Alcanbrad

I have pretty much given up on the conversion to electric shifting I started last May. Love the shifting, love the braking, CANNOT prevent the ghost shifting on steep climbs which is unacceptable. I have tried everything and am at the point where I believe that 11 speed on our Co-Mo Carrera at our weight is just not compatible for the hilly terrain we ride.

I have the choice to go back to the original drive-train (Ultegra 6700/ TRP Spyres) but I was never happy with the shifting and braking performance. We never have had a issue with the ghost shifting.

We really want hydraulic brakes and the options in 10speed road format are very limited. Shimano Tiagra 4700 and GRX-400. I have the Tiagra group on my cheapo gravel bike and am very happy with the shifting and braking performance. The interwebs say that GRX400 borrows from the 4700 and they items are cross compatible so I don't think there is any advantage of one over the other.

I have 2 items to deal with: Capacity and choice of calipers:

Capacity: Minimum acceptable gear range for us in a double is and 34-50 mated with an 11-40 giving a capacity of 45. The two rear derailleurs would have to use a Wolftooth Roadlink with the cassette. The 4700 lists a capacity of 41T and GRX-810 lists 42T (which is 1 more gooder). I know that the Tiagra and GRX-400 rear derailleurs are in fact 11 speed derailleurs with the 10 speed compatibility being achieved in the brifter, so I should be able to use any 11 speed road derailleur with these brifters.

Are there any other limitations, or issues I need to be aware of here? (Anyone run 4700 or GRX400 on their tandem now?)

Brakes: The calipers for both of these groups are flat mount and I want to stick with 203mm rotors. The pad size for the 4700's are very small so I would be uncomfortable with that at our team weight (>400lbs).

I am looking for recommendations for post mount mineral oil calipers for our use. I imagine something from Shimano's MTB line would be in contention but I have no experience with the MTB side of the equation and am unfamiliar with any of the parameter differences that might exist there (as we know, Shimano is synonymous with incompatibility)

Post note on the ETAP: I have 1 more thing to try - a different chain, however, I do not hold any hope whatsoever as I have not found any issues with the brand new chain I have currently. I did just mount my go-pro to capture video of the issue which will tell me whether the chain is trying to climb to the larger cog or trying to fall to smaller cog. After that, I may try it on our Co-Mo Primera, however, the frames look of very similar construction so I think it might just be a waste of time. I may elect to sell the ETAP to recoup some of our money or maybe look for a good deal on a nice half bike and put it there. Also, the better half keeps mentioning that we should consider a hub e-assist set up which may reduce the amount of power we need to put into the bike to muscle up some climbs but I am not sure we are at that stage of our chronological abilities.

Signed,
Desperate in NJ

DubT

I installed the X-shifter on our tandem going on two years ago. It works! https://www.sportcrafters.com/products/xshifter

Electronic shifting made simple.

OneIsAllYouNeed

Ghost shifting due to frame flex will likely be worse with a cable-actuated rear derailleur. Have you tried micro-trimming the individual gears? Most likely, they'll need to be trimmed outboard (toward the smaller cogs).

oldacura

I think I agree with OneIsAllYouNeed: "ghost shifting" is often caused by frame flex. If you are a heavy, strong team, the tension in the top span of your drive chain may be causing the rear triangle to warp to the right which causes the chain to switch gears. I'm not aware of any way to strengthen an existing frame.

Your greatest torque on the frame would be on the smallest chainring. If the ghost shifting is caused by frame flex, I would guess that it may try to shift you to the next smaller cog in the rear.

I don't know if this makes any difference but is your bike a same-side drive (right side timing chain)?

Another solution might be to revert to a 10 speed or 9 speed cassette. These have larger spacing between the cogs and might reduce ghost shifting. One other possible way to reduce the torque on the frame might be to go from an in-phase pedal position to a 90 degree out-of-phase position.

Bad1

I would say you need to be using Sram Red drive train period, On my Filament tandem on the build I used Praxis rings, KMC chain, Shimano cassette, shifting was poor. A friend in town who develops for Shimano advised me to run strictly Sram components with the thought that different components do not play nicely with each other. After swapping componets out shifting was right on no problems.

Alcanbrad

I have tried micro adjusting in, micro adjusting out, played with chain-line, B-screw, tried the Digirit large cage from Chinook, and changed cassettes. All I accomplished is moved around were and how bad the ghost shifting occurred. Basically it occurs in the 2, 3, and sometimes 4 and 5th cogs and when on the small chain ring. We have a classic left side timing belt. We have never had the issue when in the largest cog or when in the large chain ring. We routinely have short climbs ranging from 10-20% and we do literally have to just muscle over them. I don't think changing the pedal phase would work for us a team and at the cadences and speeds with find ourselves muscling through. The ghost shifting is unacceptable when moving at 2mph uphill with no momentum.

The concept of the torque being highest with the small chain ring is interesting. How much does the number of teeth of the cassette contribute? In other words, is the condition more likely to occur on larger cassette cogs? With the 11 speed, we experience the ghost shifting with 34 chain ring and 35 cog (were running 34-50 front, 11-40 rear). With the original 10 speed triple, we were running a 28 chain ring and 11-32 cassette. So with 28 front, 28 rear (2nd cog) we do not have the issue I would think that the 28 chain ring would cause more flex than the 34, correct?.

If we go back to a 10 speed, but stay with double compact crank and get an 11-40 10 speed cassette, are we more likely to experience the ghost shifting at the 35 tooth cog position? We really like and want to stick with hydraulic brakes and I haven't found a triple front brifter that drives hydraulic brakes. I am thinking about putting the original drivetrain back but keeping the stoker crank as a 34-50 double and putting on an 11-40 cassette and see what happens. The rear derailleur is more limited capacity wize, but it would allow us to try before we buy so to speak.

A 10 speed 34-50 compact, with an 11-40 cassette gives us an acceptable range, but I am several thousand dollars sunk into the ETAP and I don't want to throw another $500+ more on new brifters, calipers, and drailieurs if I am likely to experience the ghost shifting with the larger cassette. Any thoughts?

Alcanbrad

I wish it were that simple. The journey started with the ETAP Red, SRAM PC1130 cassette and SRAM Red22 chain. - all SRAM and that was the worst. Things got better when I went to a Shimano cassette, but better is a relative term.

I know many people run ETAP on their tandems with no problems and we also have no problems, that is until we do, and we do when climbing very steep climbs under high torque. If we are not in that mode, all is well.

I imagine that if we lost about half the weight, things would get better, but losing the stoker is not in cards , or I guess we could move to an flatter area, but we like it here and the cycling is awesome.

reburns

Would just like to make sure I understand; when you say ghost shifting, the behavior you are describing is that, under high torque such as steep climbing or out of the saddle efforts, the bike suddenly shifts to a different gear on its own accord, correct? Not skipping, but in a new gear. Is that new gear a higher or lower gear? Does it happen more often in certain gears than others? For instance, if you are climbing, I would expect you to be in a low gear. Does the bike then shift to a higher gear/smaller cog? Does your gear indicator change at all when this happens? Can the stoker see the chain on a different cog than you selected, or even a different chain wheel? And the bike then shifts back to the originally selected gear when you let up on pedal pressure, correct?

I’m mystified why this would happen with a wireless derailleur vs a cable actuated derailleur. Did the chain line change significantly when you made the "upgrade"?

Sorry, none of this is helpful, but curious minds are wondering...

oldacura

If your team is 400 pounds and you are muscling up a steep hill a 2 mph, you are putting a huge amount of torque on your frame.



Assume you are putting the same amount of force on your pedals regardless of big ring / small ring: The amount of tension force the upper span of your drive chain sees depends on the vertical distance from the center of the crank to the tangent point where the chain meets the ring (basically 12 o'clock). For a 52t cog this is about 4.1". For a 34t cog this is about 2.7". So the chain sees about 52% higher tension in the small ring -vs- big. The rear triangle of your frame has to resist this tension. This tension causes your frame to flex.



If you imagine that your are looking down on your frame from above, you would see that the rear portion of your frame is curving to the right as this force is applied. This would likely cause the chain to drop onto the next smaller rear cog (higher gear). I think it has less to do with which cog you are in at the rear. However as the chain moves away from the centerline of the frame (toward the smaller rear cog), the tension force of the chain has a greater moment arm against the frame and will flex the frame more.



We too have never tried an out-of-phase crank position (because we had no reason to). However, at least as an experiment, you might try going out-of-phase under the same climbing conditions to see if that fixes your problem. When you are in-phase, the highest tension on your drive chain will be when both of you are between 12 o'clock and 3 o'clock in your pedal stroke. By going out-of-phase you will shave off this peak tension and even out the power.



When a bike frame maker designs a frame, they have to strike a balance between strength and weight. As a 400# team (especially if you are strong) you must be near the upper limit of what they planned for.



Good luck and keep us posted.

oldacura

Also - I re-read your initial post. You might contact Co-Motion about this issue. I have found that they are very knowledgeable and helpful in regard to such issues.

Alcanbrad

I am mystified too. As near as I can tell it is trying to shift t a larger cog. It never makes it into a different gear but the pedal stroke snaps forward a little. The reason I believe it is trying to climb into a larger cog is we have never experienced the issue in the largest cog and I have tried 36, 40, and 42 tooth cassettes. Also, it never happens in the large chainring which imposes a greater angle away from the adjacent larger cog (however, other comments about torque on the smaller rings may mitigate this). We cannot see it happening, however, I have mounted my go-pro and intend on shooting some video next time we ride (which will probably be the first ride it doesn't occur leading me to conclude I need a camera focused there in order for it to work :-))

There is a difference in this wireless derailleur as the parallelogram has some free play it. I will post a video of that shortly. (Also, potentially by coincidence, this free play is not present in the largest cog.)

Alcanbrad

Thanks for the explanation of how the torque manfests itself. I am an engineer, but was having some sort of mental block picturing it, so this clears it up.

As near as I can tell, it is attempting to shift towards a larger cog, but I have not seen it happen. As I stated earlier, I have my go-pro set up to capture it when it happens.

One aspect that I think may be related, but Ric and House of Tandems says is not an issue, is that with the rear derailleur in positions 2 through 10, there is play in the parallelogram movement inward (towards bike centerline). With a cabled derailleur, the unit is held in a fixed position in each gear and would only move if the flex in the frame changes the effective cable length. That shouldn't be an issue with the electric, however, the parallelogram can be very easily moved with 1 finger. Here is a short video I took:


OTHERS WITH ETAP, DOES YOUR REAR DERAILLEUR DO THIS????

(It moves easily with 1 finger, I had to brace the derailleur with my hand due to the bike moving in the stand and needed the other hand to hold the phone). If I set up the derailleur per SRAMs procedure, when I push as I show in the video, it will attempt to climb to the adjacent larger cog. I don't know if this is the cause. Coincidentally the parallelogram does not exhibit this free play when in the 1st and 11th position. We have never experienced an issue in the first position.

My wife has pretty much said that she won't ride this bike anymore with this condition and I am not far behind. I would like to understand exactly what is happening and convince myself that if I go to the same gearing combination in 10 speed cable actuated that I will not have the problem.

Oh, I did call Co-Mo a few months ago. They were not at all interested in helping me solve the issue, all they would tell me over the phone is that they do not have any issues with the Ultegra cable and Di2 11 speed solutions. But, it could also be that they made changes in the frame to beef things up.

Any and all comments, suggestions, and experience are wanted.

IPassGas

Does the stoker have experience with tandems other than your co-motion such that she/he can feel the frame flexing? Your frame would flex the greatest at the stoker's BB when standing and applying power. Can the stoker sense flex? Does the stoker's sense of flex change when both or just captain are standing? Do you ever carry gear/panniers, and if so, does the stoker's sense of flex change when the captain is standing?

Alcanbrad

We do have experience with other bikes but can she can tell the difference? That’s a big ask. She is not wired that way. Also, I have not been able to get her to stand and pedal at the same time and when I try to stand and pedal she fights the bike - it just doesn’t work for us. I am pretty sensitive to minor changes in things, however, I don’t sense any flexing but our speed and cadence are very very slow when we experience the issue.

I have felt some flexiness in the frame when rolling through small dips and bobs at moderate speeds on pavement. We have done a few overnight trips with panniers, however, other than the added weight I can’t say say I noticed any different flexiness.

IPassGas

There is a trade-off with frames lacking a diagonal stabilizer, weight vs. flex. Your frame may be too far towards the flex end of that spectrum.

oldacura

You said above that you reach 2 mph when climbing - and you do this with both in the saddles. This is hard to comprehend. We don't do steep climbs much anymore but the last time we were on a very steep climb we were both out of the saddle and the lowest speed I saw was about 3 mph. I found that it was really hard to keep the bike balanced at that speed. Our team weight is about 300#. The force you are putting on your bike must be tremendous. I think your GoPro video will be very enlightening.

Alcanbrad

It might be 2 or 3mph - we grind down to the point of just pushing the bike against gravity and it only happens when we are at the very slowest. I need to try and convince the stoker to ride this bike again to capture video. I'm all set up for it, all that is left is the smoozing,

I did some chicken scratches on a pad of paper and I should have less torque going into the bottom bracket with a 34T chainring, than with the 30T I had with the triple (for the same effective gear ratio). That said, it brings me back to why we are experiencing this with 34-35 vs. 30-28? Without yet knowing if the chain is trying to climb or drop on the cassette, Is cause more likely due to the narrower spacing of the cogs on the 11 speed cassette, or the additional torque applied the rear hub due to the larger cog diameter of the 11-40 cassette (or is the free play in the parallelogram I show above implicated)?

oldacura

The torque placed on the stokers bottom bracket by the timing belt (basically the force you are applying to your pedals) will attempt to twist the stoker's bottom bracket clockwise (when viewed from above).

The tension caused by the drive chain trying to rotate the rear wheel is also attempting to twist the stoker's bottom bracket clockwise (when viewed from above).

The combination of these 2 torques could twist the chainrings quite a bit out of plane with the cassette (the chainline?).

LV2TNDM

If you can utilize some stout string on your frame to gauge frame flex, this may help solve the issue. Or at least identify or rule out flex. If you have ever used the string method to check frame alignment, then you'll have an idea of what I'm talking about. If not, check out this YouTube video:


However, this is usually done with the rear wheel removed and no one actually riding the bike. But I'm sure there might be a way to route a string from the head tube to the rear triangle so your stoker can see if s/he can observe the string and frame changing position relative to each other. Focus on the seat tube area to observe flex. Perhaps attach some rudimentary measuring cards next to the string so if flex moves the string's position, it will be easy to see. Or perhaps use a laser pointer/laser level? There must be some way to assess frame flex issues while riding. And I realize tandem frames flex in more axes than just side-to-side, so perhaps there's a more rigorous method to assess flex issues. You mentioned you're an engineer, so I'm sure you can appreciate this.


That being said, have you confirmed that your frame is indeed in alignment? I ask because if it were out to one side, this may also be the cause. It could be out of alignment an amount that doesn't cause any problems, but under frame flex, it is near and quickly passes a threshold that exceeds acceptable chainline and/or chain/cog mesh.


Also, no issues with a loose derailleur hanger or dropout somehow?


And this is the same rear wheel/hub used previously? Axles can flex under high loads, so this could be a cause as well. Have your stoker watch the rear wheel between the chainstays under high pedal loads to see if there's any movement.


Boy, this is obviously frustrating for you. Good luck with resolving this; I'm sure many of the people here are eager to see you solve it and learn from your experience.

merlinextraligh

Another, albeit expensive, answer would be a new, more rigid, frame

We had a Co-Motion Robusta. With oversized aluminum tubing, and a lateral tube, it was theoretically very stiff. Likely as stiff or stiffer than your Carrera. Riding the bike hard, there was visible flex in the rear triangle, to the point that it worried people around us, and caused us to have to back off sprints occasionally due to the flex.

We upgraded to a Calfee Dragonfly with the extra stiff option, which rides beautifully with virtually no lateral flex. With a team weight that has at times been ove 350 pounds and a combined max power over 2000 watts we can’t begin to flex it.

Again, its an expensive option, but I think you would love it.

Alcanbrad

We haven’t ridden the Carrera since before my last post, but hope to in the week or so. I am pretty much convinced it’s the frame combined with the cluster spacing of 11 speed. A new frame would be nice, but as you said, it’s expensive and we would always be “waiting “ for the ghost shifting to occur. I still will try and capture video as there should be something for us to learn or satisfy our curiosity.

as my head clears of the frustration I think the plan will be to figure a way to get 10 speed hydraulic on the Carrera and build up a nice half bike With the Etap for yours truly.

WheelsNT

Could also be the pawls in your hub slipping under max load. Whenever you get video, if you don’t see the chain moving off the cog, it might be the hub. I agree the most likely scenario is shifting to another cog induced by frame flex, but I wanted to mention another possibility.

tandem rider

Since you did not have a problem with ghost shifting with the Ultegra 6700, I don't understand why frame flex would be the problem with the ETAP. I once had the problem that you are describing going up steep hills on the Hilly Hundred. It is unnerving and hard on the rear wheel. At the first sag stop the mechanic replaced the cassette and my problem was solved. You mentioned that a Shimano cassette was a little better than the Sram, but I wonder if something is keeping the chain from fully engaging on the cog. As WheelsNT mentioned, it could be something with your hub.

Tandems East is probably close to you. They might have an answer to your ghost shifting.

Good luck fixing your problem and let us know if you find a solution.

LV2TNDM

I'd be surprised if it were a freehub body skipping problem. I've been through three freehub bodies, pawls, etc. on my current tandem hub. And that's after destroying six hubs catastrophically prior to this one. But the recent skipping I've experienced is not very disruptive to our riding. It's obvious it's internal the way it feels and is nothing like a chain skip or miss shift. It's barely a bother. But after about a year of only occasional skipping, I disassembled it and found sufficient wear on the pawls and engagement ring. Internals were just replaced. But again, this skipping is barely a bother when riding challenging terrain off road.

Alternatively, the catastrophic freehub failures are obvious and end the ride right there. If Alcanbrad had such significant problems with the freehub internals, I'm assuming he'd notice other symptoms. E

But who knows, it would be worth investigating this as well.

reburns

I think those reading this thread all hope that you can learn the root cause and let us know what you learn. I have to say that I’m skeptical that this has anything to do with frame flex. I have experienced this disconcerting skipping forward of the pedals under high load at times in the past, and it always turned out to be a worn chain ring. The fact that it never happens with the larger chainring or on the largest cog would be consistent with a worn smaller chainring, which is the first thing I would replace.