4cd6cfcfe2290b12,0.
Is that not enough to demonstrate its climbing ability? I think you should try this bike. The ride and the shifting improvements will speak for themselves.
My issues were with the marketing, not the product. Bear in mind, I am a consumer and tandem enthusiast, not a player in the cycling industry. So my view point is from purely a pragmatic standpoint based on my experiences and observations as a consumer who watches the tandem industry as a distraction from my completely un-related work life.
As for those marketing issues, right or wrong they were:
1. Over-hyping right side drive; nothing new here other than figuring out how to integrate the Gates Carbon Drive sprockets.
2. Glossing over the trade-offs of going from an 11×28 to an 11×36 cassette and giving up the 3rd chain ring.
3. Technical issues with the ‘wear and tear’ pros and cons of single side drive.
4. The ubiquitous “solves problems” that I never recognized as problems or that were problems created by other solutions looking for problems.
5. The silly pursuit of “the lightest” tandem at any cost where less than 1% of the people who buy tandems can leverage that lighter weight.
6. The other marketing hyperbole that appeared throughout the copy.
Again, I think the V2r will be every bit as good as the V2 and the 2×10 gearing will appeal to certain riders. Clearly, you and your pilot’s impressions suggest that the Paketa V2r is an ideal fit for you. I can probably come up with a list of 10 other tandem teams who would also be equally well-served by the V2r… although a few of them would likely switch-out cassettes when they weren’t riding in the mountains and still find themselves grunting out some really steep stuff in that 28.6″ ‘granny’ gear.
As I said, I prefer to use nothing wider than an 11x27t with 53/42/32 chainrings and reluctantly use an 11x32t when we head to places where we’ll encounter several continuous miles of 8-10% grades or the occasional 15%, 18%, 20% or even steeper walls that we have here in the Southeast. For example, here’s an elevation map of one of our local favorites, the 3-Gap ride in Dahlonega… a part of the 6-Gap Century.
Our tandem livery currently includes a ’98 Erickson, an ’02 Ventana ECdM, an ’08 Calfee and an ’09 Ventana ECdM on loan from MTBTandems.com; yes, they’re all sporting triples. We’ve owned a ’95 Santana Arriva, a ’98 Cannondale MT3000, a ’98 Ventana ECdM and a second ’02 Erickson. I’ve gone back and forth between conventional doubles, triples and compact drives on my single road bikes and like the compact drive, but have occasionally found myself missing the triple. That’s my point of reference… well, that and observing what other teams of varying fitness and experience levels ride and and learning what they like and don’t like or struggle with. 2×10 will bear watching, perhaps it will be the next big thing…
Appreciate your counterpoints and the time you took in making the detailed reply. It’s been a long time since I’ve been to Colorado so I don’t have a full appreciation for how the terrain may differ from what we encounter. I suspect in looking at your Garmin data that it’s somewhat different.
A few logs to throw onto the fire:
1) Once I saw the Paketa copy “The combination of 2X10 road and MTB gearing provides the same wide range as a conventional triple-chain-ring road setup” I knew Paketa was dancing around the truth as far as tandems are concerned, because most tandems do not use “conventional triple-chain-ring road setups”. Instead, tandems typically use a mix of a road triple and derailleur in front with mountain bike cassette and derailleur in back. The SRAM Red/XX combo does have the range of the latter, but not the former.
Put another way, the bottom possible combination with the SRAM XX combo is a 36 x 36 (1 to 1 ratio). The actual Paketa ratio is 36 x 39 (1 to 1.08 – slightly harder)
On a Shimano road triple, the bottom ratio is 27 x 30 (1 to 1.1 – harder still)
Many tandems will come spec’ed with the same road triple, but with a MTB cassette for a 32 x 30 ratio (1 to 0.93 ratio, easiest of all)
FWIW, my wife and I are stronger than average riders (for our ages – which happen to closely match Terry and her captain), and have modified our Speedster to a 32 x 28 ratio (1 to 0.88, easier still) We rarely use the “baby ring”, but when we want to use it, it had better be really looowww.
(My wife and I spent a lot of time looking at gear charts before pulling the trigger on a SRAM Red/XX combo to replace the Ultegra triple on her old bike).
2) Terry’s response misses the point that the Paketa design DOES lose bottom end, that’s what Mark spent a bunch of time writing about (it must be a rainy weekend in Atlanta again…)
3) Looking over Terry’s ride profile, I see a typical Colorado big climb: 3000 feet up, but over a course of 15 miles so the net result is less than a 4% grade. On a “shallow” grade like this we would be using our 42T middle ring and probably a cog or two up from the bottom. Terry’s ~ 8 MPH speed on this climb also falls in line with how we would perform on this grade.
We typically would drop into the good old “baby” ring on a sustained 10% climb or a shorter 12%+ climb. The V2r should have run out of gears long before that point…..
Put another way, the Colorado climb profile simply isn’t steep enough to test the gearing range of the Paketa.
BTW, I think Paketa may have dropped the ball on the chainring selections, the wife’s Red/XX single is built up with a 36/53 chainring combination up front and SRAM’s 11-32 cassette, helping the low end while retaining the top end, though some of the jumps are still a little nasty (and might feel nastier still on a tandem).
How’s the fire stoked now?
Update: Let me thank Terry for making me go back and re-read some of what I wrote and to do some additional homework on the gearing.
Based on that re-read I’ve softened my title and some of the comments in my blog entry. As noted in the edited copy, I was actually surprised to see that the gearing is actually pretty tight in the shorter gears, which is a good thing. I really didn’t have an appreciation for just how the 53/39 would play with the 11x36t.
Hopefully my revised text strikes a better balance and takes the edge off my original entry. I felt it was important to do that before I expose this week’s updates to a larger audience.
Hi, Mark,
Nice to hear from you after so long! Thanks for taking the time to read, think about, and respond to Paketa’s web content on the new V2r. While I’ve been at this a while, I don’t claim to know it all, and having another tandem enthusiast critique my work is very helpful: I don’t take anything you say personally. At the same time, I’ll admit you did raise my eyebrows with some of your comments. I pride myself in approaching the design issues in a scientific manner, and combine that with my own experience as an avid tandem cyclist. I’m very proud of what we’ve done with the V2r, and while I may not have time to address everything you discussed, point by point, I’ll do my best to answer the main questions and issues.
Philosophically, I agree with the second paragraph of your blog regarding marketing styles and staying away from innuendo (if not outright criticism) regarding the shortcomings of other manufacturers’ offerings. To the extent that Paketa has no plans whatsoever to discontinue the crossover-drive V2, it’s hard to argue that the marketing of the V2r constitutes “bashing” of anyone else’s tandems. That said, I can offer sound engineering explanations that back up each and every claim in the V2r description you questioned on the web site. I’ve thought about these issues for years, in some cases, and some of the details of the design may take more than a day or two (which about how long you had to analyze the V2r design as the web site just went live on Thursday) to understand fully. The one thing I will concede is that maybe some wordsmithing is in order. For example, the statement, ”The Paketa V2r tandem (patent pending) eliminates all of these problems” might be better phrased, “The V2r…affords improvements in all of these areas.” Other designs and implementations are certainly viable and can work well—our own V2 included. Another way of stating it is, “problems” translates as “opportunities,” and we’ve taken advantage of this opportunity with the V2r. As time permits, I’ll write more about each of the bullet points on the V2r web page, but let me offer up one example for starters.
The issue of belt tension is more than just hypothetical. It was Todd Shusterman at da Vinci Designs who first pointed out to me the potential problem of bearing loads created by the Gates Carbon Drive belt system because of the much higher static tension needed compared to a timing chain. That was at Interbike 2009, at which point Paketa had only recently started offering the Carbon Drive belt as an option on new V2’s and as a retrofit on older models. Sure enough, as time went on we did experience a number of bottom bracket failures. Newer, oversize-bearing bottom brackets might mitigate the problem, but there’s no doubt that the bearing loads are significantly higher with a belt versus a chain. So, how can the V2r claim an advantage here? With the crossover drive, the captain’s pedaling force (dynamic load) adds directly to the static belt force, which is then transmitted to the bottom bracket. With the V2r’s right-side drive, this is not the case, as the static and dynamic belt forces due to the captain is mostly balanced by the chain force pulling rearwards: it’s not transmitted to the bottom bracket, for the most part. If one could measure it, I’m quite sure you’d find that the rear bottom bracket load is barely any higher than the static load when no one is pedaling, or the loading due to the stoker alone when pedaling. Clearly, someone needs to do the experiment—it’s on my list.
I trust the issue of gearing has been settled now? I hope so. A couple things I wanted to point out is that our V2 (no “r”) has been equipped from Day One with a triple32/44/55T setup (FSA’s older tandem offering on the Carbon Pro model) and an 11X26 cassette for a low gear of 33 inches. It’s served us well on some of Colorado’s steepest grades, including Super Flagstaff outside of Boulder (12% max), for example. Going to the V2r’s 39/53 double with an 11X36 cassette actually gave us a lower low gear of 29 inches, so we came out ahead on that one. If that’s not low enough for some teams, there’s other options available (e.g., compact chain rings on a V2r or else a V2 triple setup).
As for the value of dropping another pound off the weight of one’s bike, I would argue that the market is always clamoring for lower weight, and obviously quite a few people are willing to pay for the privilege. The typical cost for reducing the total weight of a ~$12000 tandem (ballpark for a built-up V2r or one of the top models from several other manufacturers) is on the order of $2-4 per gram. For example, perfectly decent aluminum handlebars in the 240 gram weight range run about $80 retail, but if you want a sub-200 gram carbon bar it’ll set you back $300+. Let’s say the weight savings is 60 grams to be conservative; that’s $3.67/gram. Carbon cranksets, lightweight brake calipers, saddles, stems, etc. It all adds up—quickly! If you can knock a pound off the weight of the entire bike with no penalty in strength, stiffness, handling, or durability, that’s worth about $1-2k. While you might not find the tradeoff worth the price of admission, there’s many others who can and will. As my wife, Terry, pointed out, we have a half-dozen other tandems, and while they’re all special in their own way—we ride each and every one of them and enjoy all of them immensely—for us, the light weight V2 and V2r are absolutely worth the price. Do we need it? No. Do we like it? Yes! Apparently we’re not alone in that sentiment.
I’m actually a big fan of the Rohloff SpeedHub, and yes, I did see two Rohloff-plus-Gates Carbon Drive belt-equipped (timing and rear drive) tandems at Interbike. While I, too, appreciate all the advantages that a Rohloff hub brings to the table, there’s also a downside—two, in this case. One is simple: weight. The Rohloff hub is heavy, and as much as I admire its fine engineering, I’m put off by that. The second issue is, the Gates Carbon Drive belt is not approved for tandem use on the rear drive, so while it might look great on a show bike, it remains to be seen whether it’ll hold up in real-world riding. You could always run it with a chain to the rear, in any case.
There’s a gestalt to the V2r design that can only be truly appreciated by riding it, after which you may even come around to the point of view that the “marketing hype” in the web description just might be for real. Mark, I really think you need to visit us in Boulder and ride the V2r yourself. To paraphrase an old saying, “Riding is believing.”
Dave Walker
Paketa Cycles
Hey Dave:
The bashing comment pertained to the polarization of the proponents and critics of the brand in question, and how any critique of the marketing spin is to this day quickly labeled as bashing the brand when, in fact, it is the claims that are typically called into question not the products. Perhaps in my editing I lost some of the context, which is still focused more on the material that describes the product than the product itself.
I’m still not sure I can appreciate how a belt running on a 69t sprocket that’s similar in diameter to most sync chain timing rings can cause a marked increase in bottom bracket wear just from the slightly higher static tension required to keep the belt from skipping on the sprocket under load. I originally thought that would intuitively be the case when I first inspected Bob Thompson’s belt drive back in May 2006 and gave the cranks the old ‘spin’ test to see how much drag the belt put on the bottom brackets without any pedal loads: the drag was in fact higher than a sync drive with the typical 1/2″ slack, but not alarmingly so. However, it finally dawned on me that those static loads should be inconsequential compared to the loads that a pilot puts into the sync chain or belt when actually riding the tandem, particularly under maximum efforts on out of the saddle sprints and steep climbs. Assuming the sprockets and timing rings were the same distance from the bottom brackets, why would the belt increase the loads on the rear spindle?
Interestingly enough, I’ve used Todd’s daVinci cranks exclusively for the past 12 years or so on five different tandems; three road & two off-road. Those 34t timing rings have most certainly put higher loads into the bottom brackets vs. a 42t timing ring to no ill effect — well, if I ignore a custom Ti BB spindle that flexed so much it damaged both of the adjusting cup’s dust seal beds. However, that particular BB was a poor choice for the stoker’s BB in the first place and since then we’ve never had any trouble with excessive wear & tear on Shimano UN72, RaceFace & Phil Wood square tapers.
So, getting back to the belt drives….the 1:1 direct drive sync chains & belts clearly don’t generate anything close to the loads on the BB axles and bearings compared to the drive side given the amount of torque generated by something like 300lb tandem team climbing a 15% grade in a 30t granny / 34t rear sprocket. So, I’m really not ‘getting’ how the hand-tensioned belt causes premature wear to a bottom bracket designed to deal with massive loads from the drive side. Perhaps it’s the way that the newer outboard BB’s are designed, i.e., have they over-optimized the non-drive side for single bikes such that the non-drive side bearings just don’t deal well with constant static loading? I’d love to understand the mechanics underlying the belt-induced failures better, as it just doesn’t make sense to me.
As for the single side drive, there is clearly a mitigating effect on sync drive loads by the drive chain. But, the nature of a tandem puts a multiplying effect on the drive chain loads that exceeds the pilot’s inputs on the sync drive, particularly in those very short climbing gears. So, I don’t disagree that right side drives are easier on the rear bottom bracket — so is riding 90* out of phase. But, as I said, if you ignore the more recent bottom bracket reliability issues and look back to when square tapers were fitted to just about every tandem, crossover cranksets weren’t really tearing up the better quality square tapers like Shimano’s UN72s or even the UN52s. We and many of our riding friends put 10’s of thousands of miles on those BB’s without any failures. In fact, it would be interesting to know what kind of BB failure rate Todd has seen on daVinci’s tandems since they all use square tapers instead of Octalink, ISIS, or any of the newer outboard bearing designs. I’d ask what Todd’s daVinci failure rate was with Gates belt drives, but I’m pretty sure the belts aren’t compatible with daVinci’s ICS drive tandems.
As for the low-gearing, I guess what jumped out at me was how a lot of high-end tandem consumers who really aren’t high-end performance riders end up with bikes that are ill-suited for their needs. While there are folks out there who can climb a 12% grade in a 33 inch gear, there are a heck of a lot of them out there who can barely do it in a 23″ gear… and I see a lot more high-end, uberlight tandems hitting the road each year with those wide-range triples. So, again what struck me was the soft-sell on the gear-inch limitation where the headline for the bike was, “The new Paketa V2r Tandem is lighter than ever“. Well, technically it’s probably the lightest Paketa tandem yet, since the V2r can’t be lighter than itself… unless it was quietly introduced as a right hand chain drive 2×10 before it was offered as a belt drive.
Anyway, as to why all that struck me the wrong way, I’ve already been on ‘high alert’ by what I continue to call the silly pursuit of the sub-20lb tandem that seems to be going on. Yes, I understand that as builders you’re compelled to offer clients what they want and sub-30lb tandems, then sub-25lbs tandems became all the rage as the weight weenies turned their focus from 12lb single bikes to tandems. So, right from the git-go we’ve had the belt drives hitting the market and showing up in all kinds of curious applications where the goal is shaving off another pound for the sake of bragging rights, as it’s still the strongest teams who are winning events regardless of what they ride. For example, old school aluminum and titanium tandems with conventional wheels were used to knock-off the uberlight all-carbon bikes with deep section composite racing wheels and other high-tech gear at the Paralympic Games in Beijing. Yes, the belts are also nice in that they promise to provide lower maintenance, are quiet and may be more cost-effective than chains if the belts outlast chains (and if they don’t chew-up bottom brackets), but with the exception of a few folks who live in the PNW and ride in wet conditions all the time, the real attraction continues to be lighter weight and solving the problems associated with chain drives. So I’m looking at the V2r web page write-up and once again the spin is, “current tandems are fraught with issues and excess weight that are keeping you from being a faster team or enjoying a problem-free riding experience”. Again, it’s not the product that turned me off, its that type of heavy-handed marketing spin.
Seriously, on one hand you’re still offering the V2 crossover tandems, but on the other you’re telling potential clients who probably need a V2 that its crossover design is flawed and obsolete. That’s a bit of an exaggeration, but in essence it’s true. After that, it was game on… Yes, a lot of it is nit picking, but once again, it wasn’t pointed at the product, just the words used to promote the new product’s “benefits and features”.
As for us, the V2r wouldn’t work as configured: remember, we need that triple with at least a 32t low gear on our cassette for a lot of our steeper climbs here in the Southeast… not all the time, but when we need it, we need it and we’re still probably in the upper 10% of hill climbers among the folks who attend tandem rallies, if only because we’re a sub-300lb team. That said, it would be interesting to see how the drive train worked on less demanding terrain but, then again, we never touch our granny gear and essentially ride a ‘double’ with a 27t cassette on terrain like that today.
I guess that brings us back to the belt-drive and lighter weight as the big discriminator and that doesn’t really play either since — with the exception of an experimental belt drive we’re playing around with — the 30″ to 31.5″ stoker compartments of our tandems that Debbie prefers isn’t compatible with a Gates sync belt.
Bottom Line: I’m sure it’s a great ride… and it will sell well. The market seems to like new and improved, and I can appreciate that. And, at the end of the day, as they say any press is good press, even if a mis-guided enthusiast wanders off the reservation a bit. I’m sure you’ll correct me where my thinking is flawed… it won’t be the first time nor the last that I find myself lacking in my understanding of things that exceed my knowledge and experience.
Warm Regards,
Mark
Hi, Mark,
That was a well-considered and reasonable reply; thanks. I’ll say again that I do pride myself with integrity (leave out the guts for now) and I don’t want to slide down that slippery slope of marketing-triumphs-engineering any more than you want to see another manufacturer fall victim to that siren song. I’d love to spend some time with you to discuss the various aspects of the design AND marketing to strike the right balance.
As for spin versus substance, we both know that design and engineering is almost always incremental, and I won’t shy away from saying the V2r IS an improvement over the V2 and similar conventional crossover-drive designs, but it’s just that: incremental rather than causing obsolescence.
Gearing: Yes, you’re right that if you use standard cranks (e.g., 39/53) then you’re limited right now to a low gear of about 29 inches (depends slightly on tire size) with a 36T cog in the rear. But, as I said before, if you want or need lower gearing you can go to a compact crank on the V2r or get a triple-equipped V2 (for example). I understand fully the issues of big-jump front shifting on a double. It can work, but is it optimal? If you’re willing to live with a lower top gear and use a 34/50 (seems to be pretty conventional for compact cranks these days) then no problem. There’s always a trade off. That’s what engineers do; assess trade offs. I rather enjoy the challenge, too.
On bottom bracket bearing lifetime, all I can say is, I’ve seen failures with the Gates belt drive, but that doesn’t mean you can’t live with it: different/larger/better BBs appear to be working fine, even on our own tandems. My concern with promoting the concept of the right-side drive design and standard single-bike cranks and BB’s is that some might argue that they’re not “up to the task” of tandem usage (didn’t you say something to that effect originally?). My aim is to put that issue to rest: the V2r does NOT experience significantly higher bearing loads than a single bike, thus single-bike cranks and BBs are perfectly suitable.
One other quick comment on your original blog. If putting the Gates belt drive on the right side is so obvious, then why hasn’t anyone else done it heretofore? Even the dual-belt-drive Rohloff-hubbed tandems I saw at Interbike (both stands) had the timing belt on the left side. Furthermore, you mention Calfee as having a relieved LEFT chain stay to make room for the belt. Why didn’t they put it on the right? I’d argue that the best design improvements often are the simplest ones–and often appear obvious after the fact.
OK, it’s incremental, but it’s also useful, more efficient, and performs beautifully. A $12k tandem will never be one-size-fits all, but then again, neither is a $2k tandem. Agreed?
All the best,
Dave
Back at ‘ya Dave:
I’m not ready to say ‘improvement’ just yet, but it’s clearly an alternative approach to a tandem drivetrain that represents an innovative way of incorporating new components (right-side belt & 2×10 drivetrain in combination) in a way no one else has (at least to the best of my knowledge). That was, in fact, what originally excited me about the V2r. I’ll also agree it reduces component weight and simpifies shifting for tandems that will be used on all but some of the steepest and challenging climbs. If we had nothing steeper than 10% grades to contend with, the 2×10 would be very tempting. I’m still not fully sold on the belts, aside from folks who find themselves riding in the rain and who would benefit from the promised reduced maintenance and longer service life of a belt vs. a chain in those conditions. I say this as someone who’s only worked on one stock Gates Carbon Drive system and who’s put perhaps 700 miles on a different approach to the Gates Carbon drive as a favor to a friend. Hopefully it won’t damage my very expensive Mag-Ti BB’s in the process.
In regard to the latter, I’d be interested in learning more about BB failures attributed to the Gates belt drives. More specifically, what types of BB’s, the nature of the failures, at how many miles, etc. I’ve helped one Belt owner vanquish nagging creaks from his FSA MegaExo cranks & bottom brackets after his first set were replaced under the premise that they were trashed in just a few thousand miles. In working on his cranks it became pretty clear that unless somone was using a torque wrench and following FSA’s instructions to the letter it’s doubful that they’d put enough torque into the BB’s / cranks to meet the spec’s. When last I checked, their BB’s have remained quiet ever since. However, we also ran into a team at a tandem rally using a crossover crankset with sync chain & rings running FSA MegaExo cranks that had the exact same problems as the sync belt equipped tandem. Unfortunately, I did not have the tools I needed to tear apart their BB’s which had also been recently replaced under the same premise, i.e., premature death. Other than the long-standing issue with a certain brand of tandem using very wide 72x129mm BB’s that seem to have durability issues, most other Octalink and the better square taper BB’s seem to hold up pretty well. ISIS seems to be a mixed bag. But, again, I’ve not been privy to any failure data that is unique to the Gates belt drives that would be indicative of a systemic issue. As I said before, I can’t for the life of me understand how a bottom bracket can tell if it’s been stressed by a 180lb pilot putting out 300 watts of power through a chain or a belt on similar diameter rings/sprockets and why the minor static load from the belt’s tension is so detrimental to a BB that’s designed to deal with those huge amounts of torque.
As for certain double cranks not being up to the task, my only caution is where a crank has a given maximum rider weight, e.g., Campy typically imposes a maximum rider weight of 180lbs on their carbon cranks (as well as their wheels, etc.). Frankly, I think they’re being overly conservative given that 180lb Pro like Thor Hushovd or 150lb Mark Cavendish is clearly putting a lot more stress on cranks than a 220lb Clydesdale would. So, whenever someone is considering a crank that hasn’t been used on tandems I always suggest they contact the manufacturer to ensure they don’t have any reservations about using a crank for their teams’ combined weight, gearing and power. It’s no different that folks wanting to use a non-tandem specific forks, aero wheels, or as you note… a Gates drive-side belt on a tandem.
As for same-side drives, I don’t think anyone has done it that often because crossovers have become the ‘norm’ and, as I’ve said, haven’t really demonstated any systemic issues with boom tube deflection, premature BB failures or other things about which we’ve already shared comments. That’s not to say there are some teams out there running right-side-drives: their are. But, they typically live and ride in places like Florida where a double crankset on track and single speed tandems. In some respects, I’m sure it’s somewhat akin to why 99 out of 100 tandem teams ride in-phase instead of out-of-phase, noting that riding OOP may actually be “better” in terms of it’s efficiency and reduced drive train wear and tear. It’s also why Todd’s ICS tandems have such a hard time gaining broader acceptance. So, given that crossover cranksets have become ubiquitous, there are very few builders out there who probably gave much thought to doing what you’ve done. With the propogation of 2×10 wide range cassettes, perhaps the right-side-drive will become the next big thing and as I’ve already noted, you appear to be the pioneer in this regard.
Like I’ve always said, Like what you Ride, and Ride what you Ride, but by all means get out there and Ride.
TTFN, Mark
The problem with engineers is that they’re always trying solve problems…hopefully, ones that exist rather than ones that are merely hypothetical. I’ve done the calculations; the belt tension issue is real.
Let’s assume that a strong (not Olympic-class, but representative of real-world riders) cyclist puts out 1000 Watts in a serious if not all-out effort. That’s about 5X the average power of a moderately high effort for a 1-2 hour ride.
The average force applied by the rider through one revolution of a 175 mm crank at 90 rpm (again, a reasonable number for a hard climbing or sprinting effort) is 606 N = 136 lb-force. A 69T Gates belt-drive sprocket is coincidentally 175 mm in diameter, so the force on the belt is twice that on the crank arm, or 272 lb-force averaged over a revolution.
The recommended tension for the Gates belt is 110 lb-force, multiplied by 2 since there’s two runs of belt, top and bottom, equals 220 lb-force on the drive sprocket. And, it’s always there. Thus, the maximum BB load is about evenly split between the rider and the static belt tension.
What’s the average power output over an average ride? 100 W? That’s an average 27.2 lb-force added to the belt tension–in other words, overall it’s the rider’s contribution that is negligible, not the static belt tension.
So, the average bearing load of a belt-drive tandem is several times higher than a chain-driven tandem in normal use. Even if you make the simple assumption that the fatigue life is inversely proportional to the load for a given number of cycles (it’s something in between exponential and linear in reality), it’s not hard to see how the belt tension is a significant, if not dominant, factor in bearing life. For over-designed bearings, maybe it doesn’t matter to most people: it might mean 10,000 km lifetime instead of 40,000 km, for instance. In the former case, if you only ride 1,000 km/year a ten year lifetime is plenty adequate; for us, it would mean replacing the BB annually.
The advantage of a right-side drive train such as that on the V2r is that the maximum bearing load is approximately that due to the static belt tension only, whereas on a crossover drive the maximum load is 2X as high, maybe more. That could be significant in some cases.
Dave Walker
Where did the 110 lb-force value for the sync belt installation come from? I’ll have to do some checking on that one, as that seems like an awfully high number…
I’d characterize the tension I’ve used on the sync belts with 69t sprockets as something just a little bit higher than I use on sync chains, with a bit higher tension than that on the very small 33t sprockets I’m testing. After all, how much torque can you apply with a Park SPA-1 pin spanner before the thing pops out of the eccentric?
In use, just like a chain all of the tension resides in the upper run of the belt with very little if any in the lower 1/2. Moreover, having had an eccentric slip during a ride, we were actually able to ride and climb hills up to 8% some 10 miles back to the house on the slack belt without causing any belt slips so long as we didn’t stand and used low gears… and this was once again on the very small 33t sprockets. The 69t sprockets would generate a lot less tension and have far more “grip” under the same conditions. So, as I said, that 110-lb force number comes as a surprise to me. I’ll have to do some checking on that one.
BTW, is the new yoke on the V2r the main driver of the $900 frame-only cost difference from the V2? Or, are there some other refinements / differences in the V2r frameset?
The recommendation of 110 lb-force came directly from my own inquiry to Gates.
You comment to the effect that “…just like a chain all the tension resides in the upper run of the belt…” is patently ridiculous. With a chain, yes, the static and dynamic tension is very low (almost zero) in the lower run. With a belt, that’s not at all the case, assuming you’re following the manufacturer’s recommendation–advisable, I might add.
Dave Walker
Paketa Cycles
Perhaps I’m missing the boat completely then. How exactly is the 110 lb force value used; with a digital force gauge?
As a consumer / end-user the only “measure of force” I’ve ever found for a Gates Carbon Drive Belt — and they don’t seem to differentiate for a drive belt and tandem’s sync drive belt — is, “Proper tensioning is achieved by deflecting the belt 1/2” with approximately 5-10lbs of force. With any tensioning procedure, it is recommended to rotate the drive and recheck tension several times. This is because tension will vary within the drive due to part tolerances. ” Gates goes on to note that, “Too high of tension can cause damage to your bearings. It can also increase wear on your belt and sprockets. You may notice too high of tension through a perceived drag feeling in your drive.”
Perhaps I need to get a Gates Tension Tester, because when following their instructions and setting the tension using dead-reckoning as to what 5-10lb feels like (and something even higher on the 12mm wide belt I’m playing with), I still find if I hold the rear brake and apply downward pedal pressure on the front crank / driving sprocket to load up the drive train, the tension in the top run of the belt increases as tension drops in the lower run. No, the lower run’s tension doesn’t drop to zero like it does with a chain (an unintended overstatement on my part in my last reply), but there is a noticeable and significant change in belt tension between the two. My assumption is, this is the collective effect of various different sources of elasticity in the bottom brackets, some frame deflection, and residual belt / sprocket elasticity or modulus under full load. Does this not happen on drive and sync belts on your bikes and V2 tandems? I would expect a single speed, fixed-gear or Rohoff equipped single bike’s drive side belt to exhibit the same behavior, but don’t have one on hand to play with.
Is this, perhaps, another “improvement” to a right-side tandem sync drive system that will reduce wear and tear on the belt itself if, in fact, the loss of pre-load is occuring with the left-side drive while in use? Again, I would think the right-side drive would also have the same canceling effect of asymmetrical loads on the rear bottom bracket spindle as well as some of apparently drop in belt tension pre-load. Just curious now that we’ve gone down this road.
Yes, I’ll admit this single-rider / driving sprocket – driven sprocket example is extreme since, in a perfect world, a stoker’s pedaling inputs — if equal in effort / power to the captains — would equal the pilot’s and cancel or at least significantly diminish the driving / driven-sprocket effect. But, back to my example of hammering up that 15% grade, like it or not I know that I’m typically putting a lot more power into the drive train than my petite stoker.
Again, I’m just trying to understand and correlate the 110 lb-force figure and the behaviors I’ve observed when applying enough sync belt pre-load to achieve the specified 1/2″ belt deflection using gates single-finger, downward pressure tension check.
If, indeed, a lot more pre-load is needed vs. what I have come to believe is the proper pre-load, then I would agree that my entire premise on the potential, detrimental effect a sync belt has on bearing life is wrong. That being the case, I will certainly make amends in my original blog and a follow-up to set the record straight.
Note: I’ve revised my original blog entry twice based on the on-going dialog with Dave Walker from Paketa that has been informative but also raised more questions in my mind than it has as of yet solved.
However, I am starting to turn the corner on understanding why the single side drive drive may be a better adaptation of the Gates Carbon Drive vs. the more commonly seen crossover crankset design. So, somewhat with hat-in-hand, I’ve softened the original entry a bit and will be posting a subsequent one in the near future to relay what I learn.
I am sincerely grateful for the time Dave has taken and his patience in responding to my observations and questions.
Hopefully, more to follow.
I appreciate your comments along with your curiosity; thanks, Mark. I don’t claim to have all the answers either, and some of your questions have sure got me thinking, too, which is good!
The physics of the problem are that whatever the static belt tension is, it’s even between top and bottom runs–obviously. What makes a belt different–very different–from a chain, though, is that it’s a pre-stressed mechanical structure; like a spoked wheel, come to think of it. Both the top and bottom runs of the belt contribute to propulsion, in a manner that’s not obvious if you’re used to thinking about chains.
Step on the pedals and, you’re right, the belt tension in the top run increases while that in the bottom run decreases by about the same amount. I say “about” because, unlike a spoke (to continue that analogy) the belt is highly nonlinear in its stress-strain curve i.e., the amount it stretches depends on the initial tension or strain. If you make the simple (and not really correct) assumption that the belt IS linear, then the increase in tension on the top run and decrease on the bottom run would be equal–right up to the point where the bottom run has zero tension, at which point any increase in pedaling force is seen as a further increase in the top-run tension alone.
What does all this mean as far as a left- vs. right-side timing belt? For the left side arrangement, the bearing load is almost always just that of the static belt tension (using the simplification above regarding a linear belt characteristic). The only exception is when the captain’s pedaling force is high enough that the bottom run goes slack, in which case the bearing load goes up proportionally from there. The load due to the stoker’s pedaling only increases the rear BB bearing load in all cases, due to the bending moment.
With a right-side timing belt you still have the same load on the front (captain’s) BB, obviously. It’s the rear BB where things change dramatically. Any force on the drive chain (or belt, if you’re running a Rohloff hub or the like) reduces or compensates for the timing belt load AT THE SPIDER, since they’re opposed and all mounted at the same point. So, as soon as either rider starts pedaling, the rear BB bearing load due to the belt goes down, not up. This is true no matter how hard either rider pedals, too (within reason). The bending moment on the rear BB due to the stoker’s pedaling is also present, same as the left-side case, independent of the belt effects.
Bottom line: Generally speaking, with a left-side timing belt the rear BB load starts at some moderate level (which, by the way, can be fairly high compared to a conventional timing chain) and INCREASES due to pedaling input from either rider. With a right-side timing belt, the rear BB load starts at the same moderate level but DECREASES due to pedaling forces. Which would you rather have?
NB: The belt force pulling forward and the drive chain (or belt) force pulling rearwards due to pedaling are only exactly balanced if the belt and drive sprockets are the same diameter, but for most practical purposes that’s a minor perturbation. On the V2r, the inner chain ring (39T) is slightly smaller than the belt-drive sprocket, while the outer ‘ring (53T) is slightly larger.
So, overall, the rear BB loading is lower with a right-side timing belt compared to a crossover-drive arrangement. Does it matter? At some point, yes, but I’ll reiterate the crossover drive works, too–we have a left-side belt drive on a Paketa V2 and it’s been quite reliable.
I’m still waiting for a response to a couple more questions I’ve posed to the good folks at Gates, as I, too, question whether 110 lb.-f on the belt is really the right number. When I’ve installed tandem belts, I’ve had to use a lot more force than a pin spanner (I have the same Park tool you showed in the photo above) provides; I usually use a couple hex keys in the eccentric holes, then a big, long screwdriver across them to apply enough torque to get the belt tension high enough. The “green range” on the Krikit gauge I have agrees with the 5-10 lbs = 1/2″ deflection recommendation, but…I’m skeptical that’s anywhere near 110 lbs. of tension. I’ll let you know what I hear.
Ride on!
Dave
WOW! That’s all I can say to you engineering geeks!
Dave – bike looks sharp, and the concept of 2×10 with belt drive sounds interesting enough. I can’t comment too much on all the technical details though.
Mark – thanks for all the time you spend on the tandem blog.
Dave and Terry, we gotta get out and ride!!!! too bad you missed the mtb tandem weekend in Winter Park again this year.
Hey Dave (I think you’re still subscribed to this thread):
While Debbie and I were out for a ride from the house yesterday we were talking about your Paketa V2r, the discussion you and I have been having and at one point I found myself revisiting this invitation and comment:
Three thoughts in particular came to mind that we talking about. The first was, “Hey, I wonder Dave would be willing to loan us the V2r for a few weeks this winter?” The second was, “Of course, I wonder if it would even fit: probably not… we’re pretty small.” The last one was, “You know, we could simply demo a representative right-side / 2×10 drive train on our Calfee or the Erickson this winter“.
From a cost standpoint, the last one is probably a wash for the shipping costs since we have enough daVinci cranks and chainrings + an older XTR rear derailleur to support a chain-based right-side drive / 2×10 drive train. The only thing I’d need to get is a 11x36t cassette (which we’ll re-use on our Ventana off-road tandem) and a Jtek Shiftmate #3. So, perhaps will give it a go.
Anyway, I’m working up a new Paketa V2r entry for this week that attempts to ‘update’ readers who may have missed our dialog here in the comments section and some other thoughts I’ve had this week. In fact, I’ve had so many thoughts that it will probably take two different entries to capture that information.
Hi, Dan (& Bevin)! Yeah, sorry we missed the Big Bike mtb ride in WP, too–we’ll have to make sure we’re not off somewhere else…riding. Tahoe was great; even did the Flume and Ridge Trails on our ECdM.
As time permits, I’ll (try to) answer more questions Mark raised on the V2r. Starting with the belt tension, the follow-up to my inquiry to Gates is they’re recommending 70-80 lbs. tension on the belt, NOT the 110 lbs. they told me originally. This is still a whole lot higher than the “green range” on the Krikit, if you have one of their tension testers. I always like having the right tools, so I got one in lieu of using the simplistic 1/2″ deflection = 5-10 lbs. rule. If you went with the middle of their recommended range, 75 lbs., then the static force on the belt sprocket is 150 lbs. (again, because there’s two runs of belt, top and bottom). That’s still a lot: it’s the same (average) load on the BB bearings as a rider pedaling at 550 Watts output at a cadence of 90 rpm, to continue with the example described previously.
Mark–The offer to come out to Boulder and ride with us still stands, of course. Atlanta’s a ways away, I realize! Right now, the white V2r on the web site is the only one that’s available (that’ll change soon), and it’s our bike. Yes, you could set up one of your other bikes with a right-side drive, but there are a few caveats. You want to get the timing chain or belt as far inwards as possible, as that reduces the bending moment on the boom tube. But, if the way you do it is to use a small chain ring on the inside, that raises the force inversely proportional to the chain ring size, so try to keep it as large as possible and still clear the chain stay. Does that make sense? In order to put a belt on the inside, you probably have to design the frame for it, for reasons that should be obvious.
The other question you asked that I thought I’d answer presently is about the increased cost of the V2r. There’s several reasons for that. First, the CNC’ed yoke is a specialty item, manufactured in very small volumes, and the engineering and setup costs are substantial. Also, there aren’t very many shops that are set up to machine magnesium, as special safety precautions must be taken to avoid fire hazards caused by the chips thrown off (FYI, it’s NOT a hazard in the form of a bicycle frame or even a tube), so that raises the cost, too. Then, there’s several more hours’ labor involved in fabricating the frame for the extra alignment and multiple welds involved with the yoke. So, yes, it’s more expensive to build, plain and simple.
Cheers,
Dave
On the belt tension, that’s getting better. However, I’m still struggling a bit with the the ‘doubling’ of the tension, assuming by tension they mean the preload.
I’ve got it in my mind that if the pre-load is 80 lbs, then the pre-load IS 80 lbs where there is 40 lbs on each span of the belt when the “system” is static, i.e, there is no loading other than the pre-load specified by Gates. I tried to find an example, and the closest thing I could find is the following:
As for doing a home-built single-side drive, I understand the entire principle of keeping the sync (and drive) chain lines as close-in to the center line of a frame as practical. Given that we’re already using daVinci’s 34t chain rings and I run either a 30t or 32t granny ring, something in that range will work IF we decide to play around with the entire configuration to see if we can detect any significant difference between a crossover and same side chain as well as to get a sense of how ‘friendly’ the 11×32 is in a double chain ring configuration, i.e., how do those larger gear ratio jumps actually “feel” compared to the closer ratio cassettes we’ve grown accustomed to using. No interest in playing with belts beyond the experiment I’ve been doing for our friend in Florida.
Appreciate the insights on the cost drivers for the V2r; makes sense to me, just wasn’t quite sure what all was entailed that drove the cost difference between the V2 and V2r frames and now I do.
Thanks and sorry for the seemingly never-ending string of questions. Just trying to sort out what I thought I’ve learned vs. what I’ve seen/experienced and what I’m reading in your replies.
Help…I just sold my trek T1000 and want to build a budget lighterweight road tandem. I am intrigued by the 2X10 one side drivetrain. I have never had a a problem with the std setup. Can I simply buy a triple crank and put a larger chainring in the granny spot? Also I see many tandems do not have the extra stabilizing tube…Are there any pros and cons
Single-side drive; yes… a Triple crankset can be used with the sync chain place either in the big chain ring position or the granny chain ring position. Using the outer ring position elimiates some potential drive-train lock-up issues in the event the drive train overshifts and drops onto the sync chain and it also yields a better chain line on 145mm or narrower rear drop-out tandems. However, the aesthetics aren’t the greatest with that configuration. Using the inner-most / granny ring is a cleaner installation but requires careful attention to chain ring positioning and front derailleur stop settings.
As for the open frame designs that elimintate the internal bracing tube that became very common on most tandems produced since the 1990s, if the builder has done a good job of selecting his tubing and keeping the frame as compact as possible there aren’t too many performance issues for the small to average size teams. However, there are some upper limits and a good tandem speciality dealer and/or builder will be able to work with most clients to help determine if they are well-suited for those designs or might be better suited to have a frame with an internal bracing tube. One of the more common “Cons” of the open frame designs is the loss of water bottle mounting positions for the stoker. It’s not insurmountable, but does require some adaptation. At the end of the day, the open design really represents a way for builders to produce a tandem that weighs a bit less without giving up performance and less expensive to produce. As a consumer, I’m not sure the cost-savings washes through to what you pay for the bike unless you’re having a travel tandem made, in which case there is definitely a cost-reduction associated with eliminating a few couplers.
Hi, Mark,
I don’t see an easy way to respond directly to this latest entry via the web site, so I’m replying here; sorry if that’s a bit of a hassle.
Since all this happened in 2010 there’s been quite a bit learned along the way, which isn’t too surprising with any new design. We’ve made some slight modifications over the intervening 3-1/2 years, but the basic concept remains the same: put the timing belt on the inside of a 2X10 drive train. Since 2010, the main substantive change is that we (Paketa) now offer a compact-crank adapter so that the timing belt sprockets, which are 130 BCD, will work with a compact 110 BCD crank for the rear (stoker) position. This addresses some teams’ needs for lower gearing than what you can get with a 130 BCD crank (38T minimum).
I really don’t see any advantage to putting the timing chain (or belt) on the outboard side of the right crank+chain rings, quite frankly. It’s not so much a matter of aesthetics, but functionality. By far the most important advantage of putting the timing belt on the inside is to reduce the boom tube flexing under pedaling loads, which makes the bike feel stiffer without adding any more material to the frame. You get three advantages in one: better perceived stiffness under pedaling-induced loads, lower weight, and better ride quality (since the frame can be built lighter). Putting the timing chain (or belt) outboard would probably increase the boom tube bending load compared to a “conventional” left-side-timing arrangement since the offset from the frame center line is going to be greater, assuming you’ve done everything else correctly. Given a choice, I’d rather use a conventional left-side timing setup than a right-outboard timing setup.
Putting the timing inboard-right using a triple crank, as Brent proposes, isn’t a good idea for the reason you allude to: I’ve heard of several people who’ve tried it with a timing chain and have ended up abandoning the idea due to drive train lockup under certain conditions. Yes, it’s likely due to fine adjustments in the front derailleur and avoiding dropping the chain off on the inside, but it’s been problematic at best and a deal-breaker at worst for those I know who’ve tried it. A timing belt, on the other hand, works just fine since the chain can hit the belt during shifting and it simply slides off the smooth polyethylene belt with no sound or upset. But, Gates tandem timing sprockets only come in 130 BCD, which creates a major problem in fitting all of the hardware in the limited space on most frames around the rear BB/chain stays. I don’t know of any production frame besides the Paketa that has the special features that allow this to work with a conventional drive train (130, 135, or 145 mm with double chain rings) while keeping the chain line correct. Maybe someone else has done it, but if so it’s news to me.
The comments about the open frame design Paketa uses (along with numerous other manufacturers these days) deserves a comment or two. If the frame is designed properly, the open-frame style is generally superior to the direct-lateral style frame (that being the proximate comparison). Sure, if you take any typical direct-lateral tandem frame and simply delete the lateral tube, then of course it’s going to suffer in stiffness even though the weight is reduced; that much is obvious. The proper comparison, though, is between two frames of the same material and overall geometry of the same weight, one with and one without a lateral tube. What’s the better use of material; to put it into the lateral tube or put more into the other tubes? That I can answer unequivocally: it’s much more efficient to put the material into the other tubes (mainly the down and boom tubes) and eliminate the lateral. In the early days of Paketa tandems, we built two otherwise-identical frames, one with and one without a lateral tube. In this “experiment,” because of the tubing we had available, the lateral tube was the same as the top tube (front and rear). I made a lot of careful frame-stiffness measurements on these two frames (along with several other frames from various manufacturers) and determined that the bending stiffness (a combination of lateral and torsional bending) was 5% greater for the frame with the lateral tube, but the difference in frame weight was 20%. I.e., in a stiffness: weight calculation, the extra material put into the lateral tube was not an efficient use of material. So, following that test, we increased the diameter of the down tube, increasing the frame weight by only a few % while improving the stiffness by about 10%–in other words, the open-frame tandem we ended up putting into production is actually stiffer than the direct-lateral frame, yet lighter weight.
The one proviso in all this is that you need to have the flexibility to play with tubing diameters and wall thicknesses to take advantage of the material and optimize the stiffness:weight ratio. With steel, for example, you’re severely limited in how large you can safely go on the diameter without adding weight, required to avoid denting or, even worse, buckling under load. Titanium is better than steel in this regard, and aluminum is better still due to its lower density. Magnesium is even better: you can basically go as large as you like on the tube diameters while still keeping the tubing wall thickness adequate for good dent resistance. Carbon is a different matter. In that case, the limit becomes fracture resistance rather than dent resistance: the tubing needs to be thick enough to avoid cracking under moderate impact loads, the consequence being that the very best carbon tandem frames weigh about the same as a magnesium Paketa tandem frame. At that point, the choice becomes somewhat subjective as to which has the better aesthetics, ride comfort, and utility for the intended purpose.
Dave Walker PAKETA CYCLES
>________________________________ > From: The TandemGeek’s Blog >To: sixtiescycles@yahoo.com >Sent: Wednesday, May 28, 2014 10:11 AM >Subject: [New comment] Paketa V2r… A Cool New Offering Tests My Understanding Of Belt Drives > > > >WordPress.com >TG commented: “Single-side drive; yes… a Triple crankset can be used with the sync chain place either in the big chain ring position or the granny chain ring position. Using the outer ring position elimiates some potential drive-train lock-up issues in the event the d” >
Hi, Mark,
No problem; the goal for all of us tandem enthusiasts is to understand how they work. I found the web site you refer to, and while there’s nothing wrong with the definition they use for belt tension, it’s just that: a definition, or convention–nothing fundamental in the physical description. By their definition, the belt tension is the total load on a sprocket; i.e., the tension is 1/2 of what I used in my analyses previously, where I assumed the belt tension is in one run only. All else being equal, I’d be inclined to use whatever is considered conventional in the trade, so we’ll go with their definition.
Assuming the optimal belt tension is 75 lbs. (total), then by their definition that’s 37.5 lbs. in each run of the belt. This equates to 275 Watts power output from the captain at 90 rpm, using the same numbers as before. That’s still a pretty high power output, overall, but given that the maximum output can be much higher–as much as 4-5X as high for a typical-but-not-Olympic-class rider, this really calls into question the recommended “tension ratio,” as described in the subtopic, “Mah Harley Don’t Need No Preload” at http://www.epi-eng.com/rotorway_helicopter/rotor_drive_system/gates_belt_issues.htm#preload. Funny, but Harleys and tandems have more than a little in common. They both have high-torque, low-rpm (relatively) propulsion systems. The tension ratio on a tandem belt drive tensioned according to Gates’ recommendation is severely compromised during aggressive sprinting, but not knowing the design parameters of the belt, perhaps that’s alright given the low duty cycle; i.e., the wear on the belt caused by very brief sprinting intervals is negligible in the grand scheme of things, and is outweighed by the lower average load on the BB bearings by using a lower belt preload? OK, that sounds reasonable to me and I’ll buy that premise.
But, if you follow that logic, then the V2r design looks even better. If the preload is lower (compared to the original analysis), then the maximum load on the upper belt run is proportionally higher. For a left-side design, the bearing load is proportional to the captain’s power output once the preload is exceeded, so this results in higher bearing loads much of the time. For the right-side drive train, the bearing load is always close to the static preload, so the average bearing load stays much lower, by comparison.
Dave wrote:
… and that sounds more reasonable. I believe Gates uses a benchmark of something like a 10:1 ratio for preload based on the maximum torque the system is designed to support, which is where the sprocket size / tooth count comes into play. After that it’s the properties and dimensions of the belt that get adjusted to deal with the design loads. That why the 33t sprockets I’ve been playing with require a much higher preload (17-19 lbs to achieve .5″ deflection) vs. the 69t sprockets on the standard Gates tandem sync drive (5-10lbs to achieve .5″ deflection).
No doubt and I don’t take issue with a same-side-drive being a more efficient technical solution for linking together multiple cranks driving a rear wheel. In fact, if you run a same-side drive out-of-phase then the loads on the rear bottom bracket should — in theory — become very close to what single bikes see.
My most significant issue pertained to these two comments on your V2r webpage:
For the first one (a), I dont’ take issue with the added abuse that crossover cranksets dole-out to bottom brackets. As mentioned before, I just never experienced or observed a lot of BB failures on tandems until some of the new BBs came on the market and/or where there was clearly an exacerbating influence, e.g., low-torque/no Loctite on MegaExo cranks, Santana’s 129mm wide rear spindles, someone running really small timing rings… something even smaller than the 34t daVinci rings we’ve used since 1998 on five different tandems, road and off-road. In fact, it will now be interesting to see if folks who have adopted the 11x36t cassettes on 3×10 drive trains begin to see a higher failure rate on Bottom Brackets with yet another bump-up in the loads in something like a 30tx36t granny gear on Clydesdale teams climbing steep grades or perhaps more frequent chain breakage when shifting between the 32t and 36t rear sprockets under load. I know that early-on some of the bigger off-road teams were folding over the XTR 34t sprockets and resorted to the more robust LX cassettes. These new XT grade wide-range cassettes from Shimano look pretty robust — as are the PowerDomes & other SRAM wide range cassettes — so the 10 speed chains may be the weaker link at this point.
The second one (b), that was the crux of the issue that we’ve been chewing on for the past week: does a belt drive truly put much higher and/or detrimental loads on the bottom brackets vs. a chain? That was my assumption back in 2006 when I saw Bob Thompson’s first sync belt drive. But, after watching it in use along side chain drives, I came to believe that any difference was probably pretty minimal. Again, there are other things such as using small diameter / low-tooth count timing rings or really wide bottom bracket spindles that would be far more abusive to bottom brackets on a belt or a chain drive.
I think I’m still of that same mindset; perhaps not. Again, it would be interesting to know if the early adopters riding Co-Motion tandems with the belt are having a higher bottom bracket failure rate vs. folks who used the same cranks & bottom brackets with chain-based sync drives. That would be some very useful data, to be sure. As I said, most of the bottom bracket issues I’ve been seeing of late seem to be caused by low torque or failure to use Loctite on the MegaExo BB’s which pre-dates the belt drives.
Hey, great dialog…. I appreciate your time and interest in sharing your thoughts and it’s been beneficial.
Hi, Mark,
Enjoyed your comments, including the other blog about the 11 mm pitch (i.e., not “tandem”) timing belt on your Calfee. I’ll offer this as a perspective. I, too, find that occasionally my head hurts, but what I find is that those periods of (hopefully) brief but intense effort pay off in the long run. I find, in retrospect, it’s usually worth it, especially when you can look back on it and say, “Yeah, I’m glad I put the effort into it then, ‘cuz now everything’s working well.”
I’m struggling with the whole belt preload specification, for several reasons. The concept of “tension ratio,” defined as (high side tension/low side tension) loses its meaning if the preload is only ~75 lbs., as Gates recommends. The maximum tension on the top run due to the captain pedaling at a power output of 1000 Watts (high level but that’s just the point–you have to specify some kind of a maximum) is 272 lbs., which is way more than the preload: you’re vastly exceeding the guideline of never letting the lower run drop below zero tension. The tension ratio is INFINITE once the lower run tension drops to zero. Cycling is actually a very interesting anomalous application of belts compared to things like electric motors, as the maximum torque is huge compared to the average load, and it’s difficult even to predict what the maximum torque load might be. As an extreme example, an Olympic-class sprinter might put out close to 2 kW in a sprint, which is twice as high as the example I’ve been using above.
What this means is that the Gates Carbon Drive belt must be capable of operating even with very high tooth-face pressure in high-effort situations without skipping or ripping the teeth off the belt. It’s an impressive design, compared to older drive belts. Credit the carbon fiber reinforcement for that, which limits the stretch in the belt under load, which might otherwise cause the belt to skip on the sprocket.
So, now I’m wondering what exactly the preload specification really means? The traditional concept of tension ratio no longer applies. It seems to me the preload is a trade off between bearing load and life time (lower being better) and drive train efficiency (the preload should correspond to something in the range of average power output so that the belt efficiency is highest under normal riding conditions). An unanswered question is how much does the belt efficiency change during high-effort events compared to the average? I doubt that even Gates knows the answer to that one.
Hey Dave:
It will be interesting to sit back and watch how these new products stand the test of time. Thankfully, there haven’t been any real white elephants or seriously flawed products introduced for tandems in a long time, so I suspect all of these innovations and adaptations will, in fact, serve owners well if they are savvy enough to select equipment that truly meets their needs and follow the manufacturers recommendations for care and feeding.
Thanks again for your time and insights.
V/R, Mark
Rather than giving (yet) another dry engineering analysis, I’d like to offer another perspective as a tandem user. I’ll admit that the bearing load issue may be secondary–not unimportant, but secondary. But, having now ridden both tandems, the V2 and the V2r, side-by-side over several weeks, both my stoker-wife and I concur that the V2r is noticeably stiffer, particularly in standing sprints and climbing. The two bikes are otherwise comparable in specification; our V2 being only one pound (450 grams, approximately) heavier than the V2r, so I believe it’s more than just the weight difference.
I measured the distance, laterally, from the frame center line to the inner edge of the belt on the two bikes. The V2 is about 44 mm, and we’ve got it set up for minimal clearance on the chain stay; maybe 3 mm is all. I assumed all along, even several years ago when we built it, that this was an important parameter. On the V2r, the same dimension is just 34 mm. All else being equal (which it is, in this case), that means the bending moment on the boom tube is reduced by more than 20%. In all-out efforts, it’s noticeably stiffer. But, that’s just our impression. I’d invite the considered opinion of others; yourselves included, obviously, but if anyone else is visiting Boulder, please feel free to contact us and give it a go.
This might be a more important consideration than the bearing issue, assuming you’re starting with quality components and a not-too-wide axle, as you discussed. I don’t have any personal experience with 129 mm axles (e.g.), so I’m perhaps ignorant on that issue. I can’t see how that’s an improvement, though, given that wheel options today allow for more than adequate strength and stiffness even with 130 mm rear axle spacing. I’m sure I’ll get a lot of comments on that one, but I’ve been studying wheel designs for 20+ years and can say with confidence that well-built 130 mm wheels can be every bit as strong as most 145 mm wheels. 160 mm? That’s simply unnecessary, if you design and spec the frame and other components appropriately. And, as you point out, it can lead to problems in other areas, such as the bottom brackets. Everything is a trade off, and I think we’ve found a pretty darn good sweet spot.
I’m sure it feels very solid… I would have been disappointed it if didn’t.
Pingback: Timing Belt: Update #5 « The TandemGeek's Blog
Pingback: 2×10 Drive Trains for Tandems… My Thoughts. « The TandemGeek's Blog
My brother suggested I might like this blog. He was once entirely right. This put up actually made my day. You can not believe simply how much time I had spent for this information! Thanks!
Pingback: Other Tandems @ NAHBS… « The TandemGeek's Blog