chojn1

What about just replacing the TA 48t with another 48t? That will give you an immediate answer to your TA hypothesis.
CJ

twocicle

Yes, see above, "I now have sets of other, better profiled rings on hand to test.".

The Stronglight rings arrived by post from France (xxcycle.com) incredibly quickly. The 48T CT2 has only 2 pair of pins (as does the TA Alize), but from my observations of shifting with the Di2 FD the pins were unnecessary as most of the time shifts performed well at mid-points and not due to the pins on the TA. In fact, because the TA lacked any profiling at the pins, experimenting with the chain showed it was very difficult to ever engage the pins. The Stronglight CT2 has a smooth transition instead of the sharp lip and ramps at the pins do allow the chain to actually engage these. Also the CT2 teeth have the diagonal profiling I was looking for. Interestingly, the 50T CT2 has 4 pair of pins - double that of the 48T.

I'll get to testing these today. After getting back from a long ride. May 1st, hi of 78F. Yes!!!

fietsbob

suppose the carbon crank choice eliminated the crossover drive option, as there are no left side crank arms /spiders .

twocicle

Correct. That was one of the "pluses" in my book. It also provided as small a Q-factor as possible (can't go further due to heel clearance to chainstays being at the minumum), which was a primary decision on the crankset (DaVinci w/crossover is the only other tandem crankset w/crossover that would meet our Q-factor needs, and allow standard 130/74bcd chainrings). Both Lightening and FSA tandem cranksets are too wide for my stoker.

The FSA double (front) and triple (rear) cranksets also line up well with each other, and provide ample crankarm clearance for the SSD setup. I have to tip my hat to Ric @ House of Tandems for the sample photos on his website plus some direct answers to questions I had regarding his work in this area. Ric has been setting up 52/34T rings and SSD configs for customers in his area.

Update on the ring setup:

I got to testing the new chainring setup before heading out on my single bike ride. It seems to be working well with the combination of a Stronglight CT2 48T and a TA Zelito 30T granny. I tried using a Stronglight 30T granny, but the tooth position appears to be setup for a 16T ring difference, as a 32T Stronglight aligned properly with the shift pins and worked well. Luckily, the TA Zelito 30T teeth are rotated nearly a 1/2 tooth over and align the chain links perfectly with the shift pins on the 48T Stronglight.

From the referenced article "Chainring Choice or Shifting Ramps?", here is a photo showing how chainlinks should align with the shift pins...



They usually provide 2 pins close together, because the spacing between them will guaranty alternately hitting either the inner or outer chain links.

TBD... actual road testing.

Chris_W

I now have my single bike set up with 30-46 chainrings and a 11-32 11-speed Shimano cassette, and Ultegra 6870 Di2 derailleurs. Every front shift has gone perfectly so far. Chainrings are a stock Shimano 30 tooth inner ring and a TA 46-tooth middle ring (the version designed for the middle of a triple setup), mounted on a Shimano 5703 triple crank. I have Race Face tabs to blank off the outer ring mounting tabs, which simply makes the crank look more slick.

The FD mount is a braze-on, so I can't get the FD as low as it should be, there is about 5 mm between the 46-tooth ring and the bottom of the outer plate of the FD. If you haven't already tried it, I would try some unusual positions for the FD - too high, not totally parallel, etc.; occasionally I've had success solving FD issues by using some non-standard positioning of the FD after all regular methods have failed.

waynesulak

Chain line questions. You mentioned having 145 rear spacing.

1- What rear hub are you using, including year of manufacture or model ( some manufacturers change flange spacing over time)?

2- Where does the center of the cassette line up on the rings?

comment on experimenting :

Lots of fun to be had trying new things but not for those that just want it to work with no adventure. Same side drive is old tech and will work easily with a chain. Belt drive tech is proven and has begun to be understood in the bike world. Using together is experimenting and adventure which means the possibility of failures rides delayed, changed or terminated. It is nice to have a stoker that understands the fun of trying new things when problems occur on the ride.

Q-factor comment: chain appears to allow for different ring alignment between captain and stoker not available on belt drives. Example: We use DaVinci cranks and are trying 113mm captain BB and 106 stoker BB. Sync chain has zero problems and looks like it would have no problems with a 126mm and 106mm BBs.

chojn1

This is a chain line diagram from Sheldon Brown which I think is from Stronglight:



To place the inner two rings of the triple in the position of the double for which the DI2 is designed, you have to laterally displace the right spider/crank by ~2mm. Additionally to compensate for the 145mm rear dropout spacing versus the 130mm for which the DI2 is designed, you would have to technically displace the right crank an additional 4mm.

Also, the DI2 front derailleur over shifts slightly before settling back ~ 1mm toward midline.

My guess is that your chain drop is due to these two factors rather than the TA ring's ramp and pin which could affect shifting smoothness but not chain drop. Changing the chain ring to another 48t will give you no improvement in this case.

Alternatively, if there is not enough spacing between the middle (now large) chain ring and the belt pulley, the over shift of the DI2 could cause the drive chain to catch the timing belt. Running your setup without a belt will give you an answer here.

Your empiric testing should give us an answer soon.
Looking forward to your results.

CJ

Bad1

We'll stated chojn1,with two cycle set up a proper chain line will never be achieved, DI2 with a 10 speed cassette should yield a 20 gear selections in shifting combinations. As I see it the short comings of this set up is your limited to a range of gear combinations but not the whole enchilada!

twocicle

Hi CJ,

Agree, that is a good, basic reference from Sheldon. Providing the link for others to follow here.

Earlier in this thread I mentioned placing the FSA wave washer on the right side to improve chainline. This puts the triple cranks very near optimal, but the big ring is still slightly biased (a mm or so) toward center which is what I wanted since we use that ring most of the time. It is not a fatal problem to shift the chainline by this very small amount, it just depends on what ring/cog combos you intend to use and not use. Regarding timing ring alignment, the spider positions of both the double and triple are well matched and confirmed once the belt is installed because if the eccentric is moved 1-2mm right or left the belt noise lets you know when the alignment is off. I also used a digital caliper to measure

I have the FD outer limit screw set to just allow clearance of chain rub when in big ring -> smallest/outmost cog, and with this limiter set there cannot be any gross overshifting by the FD to cause misshipping the chain. The 1mm overshift to which you mentioned is equal to the FD position when using the outermost cogs. On the inside, the limit screw is set there too. The FD range can be tuned to work between these two limit screws and will adapt to work with chainlines within that range.

Anyway, misshipping the chain is not the issue I was seeing, but where the chain would shift between pins (where not pins or tooth profiling existed) it would get caught up on the ring teeth with this new chain. With other rings combinations and mfr, I am seeing different behaviors, I just need to settle on the gear selection now.

twocicle

CJ, here's an update especially for you

I've experimented with every ring on hand. Best shifting and gear combo for us turns out to be the Stronglight 48T/32T with a 11-32 XT cassette. I wanted one more lower gear than what we previously had (48/30 w/11-28), to coax and reassure my stoker to start doing some longer climbs during our rides. With the 32T granny, that eliminated the need for any spacers and moved that ring closer to the 48T mounted on the middle position. Without the spacers the chain starts to chime on the big ring pins when in the 8th rear cog, but I do not intend to use cogs 8-10 from the granny anyway. Aligned as it is, the big ring works very well with cogs 2-10. So, this keeps the typical 16 out of 20 cogs to use without excessive crosschaining.

Info about other rings:
Due to the chainline I have setup which is based on the inner two chainring mounts on a standard triple and the 145mm rear spacing, neither a 50T or 52T work well enough to use. The 50T maybe just ok, but the FSA 52T is a definite no - the more severe chainline and tooth profiling on that ring is such that when the chain is in the smallest rear cog it wants to climb to the outside on the chainring. I'm 100% certain either of these rings would work on a wider spaced tandem crankset, or with a smaller rear frame spacing, with either mechanical or Di2 shifting.

This reinforces my future preference for a standard 135mm or a 142mm X12 rear spacing, as we need "skinny" stoker cranks and I absolutely will keep using electronic shifting. Perhaps if/when the eshifting is available with a triple FD, I might consider finding a skinny crossover setup (ie: DaVinci) at that time. At the time we had this frame built I had no intention of ever going Di2, but here we are now.

FWIW, both sets of FSA SL-K Light cranksets with BBs were a total of $550. That is close to half the cost of their equivalent (but wider) tandem set and could possibly recover the full cost if I ever decide to swap out to different cranks.

chojn1

Twocicle,

Thanks for the update. I was hoping you could make the 30/48 crankset work, as I really like this gearing combination. So far I have not had any problem with this combination, but again, there are no hills around here.

CJ

twocicle

I didn't say the 30t granny would not work, just that I selected the 48/32 with a 11-32 cassette to get just one gear lower. I suppose the 30t granny could have been reinstalled after my last round of pulling rings & cranks... and maybe will do that in the future after we get a better feel for this current gear combo and more hill climbing. FWIW, the Di2 FD is shifting to the 48t from the 32t (16t gap) with no apparent difference than the 30t (18t gap), though it does appear that with the Stronglight rings, the shift pins and chain line up better with the 16t gap. Even so, the FD can (and does) frequently complete upshifts at mid-points between the pins which suggests perhaps that the pins are not needed at all.

Whichever rings you do select, my gut tells me results will vary depending on the precise setup. For example, I never had a single issue until swapping to a new CN-6600 chain (and after all the pains of retesting multiple setups, I never tried a different chain even though I have a bunch of them in boxes!).

twocicle

Update on the 10spd Di2 6770 FD shifting problem (over-shifting to the outside) that I had experienced,

Upon reconfiguring our drive components now with 135mm rear spacing, I narrowed down the FD problem to the fact that we are using the inner 2 chainring positions of a standard FSA road triple crankset.

With the standard BB and triple road crankset installation, the big ring in the middle ring position, is sitting only ~45mm from centerline. The Di2 FD is designed to work with a road double, of which I measured 50mm to centerline from the big ring on my single, so I assume that is the expected distance from center. This means the standard FSA triple middle ring position is some 5mm further inside than expected. As I have the Di2 6770 FD limit screws set to the innermost adjustment, the only recourse I have is to space the drive crank further outboard.

This issue would not effect people using the FSA tandem triple crankset, because that is spaced much wider than the single crankset (which we are using due to stoker Q-factor issues, ie: need narrow as possible).

According to Todd at DaVinci, his triple crank spider (130/74 BCD) combined with the smallest axle (103mm) would result in a 46.5mm middle ring position. His double ring spiders, while achieving the correct distances (48.8mm centerline), are not 130/74BCD, which is what I want to continue using (running 48t/30t rings).

At this time, I do not know if the 11spd Di2 6870 FD allows for a more inward cage position. If it does, then our moving to the 11spd derailleurs and 11-32 cassette would be a next step.

--
Followup: Shimano support says the 6870 FD probably does not have more inward adjustment range.

twocicle

Newly spaced 135mm rear wheel info posted in this thread: https://www.bikeforums.net/tandem-cyc...l#post16899388

After installing the new Saint RT99 ICE rotor and observing the TRP Spyre pad markings on that, it became apparent the 203mm Avid adapter I was using needed some 2-3mm of spacers to lift it further outboard (it was sitting too far to the middle of the rotor). After repositioning the caliper, the TRP braking is feeling very nice indeed. I went back to TRP to order their taller 203mm adapter and found it only available from the Teckro (parent) website.

Chris_W

I'm not so sure that it is a chainline issue. As I've said before, I'm running Ultegra Di2 (6870) with 46-30 chainrings mounted on the inner and middle positions of a Shimano 105 triple crankset (which should have the same 45mm chainline as the FSA triple cranks), but on a single bike (with 130mm rear spacing), not on a tandem.

After 3,000 km, the setup hasn't missed a front shift once. I did have to be careful about the angle of the FD so that the chain didn't rub on the inner plate when in the lowest gear (small ring-big cog) - I angled it a TINY bit inwards from being parallel, but it's BARELY noticeable even when looking closely. I've never had any issue with the outer position or overshifting to the outside.

The chainrings are a TA Specialites 46 tooth middle ring, and the stock Shimano 30 tooth inner ring. I've been entirely happy with this setup and would recommend it to everyone (especially if you don't mind coasting when above 50 kph / 30 mph); I had been using it with an Ultegra 11-speed 11-32 cassette, but I've just combined that with the first four cogs from a 12-25 11-speed cassette to give me 12-32, so I've basically switched the 11-tooth cog for a 15-tooth, which is more commonly useful for the kind of riding that I like to do on my single bike (long endurance rides of 200-300 km in the hills and mountains of Switzerland and France).

chojn1

2cicle,
I do admire your perseverance. I would have given up a long time ago.
That said, the 30-48 chainset you recommended worked flawlessly for me.
And I did have a chance to take it into the hills of the Adirondacks. No overshift and no chain drops.

If I may, I like to propose another hypothesis for your dilemma:
Flexion of the frame, bottom bracket, spider, and chainwheel/belt pulley especially when you are pedaling from the captain's crank could rotate the front of the middle(large) chain wheel in toward midline. That combined with the DI2 overshift could be the cause of your shifting problem. This is not a problem with the left side timing drive as it would rotate the right crankset out.

By the way, which front derailleur hanger are you using? I now have 3 different sizes on different bikes to get those front derailleurs where I need them to be.

Also, you may want to wait for the new XTR before upgrading again especially with your new 135 spacing and concern for narrower Q-factor as it moves your rear cassette slightly in compared to the ultegra.

Then again, you seem like you like to experiment even more than I do.
CJ

twocicle

CJ,
Frame flex isn't a factor here as the overshifting can occur on the workstand.

This was my first foray into Di2 and the FD cage inward limits are limiting its implementation more than I ever expected. The adapter clamp I am using is the 34.9mm version of Ultegra Di2 SM-AD67. With our previous mechanical shifting setup I used a K-Edge clamp, but IMO that is inadequate for the Di2 FD. The SM-AD67 provides a very good mounting platform for the Di2 FD and it has the stop plate required for the "support screw". There is no way the FD could be mounted further inboard because even as it is now, part of the derailleur mechanism is very close to touching the seattube.

The FD is now tuned so the movement to the outermost limit is just a hair off the chain when in the big ring, and I could only get it to that position by cheating the cranks 2mm to the right (placed the wave washer on the drive side instead of the non-drive as is typical for the standard single bike FSA cranks). This put the chainline (btwn granny and middle ring position) at ~42mm. Shifting seems to work perfectly at this point and combined with the 135mm spacing (moved the cassette 5mm inboard) we now have access to the full range of cassette cogs (barring crosschaining).

Looking at all the tandem cranksets available and our checklist of needs (narrow stoker Q-factor, plus a granny in the 30t range), it seems a Davinci crankset with a 103mm axle would be the only alternative for us. Todd's triple spider would place the inner/middle chainline near equal (maybe +1mm further outboard which is even better) to what I have implemented now with the FSA cranks and the Q-factor would be 158mm which is at our widest target limit. I don't really care about staying with SSD (same side drive) or not, just that we keep within the factors stated above.

twocicle

Updating thread

Over the last month or so Calfee respaced our frame to 135mm and got our rear wheel rebuilt with a WI centerlock disc hub (hub/wheel thread:
https://www.bikeforums.net/tandem-cyc...e-ind-hub.html). The rear wheel is still very stiff laterally, much more so than the Rolf Vigor Alphas I have on my single. We've been steadily upping the torture test on the wheels, driving them into corners at higher and higher speeds - they feel very solid indeed. No issues. With 25mm Schwalbe ZX tires, the only sound we hear from the wheels is the tires humming on the road. Zero wind noise, nada.

Among other added bonuses, this hub is 11spd compatible. Hmmm, couldn't resist the upgrade from 10spd... so 11spd Ultegra Di2 is on the way with new derailleurs (GS cage on the RD), cassettes (11-28 and 11-32) and chains (keeping the same shifters). The 10spd stuff is headed to my stoker's single as an upgrade to her 13 year old components.

I got around to creating a stoker bottle cage riser of my own design, using cheap AL stock from Home Depot and matte black spray paint. It weighs some 10g less than the steel version I posted about earlier, and won't rust. Also the cage mounting is easier to do and the whole thing sits better on the bike without spacers under it. I really should start learning how to make carbon parts.



We have also been finding that as our Camelback bottles soften with age, they were not held as firmly by the Arundel Sideloader cages, so we changed those out for the matte carbon Mandible version that holds bottles with a death grip (bunny hop and cobble proof). Sideloaders cages headed to the parts bin.

twocicle

Ultegra Di2 11spd (6870) upgrade feedback:

Great stuff! Silky shifting, precise chain alignment, slightly narrower chain. Smiles

Derailleurs:
The 32t capacity of the 11spd RD-6870 GS cage is a great option to have available - no cage hacking as I had done on the 10spd RD-6770. I'm sure this new RD would handle a 34t, but we have no need for that, and besides, 11spd Shimano cassettes are pretty much limited to 32t anyway (future XTR 11-40 aside... not needed here).

As with the 10spd Di2 front derailleur (FD-6770), the 11spd version (FD-6870) just barely handles the range needed for our narrow rear crankset (FSA single bike triple with 48/30 rings mounted on the middle/inner positions). Our big ring chainline position is right at the innermost limit of the Di2 FDs, so very careful adjustment of the outward FD movement is required to keep from overshifting. I've found that setting the limit screws to the minimum outward travel + tweaking (ie: bending) in the front lower tab of the FD prevents the dreaded overshift on our setup. This would not be much of an issue with a standard single double chainring, or a wider tandem triple crankset, both of which have their chainlines further outboard than ours.

Surprise (to me), the 11spd FD-6870 now has an electronic trim adjustment feature, just as the rear derailleurs always have had. This feature is not available with the 10spd FD-6770. Maybe I'm a little slow on the uptake, but I only learned of this feature while messing with the Etube software, then eventually figured out how to perform the task manually too.

The FD-6870 trim adjustment effects the auto-trim position. So for example, if the FD outer plate rubs the chain when using the 6th largest cog, you would want to adjust the FD trim outward a couple ticks. However, you can only perform this adjustment while in the big-ring/big-cog combination. Etube forces the derailleurs into this maximum cross-chained position, whereas during the manual method you must shift to these positions yourself. The adjustment feature is accessible via either the A-Junction box, or the “Front derailleur adjustment setting” in the etube software, and should be used only after you have already set the physical inner/outer limit screws.

That you cannot perform this adjustment while in the 5th or 6th rear cog or the small chainring, is a bit of a nuisance. I have no idea why Shimano created this restriction.

Levers:
Initially I just installed the 11spd derailleurs, chain and cassette, leaving the 10spd shift levers on the tandem for a few rides. These operated the full range of 11 cogs without any software changes. Everything worked perfectly. Then after making the decision to reuse all the 10spd components on Linda's single (new build), I ordered up a set of new shifters for the tandem and finally got around to installing them yesterday. The 10spd vs 11spd levers look almost identical and perform exactly the same. The only notable functional difference in the new 11spd levers is the 3 wire ports per lever (vs. 2 on the 10spd), plus some very slight cosmetic diffs (ie: darker levers). The front of the 6870 hoods feel a touch narrower than the 6770 hoods which I prefer.

In a nutshell, this confirms the reports that the Di2 shifter model does not matter as far as cassette cog count is concerned so you can use either if you wish - ie: save some $$ with the 6770 shifters.

Cassette/gearing:
I setup the 11spd cassette similar to what we had been running on the 10spd gearing... combining cogs from two cassettes (11-28 and 11-32) to achieve optimum gearing for us when using the 48/30 chainrings. However, while the 10spd cassettes can be setup as 11,13,15..., the 11spd cassettes now have a recess groove in the 12t for the 11t to nestle in. This means that we can no longer mess with the 1st two cogs. So, our 11spd mixed cog cassette is now setup with 11,12,14,15,17,18,20,22,25,28,32 with our primary riding gears being the 48x15-20. A 13t cog is missed when rising above 29mph, but we can live with that and the 14t seems more usable to us.

The 48x28 takes us down to around 12mph (bottoming out at a 90rpm climbing cadence) without needing to drop to the 30t granny. Then at the other end, the 48x11 (which is slightly taller than a 52x12) takes us up to ~35mph which is adequate for our needs. For the granny setup, I can use all cogs except the 11t without issue, but refrain from venturing to the 12t as well. That still leaves us with a 30x14-32, which is a lot of gears. The 30x32 "bail out", is more gearing than we ever used before... riding at 6.5mph at 90rpm feels like the world is standing still :/

Battery & junction boxes:
As we have a coupled tandem, I've relocated the internal battery and 2 junction boxes to inside the bottom tube. This allows for very easy access and wire routing instead of having the battery and 1 junction box in the stokers seat tube/post which was the initial setup and could be a real pain to access in a pinch, needing to remove the rear cranks just to get the junction box and mess of wires out of the stokers seat tube. Having everything in the bottom tube is a piece of cake, with 1 junction at the front coupler and the other box at the rear coupler, battery simply wrapped in foam and stuffed in the middle.

Summary:
Electronic shifting seems to be the only way to go once you get up to the 11spd count. I've been running Dura-Ace 9000 (11spd mechanical) on my single for 2 years and find the derailleur tuning (both front and rear) very finicky and probably a nightmare on a tandem. By comparison, once setup and dialed in, the Di2 shifting is flawless and zero manual cable/derailleur trimming required.

chojn1

Twocicle,
Thanks for the update.
Planning for our next trips, I think we're going to need some gearing changes.
Looks like moving to the 11 speed would be the easiest transition.
CJ

twocicle

3T Rigida Team fork experiment:

Tapered 1.5", 46mm rake, 367mm axle to crown (3mm shorter than the 370mm ENVE tapered which is advertised as 367mm).

The tire clearance above tread of a 25mm Schwalbe Ultremo ZX mounted on a 23mm rim is tight. 3T product manager said this fork would accommodate a 28mm tire, but not sure about that. I'll need to test it with a 28mm Continental 4-Season, which seem less tall but wider than the Schwalbes.

Just wheeling the tandem around the house with the Rigida fork feels less "floppy" than with the 43mm rake ENVE. ie: handlebars don't fall over to the sides nearly as much.

Ride feedback TBD.



Album of the above Rigida photos is available here.

DoubleDiamonDog

Glad to see that this is working out for you twocycle - way to stay with it - looks like a sweet ride, enjoy!

twocicle

Forking feedback (excuse the word play, couldn't resist):

re: 3T Rigida. YES!

After 2 rides including a 60mi torture test today (lots of hills and tight twisty corners), the verdict is in. The Rigida completely eliminated the "understeer" and dialed in high speed tight cornering. No longer the feeling of wanting to go straight, no fight to hold the tandem through fast corners. Smooth & natural the way it should be. Perfect!

As a secondary benefit to the handling, the Rigida (1.5" tapered) is also more compliant/smoother than the ENVE (1.5" tapered), but at the same time does not feel at all flexy. IMO the ENVE tapered w/43mm rake is not the best match for the standard geometry of a Med/Sm Calfee.

With the Rigida fork rake of 46mm, the performance is transformed into everything we were hoping for. Hopefully this feedback helps others with their setup choices too.

Bad1

Just hope it's strong enough for prolonged tandem usage for your teams sake! I'd be a little hesitant on a unproved fork for tandem usage considering the consequences. So dialed in high speed tight corners doesn't indicate what kind of speed are you talking about? Also a report on big sweepers like letting it run and getting into 40/50mph range, CoMotion rated fork nails it + piece of mind.

twocicle

Discussion of "tandem rated" components has been covered ad nauseam elsewhere. There is no formal tandem certification in any country or region, so any claims or inferences to that seem pretty much pure marketing. As I mentioned in this other forum thread, I had contacted the 3T product manager to inquire about this fork. Their forks pass all the stringent CEN (Europe testing) which is something I'm not aware of as far as either ENVE or Co-Motion's "tandem rated" fork.

Reposting the 3T PM's response here for convenience:
Skill level is difficult to relate in a forum thread where experts abound. I can say that the numerous sweepers in the 40mph range (and tighter 22-25mph 180 degree switchbacks) we just completed on the test ride were surprisingly easier than with the ENVE fork. Higher speed testing will come along as we adjust to the new handling characteristics and avoid loose chip seal. Safety first (like avoiding one stupid mini van doing a u-turn in the middle of a twisty blind-corner-riddled downhill section).

Anyway, just reporting my results and opinion. At a team weight of ~255lbs and no loaded touring, this fork is looking good for our use.