5

I recently set up a new gravel bike with all new components, including a 2x11 GRX DI2 drivetrain. 31/48 and 11-36t. It has ridden under 600 miles and only seen rain once. It is accident free and has never been dropped.

When in the large chainring, and shifting up into a smaller sprocket quickly, starting around the 7th or 8th sprocket, the chain bounces from side to side. While pedaling, the outer face of the chain links occasionally hit, and get stuck on, the outer plate of the front derailleur cage.

Last week, when pedaling at high speed and force, it even broke off the derailleur braze-on clamp.

  • The chain length is according to the instructions (4 links longer on big big combo). I also tried a chain with 4 links less, but it did not help.

  • The b-screw is set according to spec - with a ~5mm gap between the pulley and the large sprocket. I also tried increasing the chain tension (gap between pulley and largest sprocket was close to 20mm), but it did not help.

  • The front derailleur is set up correctly (aligned with outer chainring, and about 1 mm gap between outer plate and chain when in highest gear, and 3mm above the large chain ring)

  • The rear derailleur clutch is engaged and the torque is around 4Nm. I already cleaned and lubed it, but it did not fix the issue. I also tried increasing the tension to 5.5Nm, but it did not help. Turning the clutch off does not help.

  • I serviced the rear hub. It did not help.

  • I installed a brand new chain and cassette, but it did not fix it. For clarity, I had a 11-34 Shimano cassette with a 1.85mm spacer originally when this issue first occurred. I am now running a 11-36 Sram PG-1170 without spacer.

  • We tried out a different GRX crankset to rule out the likelihood of the chainring being bent or worn. It did not help

I had two bike shops look at it, but they could not find a solution (but validated that there is an issue).

The bike shifted fine for the first month. Or at least, it was fine enough not to bang the front derailleur cage. I do not know what has changed since then.

My next steps are

  1. upsize the pulley wheels
  2. replace the rear derailleur
  3. go back to riding my old bike :cry:

Do you have any ideas on what else might fix this issue? I really want to jump back on this dream bike but don't know how to fix it. Thank you!

—— Update: I tried out a friend’s mechanical GRX810 RD, and it worked like a Charme. We compared the geometry and noticed that mine doesn’t spring back nearly as far to support the smaller sprockets. Mine looks like the one in the first picture in its “neutral” state: enter image description here

While his looks much more “collapsed”: enter image description here

I am contemplating replacing my cage with one for a GRX810 or similar.

5
  • 3
    That it worked before and not now is indeed strange. 2 thoughts, but not answers: Shimano doesn't have a 11-36 11 speed cassette (they have 11-speed 11-34, 10-speed 11-36, 12-speed 11-36, you of course need a 11-speed cassette), and this cassette requires a 1.85mm spacer on the freehub body before inserting the cassette (not an issue on mechanical derailleurs, but depending on how an electronic derailleur is adjusted, may be an issue). What is the reference of the cassette?
    – Renaud
    Oct 12, 2023 at 20:49
  • Hi Renaud, thank you for your comment. I set it up with a 11-34 Shimano cassette initially, but it had the same issues there. I only switched to 11-36 recently thinking that the cassette was causing it, but the issue persisted. I had the spacer when using the 11-34 cassette, and since removed it for the new Sram cassette. I updated the post to include this detail
    – jonatans
    Oct 12, 2023 at 21:47
  • If you have all new drivetrain parts mentioned and correct setup, there are ony a few items left. 1--Bottom bracket: maybe a defective unit? 2--Frame: Alignment error? i.e. manufacturing fault. All your new parts are under warranty and under some circumstances it's possible to get Shimano to inspect them. But i think you will find the problem is not the individual components or even the set up, as you have had this independantly inspected.
    – Noise
    Oct 13, 2023 at 7:54
  • 2
    FYI, I am running what you are (48/31 GRX with an 11-36 PG1170 SRAM cassette) without the issue you are describing. I am using the first generation Ultegra RX rear derailleur, RD-RX800-GS (39T capacity, 16T max front difference). No issue with the RD clutch on or off. No problem with RD capacity. I wanted to throw this info in there to give a data point where this did not occur with the chainring/cassette setup you are using. Sorry to hear of the issue you are having. The gearing is truly a sweet spot for me, so I understand your passion in getting it working. Note: I use cable shifting.
    – Ted Hohl
    Oct 13, 2023 at 20:06
  • 1
    In fact another user of this stack (I can't remember who) indicated that they are using 11/42 cassette on RX815 (with a wolf tooth adapter), and using the "synchronized shifting" of Di2 to avoid the big-big combinations. Note that I'm also running a 46/30-11/42 combination on my fun bike (not a gravel, but same use) and find it excellent. Gravel bikes are among the bikes who would be benefit the most from large ranges and yet the offering is officially limited to 500%... (including the newest GRX 12-speed, except maybe by taking the 1x and a 9-52 cassette - with its own limitations)
    – Renaud
    Oct 14, 2023 at 7:13

4 Answers 4

3

You've cheated the drivetrain setup. Presumably this is RD-RX815, which has a nominal total capacity of 38, max low tooth count of 34, and min of 30. As such your chain is longer than intended and the b-gap in the trouble area on the small end of the cassette will be much wider than intended, since on the low end it's set to clear the 36.

I recommend trying it with the 11-32 that's the intended match for the 31-48 front. Set the chain length to match. If nothing else it will give you some more data.

Make sure the back support screw of the FD is contacting a solid frame member as intended. The FD braze-on "hanger" getting ripped off has a corollary with back support screw issues (or issues with the thing it's contacting being stable enough).

12
  • Thank you Nathan, this sounds feasible indeed. I shall try a 11-32 with an appropriately sized chain next. You are correct, it is the RD-RX815
    – jonatans
    Oct 12, 2023 at 22:17
  • @jonatans One of the pieces of what you have going here that may be helpful to think about it is that RD's guide pulley has no offset to the cage pivot. What that means is the effect of excess b-gap can be more pronounced than on the more common design of an offset pulley, which many might use as a basis for comparison without thinking about it. With the pulley offset, some of that distance will be taken up when you're in the high gears. Without it, the pulley will be a lot further away. Oct 13, 2023 at 2:26
  • Thank you @nathan-knutson. Do you have a recommendation on how I can modify my RD to offset the pivot to accommodate the 11-36 cassette? I will test out a 11-32 tomorrow to validate your theory, and also verify that my current RD in fact works as expected with the recommended capacity.
    – jonatans
    Oct 13, 2023 at 6:06
  • Note that Canyon sells their gravel bikes with the RX815 with 11-34 cassettes (latest Grail and Grizl). I don't think they would still do it if it causes problems of this kind.
    – Renaud
    Oct 13, 2023 at 6:52
  • 1
    Ok, I tried a 11-32 cassette with a correctly sized chain as per your suggestion @nathan-knutson, which improved the shifting, but did not eliminate the issue of the chain hitting the front derailleur. Tightening the clutch tension to 5Nm further improved it. And raising the FD to 5mm above the chainring further improved it. All this makes me believe that this is not a capacity issue after all.
    – jonatans
    Oct 13, 2023 at 18:17
2

After thinking about this for a bit, the questioner revealed a clue that revealed the solution to their problem. The cause of the problem is the poorly functioning, sticky rear derailleur cage, not travelling into its fully retracted position under spring tension alone. Cleaning and lubing/greasing the cage pivot on the rear derailleur to return the full range of travel under good spring tension will resolve the problem with the chain bouncing in the smaller cogs, especially on the large chainring. The rear derailleur could also be replaced, but that would be much more expensive.

So, how does the lack of full travel of the RD cage cause the chain to bounce significantly in the big ring/small cassette cog combination? It is caused by several factors.

  • First of all, lack of chain tension due to the sticky RD cage, with even less tension as the lower run of the chain is shifted to smaller cogs.

  • The larger rear cassette plays a part too. The larger cassette has a larger diameter, and more mass further from the centerline of the rear axle. At higher wheel speeds these larger cassettes have a lot of momentum, and if one were to stop pedaling abruptly at this high wheel/cassette RPM (i.e., when in the big ring/small cog), that cassette momentum will somewhat continue to pull a few links of chain through, causing the top run of chain to slacken, and chain slap on the top of the chainstay is likely. It literally will rotate the RD cage forward for a moment until it recovers to remove the momentary slack from the top run of chain and return it to tension. I discovered this phenomenon myself on my identically geared bike, and it only occurred when stopping pedaling at higher road/trail (i.e., wheel/cassette) speeds (20+ mph/32+ kph). I had to actively change my riding habits to not stop pedaling quickly in order to remedy this. Turning the RD clutch on stopped this behavior because the RD cage would not move as freely with the clutch on (I like the smoother shifting feel – mechanical shifting- with the clutch off). Also, my freehub and RD cage are functioning well. Note that a sticky/dragging freehub will also exacerbate this issue. The momentum of a large cassette at high speeds still will cause this momentary chain slap if pedaling is stopped quickly. With a smaller cassette, the issue is reduced significantly, to the point it is not an issue.

One can test the cassette momentum factor while riding or with the bike in a work stand. At a good road speed and in the big ring/small cog pedal the bike. Then abruptly stop pedaling, while observing the top run of chain to see if it slackens then tensions back up. Doing this in a work stand is safer, but it can be done while riding as well. In my personal experience I discovered this by the sound of the chain hitting the chainstay before it tensioned back up. Switching to a smaller (11-32) cassette minimized the issue that I experienced with a 11-36.

Now contemplate this high cassette momentum with a RD cage that does not spring back freely due to the pivot being sticky. That lower run of chain is not adequately tensioned due to the pivot issue, and at higher speeds, the RD cage is not doing the job it is supposed to do.

So put these components and conditions together:

  1. Poorly functioning RD cage with lack of spring tension (especially on the smaller cogs), and possibly a dragging freehub as well
  2. High cassette momentum (large diameter cassette) while pedaling at high speed (big ring, small cogs)

As long as firm tension is maintained on the top run of the chain, it should not bounce significantly. However, if the rider is not pedaling well throughout the pedal stroke (pedaling squares instead of circles), or standing on the pedals (much harder to maintain power through the dead spots without specifically training to minimized the dead spots), there will parts of the pedal stroke that may allow the top run of the chain to slacken due to the momentum of the cassette exacerbated by the lack of tension in the RD cage. With the poorly functioning RD cage tension, a bouncing top run of chain can occur.

Fixing the rear derailleur pivot to return it to freely springing through the full range, should solve most of the bouncing chain while pedaling issue. Checking that the freehub is not dragging is also important. If these two items are in top working order, the top run of chain should stay in reasonable tension enough that the chain will not bounce, even with a inefficient pedal stroke with major dead spots. Fixing the source of the problem will eliminate the need for additional band-aids to be implemented as the drivetrain will work as designed. In addition, working on smoothing out ones' pedal stroke to minimize dead spots will do a lot for the performance they can achieve while riding, and that applies regardless of a bouncing chain or not.

2
  • I had an Ultegra 6800 SS rear derailleur that had a sticky pivot for the cage that would not retract completely, leaving no chain tension when in the smallest cassette cogs. I CAREFULLY disassembled the pivot, cleaned and re-greased it, and it is still working great to this day.
    – Ted Hohl
    Oct 17, 2023 at 6:36
  • 1
    This problem of cassette inertia is well known on MTBs, and can cause derailments on flat terrain if the clutch is disabled (because it's possible to ride faster on flat terrain). That being said, my experience with it is that the derailments happen at the front, not the rear. Overall, I concur that checking the pivot should be the priority (especially if the derailleur is under warranty), that would be the most consistent considering that it was working fine initially (a 2x setup should work without clutch - on my funbike I have disabled the clutch, despite the 11/42 cassette)
    – Renaud
    Oct 17, 2023 at 7:18
2

The cage of the GRX 815 DR has holes for two different spring tension settings. The spring was originally mounted to the softer one. I changed it to be mounted to the second hole. This noticeably increased the tension, and the chain runs much smoother than before. I could bring the b-gap back to around 6mm, which further improved the shifting.

It also didn’t help that the first hole was slightly worn, which meant that the tension spring did not immediately engage when the derailleur cage moved forward. So this was likely the real culprit.

enter image description here

1
  • 1
    Nice find. The location of the spring engagement will resolve your issue, and since you have it apart, a little cleaning and fresh light grease will keep its movement fluid. You may be to remove some of your band-aids you were making to your drivetrain (e.g., FD positioning) if your bouncing chain is resolved. Even more improvements in shifting are possible getting those back in spec.
    – Ted Hohl
    Oct 17, 2023 at 23:17
1

A "thinking outside the box" possible cause could be a harmonic issue with the rear wheel. Since you have already swapped out the chain, cassette, crankset without resolution, it may be that the wheel/tire/sealant (assuming tubeless) at the wheel speed when in the 48 on the front and the small cogs in the back is creating a harmonic vibration that is manifested in the chain at those higher speeds. Tire sealant can dry out some in one location within a tire and create an imbalance that would not be there when the tire is fresh with new sealant. Possibly trying a different rear wheel to test this theory would prove or disprove it.

In addition, is there any play in the freehub, or is the freehub misaligned? Testing using a different wheel (different freehub) would also test this possibility.

Granted, these are not the first things one would think of to resolve this (I have never seen a physically maligned freehub, so that is really a shot in the dark), but you have covered a lot of ground already with your troubleshooting, hence, trying to think outside the box here.

1
  • 1
    Wow, thank you @TedHohl for these suggestions. I am in fact planning on visiting a friend that has a 11-36 cassette on his wheel tomorrow to try exactly this :)
    – jonatans
    Oct 14, 2023 at 5:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.