New cannondale SYSTEMSIX road frame!

[CallumRD1]

On the non- Himod models, with mechanical, anyone know if the two cable/wires we see externally is the brake hoses or the gear cables? I wondering if I should strip the Duraace mechanical and build it up with Etap.

Cannondale has always ran the rear brake cable/hose through the top tube and any gear cables through/under the downtube so it’s PROBABLY gear cables

This makes little sense with flat mount disc calipers as the hose needs to come out of the chainstay. Every flat mount frame that I've ever seen thus routes the rear hose through the down tube and into the left chainstay.

[DamonRinard]

Hi fosen,

The brake hoses are always internal (entering at the top of the head tube just below the stem). Di2 wire also goes inside there.

Mechanical shift housings can also fit inside the frame there, but the bends lead to poor shifting, so we recommend the routing you see in the catalog photos: mechanical housing entering the frame at the switchplate in the down tube.

The bikes include a blanking plate to neatly close the switchplate in the down tube if you want to install eTap.

Cheers,
Damon

[CAAD8FRED]

Hi fosen,

The brake hoses are always internal (entering at the top of the head tube just below the stem). Di2 wire also goes inside there.

Mechanical shift housings can also fit inside the frame there, but the bends lead to poor shifting, so we recommend the routing you see in the catalog photos: mechanical housing entering the frame at the switchplate in the down tube.

The bikes include a blanking plate to neatly close the switchplate in the down tube if you want to install eTap.

Cheers,
Damon

So that’s why running cables with a Metron 5D is hard

[DamonRinard]

So that’s why running cables with a Metron 5D is hard

Hi CAAD8FRED,

Internal routing is always fussier than external, but give the SystemSix routing a try. It's a lot less fussy than most superbikes these days.

EF-Drapac p/b Cannondale pro mechanics often assemble the Vision Metron 5D handlebar on the SystemSix, it's a good match both from an ease-of-routing perspective and an aero-performance perspective.

Cheers,
Damon

[themidge]

I'd just like to say, the EF Drapac Pres...blah riders look like they're riding spaceships at the TDF, even when Lawson Craddock was going into the pave sectors at 25kph at the back of the bunch. Hopefully the MAMILs' spacer towers won't kill the look too much

[romanmoser]

They were riding the synapse disc today from what I saw ...
Maybe some on the supersix rim brakes but I doubt

[themidge]

Oops Well the synapse looks like a spaceship too

[Calnago]

Damon,
Thanks for all your answers.
Looking at the geometry I see you keep the 405mm chainstay length despite the wider rear hub. Isn’t Shimano recommending something longer (415mm?)

...
Yes, but we like the handling with 405mm chainstays better (it is a racing bike after all), and you only give up (maybe) one gear in the small chainring.
...
Cheers,
Damon

Hi Damon, the "one size fits all" approach is something I've never been able to appreciate. We've already seen how the S/M/L approach was far too limiting when Giant first introduced their "compact" frame geometry and tried to pawn it off on the public as being overall smaller/lighter/stiffer etc., when in reality it was a means of reducing production costs through reducing the number of different size frames that had to be made.

But let's just for a minute look at the chainstays which, when taken in the context of a complete bicycle play such an important part in determining the overall handling characteristics, that to say one length is good enough for the smallest of frames and the largest of frames is really like saying "Hey, as long as your pants cover your knees, they're the right length".

Small bikes have always been the most difficult challenge from a geometry perspective for frame builders in the context of 700c wheels. And for these bikes, the shortest chainstays possible are probably the way to go, short of also going to 650 wheels etc., and designing the frame around them. So, short chainstays for the tiny bikes are a good thing handling wise, if not so great for the drive trains. As was pointed out, minimum chainstay lengths (particularly for disc braked applications) have been specified by the likes of Shimano and Campagnolo, typically 410mm in Shimano's case, and 410mm for Campy disc (405 still for rim brakes and 130mm dropout spacing). But let's ignore drive train issues for purposes of my question and only focus on handling issues.

I ride a larger frame, and for reference in manufacturers terms, that would ideally be a 60cm Trek, a 58 Specialized, and any Colnago traditional frame from 59-61. And in that size of frame, the minimum chainstay specs do not really come into play as any frame that I've really liked the handling of in that size has chainstays in the 410-415mm range. I've found that frames that size with super short chainstays have a more "nervous" feel to their handling (steering geomtery aside) simply because it places relatively more weight towards the rear, and that increases with both shorter chainstays and higher saddle heights. With seattube angles typically in the 72-73 degree range for larger frames, it just makes intuitive sense to me that as the taller riders seated weight moves further back over the rear wheel, a bit of compensation for that in the form of longer chainstays is appropriate to maintain a constant feel over the size range. Specialized, for example, is all about touting their "rider first" approach to frame specs these days, meaning a small rider should get the same feel on the bike as a taller rider. If that's even possible is another question, but I do like the concept at least. But regardless, both Trek and Specialized have chosen to go with a minimum chainstay length of 410mm to appease Shimano. Specialized has chosen the same approach as Cannondale with respect to a "one size fits all" approach and have settled on 410mm as their one size. Trek on the other hand, while they have chosen a minimum chainstay length of 410mm to meet Shimano's specs, they also acknowledge that larger frames should be made proportionately larger in chainstay length as well. So, while their smallest frames used to have shorter chainstays, they now all have a minimum length of 410mm but start getting longer once a certain size frame is reached (58, 60, 62). And Colnago has always had very incremental differences all the way through their size ranges, and still do, incorporating no less than five different chainstay lengths across their size range.
I have ridden larger frames with short chainstays (in the 405mm range) and personally I do not like the handling in comparison with larger frames having chainstays in the 410-415mm range. For purposes of this discussion I'm using a range of 405-415mm for chainstay length becasue that's the range the vast majority of road race frames would fall into, barring longer chainstays to provide clearance for huge tires, etc.

So, I guess just answering "We like the handling of short chainstays" in response to the question posed seems a bit lacking for a couple of reasons... 1) You are ignoring the drive train manufacturer's specs altogether, and yes, while I don't care if you can't use a couple of the smallest cogs while in the small ring, some drivetrains can be extremely sensitve to front derailleur setup, in particular to the angle of the chain from big to big when crossed, and that is very much a combo that many people would use, myself included. And 2) It's basically saying that you don't really care where the riders weight is situated in realtion to the rear wheel. I think you understand it, but maybe just decided it's not something you're going to worry about, as who is really going to notice that, anyway. I think I notice how the wieght distribution feels between the wheels very quickly.

[bikesrdangerousmmk]

[DamonRinard]

Hi Calnago,

Bit of background:

Way back in the day, I helped make the Trek geometry with increasing chainstay lengths.
After that, I helped study CS length changes at Cervelo.

Based on bikes you mentioned, I just looked up Colnago's increasing chainstay lengths, and I believe you when you say Specialized uses the same length.

What's going on? There are several factors at play.

1) No one can tell the difference. (For one thing, riding different production bikes, there is always something else also different. More about non-production bikes later.)
2) "Everyone" likes short chainstays. The main reason they get longer on bigger sizes is the seat tube angle gets slacker. Traditionally the tire would hit the back of the seat tube if you didn't make the chainstays longer. This applied to the aluminum Treks and was held consistent on the carbon models (H1, H2, etc.) and looks reasonable on the lugged Colnagos (ST leans back, tire has to get out of the way), but isn't needed on the molded bikes (Cannondale SystemSix, Specialized) which have the luxury of any chainstay length we like - and we ("everyone") like them as short as possible. Proof: the Colnago CS cannot get shorter, even on the big sizes, or the tire would hit the seat tube. That means, like us, Colnago uses CS as short as possible (for their chosen frame assembly method).
3) Other dimensions are far, far more important to handling than CS length.



When I worked at Cervelo we had to lengthen chainstays on the TT bikes, which were centimeters (not millimeters) shorter than Shimano's recommendations (380mm from memory!). We were losing too many gear combos to keep selling more bikes - we just couldn't explain to enough people the benefits in order to accept the gear limitations. But the proposed change was a big struggle internally, with even Phil White demanding a science experiment before allowing a change.

So we made prototypes with different chainstay lengths, even adjustable bikes. Lots rode them, some blind testing with changes (or no change!) and the rider not knowing which length they were riding. Even Roger Hammond (Cervelo TestTeam), bike handler extraordinaire, got in on some geo testing in those days.

Result: no one could tell. And it makes sense: 2-3mm CS length difference out of a wheelbase of ~1000 is less than half a percent: nowhere near the JND, just noticable difference level humans can detect.

Same at Trek before that: when we did protos of the bikes with different CS lengths, even the prototype weld shop couldn't hold the tolerances - CS length varied within a size more than between sizes. (This is less of an issue with molded frames, but the bond joint(s) still allow some tolerance).

Basically, it's just not worth the effort for a few millimeters between sizes, when no one can tell.

Okay, back to the Cannondale SystemSix.

For small frames: instead of a few mm in CS length, to address geo we've cleaned up our stack & reach, and yes made our smaller bikes fit and handle better. Steering geo also includes a second fork offset, something that really *is* important to handling. Saving a few molds (where before, we had S&R very very close in a couple of sizes, so not really that valuable for fitting) means we've been able afford a second fork offset and size-specific layups, both of which make a bigger difference - but more on that below.

For larger frames: instead of a few mm longer CS, we, like Specialized, have focused on both steering geometry, together with frame stiffness. As a taller rider too, no doubt you (like I) have ridden frames that are too flexible; the SystemSix has size-specific stiffness, with bigger frames appropriately stiffer (but not overkill), and smaller frames less stiff (but not too flexy). This is a real engineering feat, not easy to accomplish at all, since as we all know a smaller structure is inherently stiffer (all else equal), yet with great effort we've accomplished exactly the opposite. Check out the white paper to see the measured data.

For the future: We may again find that the market prefers to use all their gears and change to 410, but for now we have possibly the shortest (best handling) bikes with 405 CS.

Cheers,
Damon

[TobinHatesYou]

Very informative, thank you DamonRinard.

Perhaps a few extra millimeters would make a difference on touring bike where the rider is more upright and the weight distribution is different, but with “race” bikes meant for aggressive body positions, I’m with you, it really doesn’t matter unless the cyclist is insanely tall...7’ NBA stars who buy Dirty Sixers.

[Calnago]

Hi Calnago,

Bit of background:

Way back in the day, I helped make the Trek geometry with increasing chainstay lengths.
After that, I helped study CS length changes at Cervelo.

Based on bikes you mentioned, I just looked up Colnago's increasing chainstay lengths, and I believe you when you say Specialized uses the same length.

What's going on? There are several factors at play.

1) No one can tell the difference. (For one thing, riding different production bikes, there is always something else also different. More about non-production bikes later.)
2) "Everyone" likes short chainstays. The main reason they get longer on bigger sizes is the seat tube angle gets slacker. Traditionally the tire would hit the back of the seat tube if you didn't make the chainstays longer. This applied to the aluminum Treks and was held consistent on the carbon models (H1, H2, etc.) and looks reasonable on the lugged Colnagos (ST leans back, tire has to get out of the way), but isn't needed on the molded bikes (Cannondale SystemSix, Specialized) which have the luxury of any chainstay length we like - and we ("everyone") like them as short as possible. Proof: the Colnago CS cannot get shorter, even on the big sizes, or the tire would hit the seat tube. That means, like us, Colnago uses CS as short as possible (for their chosen frame assembly method).
3) Other dimensions are far, far more important to handling than CS length.



When I worked at Cervelo we had to lengthen chainstays on the TT bikes, which were centimeters (not millimeters) shorter than Shimano's recommendations (380mm from memory!). We were losing too many gear combos to keep selling more bikes - we just couldn't explain to enough people the benefits in order to accept the gear limitations. But the proposed change was a big struggle internally, with even Phil White demanding a science experiment before allowing a change.

So we made prototypes with different chainstay lengths, even adjustable bikes. Lots rode them, some blind testing with changes (or no change!) and the rider not knowing which length they were riding. Even Roger Hammond (Cervelo TestTeam), bike handler extraordinaire, got in on some geo testing in those days.

Result: no one could tell. And it makes sense: 2-3mm CS length difference out of a wheelbase of ~1000 is less than half a percent: nowhere near the JND, just noticable difference level humans can detect.

Same at Trek before that: when we did protos of the bikes with different CS lengths, even the prototype weld shop couldn't hold the tolerances - CS length varied within a size more than between sizes. (This is less of an issue with molded frames, but the bond joint(s) still allow some tolerance).

Basically, it's just not worth the effort for a few millimeters between sizes, when no one can tell.

Okay, back to the Cannondale SystemSix.

For small frames: instead of a few mm in CS length, to address geo we've cleaned up our stack & reach, and yes made our smaller bikes fit and handle better. Steering geo also includes a second fork offset, something that really *is* important to handling. Saving a few molds (where before, we had S&R very very close in a couple of sizes, so not really that valuable for fitting) means we've been able afford a second fork offset and size-specific layups, both of which make a bigger difference - but more on that below.

For larger frames: instead of a few mm longer CS, we, like Specialized, have focused on both steering geometry, together with frame stiffness. As a taller rider too, no doubt you (like I) have ridden frames that are too flexible; the SystemSix has size-specific stiffness, with bigger frames appropriately stiffer (but not overkill), and smaller frames less stiff (but not too flexy). This is a real engineering feat, not easy to accomplish at all, since as we all know a smaller structure is inherently stiffer (all else equal), yet with great effort we've accomplished exactly the opposite. Check out the white paper to see the measured data.

For the future: We may again find that the market prefers to use all their gears and change to 410, but for now we have possibly the shortest (best handling) bikes with 405 CS.

Cheers,
Damon

Thanks for the response Damon,
I'll just say a couple of things then move on, as I think you've answered what I wanted to know..
Firstly, while I agree in general that "No one can tell the difference", that is a very general statement and most people don't even have a most basic understanding of how different geometry decisions affect a bike's handling, let alone are able to tell which of those factors is making a bike handle the way it does. It's complex, and everything interacts with each other, but to say "No one" can tell is simplifying things a bit too much for my liking. There are people, granted probably not your average cyclist or magazine reviewer or even pro rider, with enough experience and sound understanding of geometry, that when faced with two very different handling bikes, I would bet could come in and sit down after spending time on the bikes and when given basic questions like "Which has a steeper headtube angle?"..., or Which has a longer wheelbase?..., or Which one has longer chainstays?... would in fact more often than not get the answers to those questions correct. I suppose I also wonder if testers can tell the difference between a bike that is 2-3 watts faster than another. Or if they could tell which bike had a 3mm different fork offset, and if so, which one had the lower fork offset. Point is, and this we certainly agree on, is that the bike and how it handles is ultimately the sum of everything, not the least of which is how the rider sits on it and rides the thing. An awful lot of individual preference comes into it, which more often than not gets clouded by the marketing spin.

Secondly, does "Everyone" really prefer shorter chainstays? Who's everyone? I'm someone... and I do not prefer 405mm chainstays on my size frames. Short people on the other hand probably prefer short chainstays. Also, the chainstays on my Colnagos, even the 59T which is the smallest I have, could most certainly be shorter (even with the 25mm tires I have on) by a good 5-8mm if they so chose to do so), without altering the seat tube angle. And yes, I get how the seat tube angle plays into the chainstay length and that was often the most limiting factor to chainstay length in days of old when a 22mm tire was considered big by some. But those days are gone, and the limiting factor now, I would venture to say, is the massive tires that some people want to throw on a road bike, creating the need for more clearance up front.

You mention your days at Cervelo, and yes, they were one of the companies who I can remember first started the one length fits all approach. But in their tt bikes etc., (I have a 58cm 2010 P3 by the way), I could turn a blind eye to that simply because it was a very different geometry than a road bike, and the weighted position is so far forward and the front center so long that yes, short chainstays even on the larger frames were just fine, perhaps preferred (all, or at least most, of the Cervelos back then had 399mm chainstays). So, ok for a tt bike geometry, but imo terrible for a road bike with more standard geometry. No doubt Vroomen had a lot of control over what ultimately got produced and what didn't. And his influence over at 3T is imo painfully apparent as well. Just ask the AquaBlue guys how much they like the bike. It's not just the one-by drive train, but look at the clearance (or lack thereof) those bikes have for the rear tire. And again... their entire size range... 405mm chainstays. But with 1x drivetrains, those short chainstays are surely not doing them any favors either.

You then go on to talk about other factors being more important than chainstay lengths. I would propose that all of the factors together make up the final bike, and to completely deemphasize one or more of them because you can get away with fewer molds in the process is great for the bottom line, but maybe not so great for the end user, whether or not they can tell the difference (again, some might, some won't). You say that fork offset is so much more important that chainstay length. But is it? By how much? What about wheelbase? Which affects wheelbase more? And the thing is, they all work together. Trek was a very interesting case for me when i got my Koppenberg, mostly because other than the front end, most of the geometry was etremely close to my colnagos, but the front end geometry was very different. Trek is really the only significant manufacturer I'm aware of that uses a fork offset of only 40mm on their larger road bikes, and they couple that with a relatively steep head tube angle to come up with the trail/wheelbase they're after. And both my Colnago of the same size and the Koppenberg handle fabulously well, but can I tell the difference up front?... you bet I can. And that was a big reason I wanted to try out the Trek geometry in the first place, almost an experiment if you will. And conversely, bikes my size with a very short chainstay tend to feel a little bit more nervous handling even with similar front end geos. Why is that... my best description of how it feels is (and you have to be able to extrapolate here a bit), is that the shorter chainstay bike coupled with a tall rider set back will have, relatively, a "pivoting" effect while steering, in comparison with a bike where the rider weight is just a bit more forward between the wheels, as would be the case with longer chainstays. To take the weight distribution thing to the ultimate extreme, imagine a unicycle effect of weight distribution but applied to a road bike... as the seat gets pushed farther and farther back and up on the road bike, with no corresponding increase in chainstay length, you get closer and closer to the weight distribution of a unicycle. Of course, that example is merely for illustration purposes but I think it's accruate as to what's happening from a "feel" perspective.

So, that's all I wanted to say, but appreciate hearing the design philosophies of different manufacturers. You're all obviously quite different in your approaches. And obviously there is no one right answer. So... may the best marketing win.

Cal

[wingguy]

1) No one can tell the difference....
no one could tell....
no one can tell....

We may... change to 410, but for now we have possibly the shortest (best handling) bikes with 405 CS.

I'm confused

Thank you for your post, it's extremely informative - but I can't follow the logic. If 2-3mm is "nowhere near" the humanly noticeable difference in CS length, why are 405mm chainstays better than 410?

[fosen]

Hi fosen,

The brake hoses are always internal (entering at the top of the head tube just below the stem). Di2 wire also goes inside there.

Mechanical shift housings can also fit inside the frame there, but the bends lead to poor shifting, so we recommend the routing you see in the catalog photos: mechanical housing entering the frame at the switchplate in the down tube.

The bikes include a blanking plate to neatly close the switchplate in the down tube if you want to install eTap.

Cheers,
Damon

Awesome, thanks. So it is also possible to remove the DuraAce mechanical, and run for example Ultegra Di2 with internal cables, on the same non-hi Mod model? With the junction box in the frame on the downtube? Or maybe in the handlebar?

[fa63]

Thank you for your post, it's extremely informative - but I can't follow the logic. If 2-3mm is "nowhere near" the humanly noticeable difference in CS length, why are 405mm chainstays better than 410?

Same here; seemed strange to argue that a few mm make no difference, then to say 405mm was chosen instead of 410mm etc...

That said, the "snappiest" bike I remember riding was a Giant TCR with 405mm chainstays, so maybe there is something to it