As with most "normal" double cranksets (my front cranks), there is no wiggle room to add BB/axle spacers, so it is not possible to add 3.6mm worth of spacers. Also, moving the eccentric to one side by more than a mm or so is not a good thing to do, especially if using a new-style Bushnell that grips mostly around the outer edges - you wouldn't want the eccentric sticking out and not gripping. Since neither of these is an option, attempting to shift the crankset and then equalize one side is not doable (or wanted).
This does bring up one thing, I installed the FSA wave washers on the drive side (instead of the non-drive) in order to improve the chainline by a little bit. I figure as long as the wave washer is installed somewhere, then there will be the appropriate preload and hopefully no problems in doing that. To compensate for the cranks being shifted to the drive side by a mm or so, I removed a FSA pedal spacer from that side (Speedplay pedal axles have a small flange that does not touch the carbon arm material, so the spacers are not required in this case).
Just a quick update on the Specialites TA rings. Heads up for anyone thinking about using the TA rings.
We have been experiencing an over-shift problem when upshifting to the big ring while using the inner cogs on the rear. It seems to create a chain angle that causes the chain to walk over the TA Alize teeth and drop to the outside... especially bad situation with the same-side Gates belt as that has caused a couple chain/belt jam situations. This is not a problem with the Di2 FD limit screw setting - that is set to the minimum distance needed to prevent chain rub on the FD.
While the TA Alize big ring has upshift pins to assist with getting the chain up to the ring, it does not have ramps or other tooth profiling to get the chain to sit down in the teeth valleys. This causes the chain to "skate" in some cases. I have been trying to re-profile (file down) the teeth to fix this issue, but for lack of a 100% solution (and no tech info explaining what profiling is required) I'm giving up on the TA rings for now and ordered sets of Stronglights instead (as recommended by Chris_W, xxcycles.com, good selection and much lower pricing than TA).
I have a discussion in the Mechanics forum about this, but no real fix found...
All the fun of experimentation. TBD...
A series of very small incremental adjustments to the limit screw has so far always found a setting that does not derail, has no chain rub in the small cog, and still lifts the chain easily. This is on a Santana with a 113mm square taper BB which I would think places the small cassette cogs more outside the crank than your Calfee. The trial and error process does however leave chain marks on my nice polished daVinci crank arms. :(
Please let us know how the stronglight rings works out. It would be nice to avoid the process.
Meanwhile, our Gates belt is getting chewed up from the derailments and chain jams. During the initial build with Di2 and SSD, I thought the belt might help prevent any outer chain drops, but no such luck, and in fact can result in a fairly ugly mess between the chain getting stuck under the belt and the Centertrack belt derailling. I'm being stubborn about replacing that (have a spare) just to see how long we can go before a failure appears imminent. Hopefully no sudden loss, otherwise my 100 watt rear engine will be getting us home by herself.
Thanks for the feedback. No such alignment in the "Mechanics" forum yet.
As I had previously posted about my new Macchaito build and the problems associated with the front derailleur on the DI2 be being so powerful causing the Praxis chain rings to flex and deflecting The chain off outward bound , i'm sure your TA rings were not built for DI2 and you're having similar problems, the DI2 FD slams the chain up to the big ring so fast that your rings have to be flexing and throwing it right off, without giving it any chance of settling in. FSA Super Type Chain Rings N10-11 are built to handle DI2 shifting. I don't know in your case if 50/34 compact is too large up front with whatever cassette is in use.but a 34 up front & 34 in the rear gives you 26.3 gear inches plus get rid of all your mix & match on chain rings that are out dated for DI2 usage.
Did you use spacers to move the 48T inboard? Maybe replace the spacers with a "catcher" or dummy ring that will prevent the over shift? I'm not an engineer or fabricator, and the dummy ring idea is probably just dumb, but I seem to remember something like that on cheaper cranksets (more likely there to protect chainring teeth...)
As I will soon also own a set of TAs, I may end up joining the "what the.." club.
I had also thought of using a ring guard (FSA makes a nice carbon version) as a outer chain catcher, but there is insufficient room to do that with the SSD Gates Centertrack rings & belt. Even the older CDC ring flange would sit too low to make any difference.
Sadly, thought I had this nailed down. We rode for over a month before the issue began.
EGAD! I just realized the problem only started after installing a new Ultegra chain. New chains are laterally stiffer which explains why the shifting issue occurs when upshifting while cross-chained.
I guess one could say a workaround is to use an older chain, but that would be just silly.
For decades my stoker has had a waterbottle mount on her handlebar with a plastic tube in place of the push/pull gizmo. Similar setup on pilot's handlebar.
No need to remove the bottle to drink
Rudy and Kay/zonatandem.
I am not clear how the ring flexing away toward the outside would make the chain jump off the outside. It seems to me that if the ring was pushed outward there would be less room for the chain out there.
I seem to recall your Di2 is 10 speed. If so I suggest trying KMC and Wipperman chains as alternatives. No Shimano pins. If you have to stick to Shimano then try an 11 speed chain.
If Di2 ends up limiting ring choice to Shimano dictated ratios that is very bad thing in my book.
Luckily I can get my mechanical setup to work great, it is just tricky and time consuming to set up. I look forward to reports on your new rings. Maybe that will fix the problem.
While waiting for the Stronglights to arrive, I am continuing testing with what we have and looking for more suspect causes. The upshift problem really does appear to be a combination of the chain angle (ie: when cross-chained to the inside on the back) plus the chain angle during the upshift as it rides the bottom of the granny ring while the top of the chain loop is climbing the big ring. Part of this problem may be that the granny and big ring are spaced a little too far apart and that is contributing to the greater chain angle along with the cross-chaining scenario.
Studying the 48T TA Alize ring and comparing it to some FSA Super Road rings (53T, 52T), the former (aka trouble maker) is perfectly flat on the inside then up by the teeth has an abrupt shoulder where they CNC the teeth. By comparison, the FSA rings (and my Shimano DA 9000 11spd) all have very gradual slopes on the inside faces.
Have you stood at the rear of the bike and have someone shift gears and spin the crankset so you can see what the front chain rings are doing in relation to the rear cassette, this will give you a whole different perspective of what you're looking at, hopefully you can see something you may be missing.
I agree difference in ring size may contribute to the problem as you described it. In addition since you have a large gap between rings, you may front shift when in a smaller rear cog than you would if the rings were closer in size.
Is the FD cage inner plate parallel to the outer ring? I angle mine slightly and if I angle it too much that will tend to contribute to my problems. Moving it closer to parallel makes it less likely to over shift and not shift quite as quickly from my middle to big ring.
Your comments remind me of an article about ramps and pins. I disagreed with it when originally read it but your comments are making me reconsider. You might be interested in the article.
Chainring Choice or Shifting Ramps? | Off The Beaten Path
The article seems to suggest a tooth profiling approach they might help.
More detail on the tooth profiles are in an article in The Winter 2012 issue of Bicycle Quarterly.
"The inner edge of the chainring tooth is chamfered to provide more room, anthe backsides of teeth are cut away slightly to make it easier for the chain to enter"
Some dislike the magazine but it is the only bike publication I know that attempts to analyze practice issues without marketing hype. It provides opinions that can be accepted or rejected based on detail data.
I went back and reread my BQ for that article and put our tandem in the workstand to see what is happening. Even though mine shifts to the big ring without dropping the chain, I see the same issue. The chain does not derail but falls into the teeth after about a quarter revolution. I moved my limit screw ever so slightly and the chain falls to the outside quite regularly. My finicky adjustment procedure must serve to get the cage in just the right position to prevent the chain from falling while it is on top of the teeth.
Thanks a lot for the links to the Jan Heine articles, they were very interesting to read. I had already discovered that using a Campy shifter works much better when using custom rings than does a Shimano triple STI shifter, and now I understand better why that is. Plus, it taught me why the size of the inner ring makes a difference to the performance of the shifting ramps, and why rotating one of the rings can help.
I would subscribe to BQ if only they made a digital version which didn't need to be physically shipped across the Atlantic, but unfortunately they weren't interested when I wrote to them with this request a couple of years ago. Feel free to let them know if you'd also be interested in having that option.
I could deal with the shifting the way it is (chain does settle onto the big ring after another 1/2 revolution), but the critical issue in my case is with SSD. While the chain is hung up on the teeth and protruding to the outside, rotating that mess around will catch the SSD belt (probably the same or worse case if a SSD chain was used instead of a centertrack belt). This is no fault of doing the SSD, but just sorting out the correct ring combos and shifting characteristics. When I commenced this "experimental" setup, I had some anticipation of a hiccup or two to solve.
Delving into the ring/pin/ramp/combo knowledge area is just another learning process... simply growing pains.
I put my single bike on the stand to see how it shifts. It has a cheap Sugino touring triple crank with 26/36/46 and ramps and pins. The pins are very small and look identical to the pins on the TA rings. It uses a 2006 Campy triple Centaur FD with Campy Record Ergo shifters. The FD is designed for 52 tooth rings and is a horrible fit for the 46 "big" ring. To my surprise it shifted perfect every time with the chain engaging both rings at the same time and immediately.
I suspect that relative disk size may be an issue. The BQ article makes a point that for optimum shifting the size of the inner and outer ring should be compatible with each other so that the chain can engage on both rings at the same time. The pitch of the chain requires that the distance from the tooth engaged on one ring to the tooth on the other ring be a multiple of that pitch. We should be able to manually lay the chain in a position so that it fully engages teeth on both rings. If not then there is no way it is going to happen when actually shifting. Once stated this seems obvious but I had never thought of it that way before. I will need to go back and check the distance between by 50 and 36 tooth rings to see if that is an issue. If so I may consider adjusting on or the other ring sizes or decide to live with the current setup.
Maybe we need a specific thread related to SSD?? I know that Calfee has done a fair bit of SSD, including a mtb set up with electronic shifting. They may have some info to share regarding the best chainring combinations that do not hang up on the teeth of the big chainring? Or anyone else that has had good luck with SSD?
Again - good info!!!
NP. While these issues may be more critical to SSD setups, the issue itself is more generic.
With my 48/30 rings, I am seeing good chain link alignment with the ring teeth. Also, shifts are happening regardless of pin location or engagement. In fact, with these TA rings I am seeing more shifts engage the big ring half way between pins than at the pins.
After removing a set of .6mm granny spacers (to test reducing the gap between rings), that had no beneficial effect. The reason for that is the inner TA ring shoulder bevel is playing a primary role in preventing the chain from engaging at a more parallel approach to the ring. I can now see the chain come off the granny, then bend over the TA's pronounced inner beveled shoulder and then catch the teeth at a sharper than necessary and problematic angle. For the most part, the pins on the TA rings cannot be engaged whatsoever due to the shoulder bevel sticking out and interferring with the chain.
I can't see any way to resolve this issue with the flat-sided TA outer/big ring except to use different rings that have smoother (or no beveled shoulder) transitions. I'll have to wait for the Stronglight rings to arrive and test those. If the Stronglight 48T or 50T also fail, then maybe I'll pick up a Dura-Ace 7800 50T, or use the FSA Super Road 52T I have on hand.
So I have a few questions, before things got to the Point where they are now could you use all 20 gears without any chain rub on derailleur cage, or chain pinging on front chain ring, when in the small ring? Your current rear spacing is 145? Fsa triple your using is for a 130 spacing? Using this set up I could see where you could potentially have the problems of being to narrow & without the correct chain line & to much cross chain angle, compound with your use of a 18 tooth jump in chain rings. You will probably be able to make adjustments in your chain ring choices to make your shifting work, but I question if you'll be able to use all 20 speed gears with the cross chain angles you may have which is not a optimal set up when you had the potential to have one. I can understand why Paketa runs there gates belt ring on the inside, to achieve proper chain line with their design Good luck!
Thanks for the heads up. I am a little worry about my upcoming trip, as I have a similar setup as yours (DI2 30/48 crankset) and I don't want to alter my set up right before the trip. So far I have not been able to reproduce the front chain drop off. But around here there is not any hills to fully test it. Also, I can't remember the last time I had to shift from the granny to the 48 while on the larger rear cogs. While you are experimenting, what about using an extended front derailleur clamp like the FSA tandem clamp along with a spacer at your bottom bracket (changing the chain line)?
it looks like a whole lot of time energy and money has been spent to experiment with a system that is plagued with issues. I talked to Michael at Calffe a few weeks ago about drive trains and in the conversation he stated that they have for the most part gone back to crossover drive, even on electronic shifting units. Too many non standard items and hard to provide proper OEM support for.
There is a past thread regarding Paketa custom mods to Lightning cranks causing spider failure on a right side drive. We have over 5000 miles on our DI2 left side belt drive with zero shift problems with lightning cranks 50-34 and 11-36 cassette. Sometimes reinventing the wheel is not worth the effort but I applaud your efforts Bryan.
1. While the drive cranks are standard FSA SL-K Light road triples, with the SSD setup I have the "big ring" installed in the normal middle ring position. This does not exceed the range of the Di2 FD.
1a. There is no derailleur cage rub. Never said there was. The Di2 FD works perfectly in spite of the extended 18T range I have setup. A key point to this is that the double cage design is pretty much uniform and not ring size specific as some Shimano triple FDs.
1b. Apparently Calfee liked my build so much that they put it up on their FB page. Back late last year, they said they had not figured out how to do the centertrack because of the ring width, whereas the CDC had the flange and built in offset which made that easier to do. I showed them otherwise. It works great with just a ring selection issue to figure out (noted below).
1c. List of non-OEM parts... 2 rings, longer chainring bolts, and some washers are all that is needed. Oh yes, do delete 2 tandem specific cranksets and some excess weight.
2. With a 145mm rear spacing the above chainline is biased by a few mm to the big ring. Since the bulk of our riding is easily done within cogs 3-7 (inner to outer) and the big ring, this chainline is optimal for maybe 90% of our use. Granny is only needed for cogs 1-7, so no crosschain issue there.
3. The only currently problematic issue found was while shifting up from the 30t granny to the TA 48T, while in cogs 1-3 (dubbed as cross chained). It was since pointed out that the lack of tooth profiling on that TA ring is causing the chain to hang up when shifted in that cross chained scenario. A little custom tooth profiling (filing) has helped a bit, but that ring has other non-solvable problems such as the inner shoulder bevel/lip that interferes with the chain achieving a proper angle to settle in. I am very confident the issue was the TA 48T ring - took photos of the case for later comparison.
Oddly enough, this entire setup worked like a charm with a used Ultegra 10spd chain installed. Once that chain was replaced, the new chain revealed an issue with it vs the TA 48T ring. As shown by the article provided by Wayne, this is not new and certainly nothing to do with SSD or Di2.
I now have sets of other, better profiled rings on hand to test. Hey, I did say at the top this was something of an experiment :)