Hi,
If anybody solved an issue like the one below, grateful if you give me a hint:

I replaced SRAM Force 22 rear derailleur (11 speeds) - short cage with a new one with medium cage. I used new cable, housing and ferules. I also replaced the old 11-28 cassette with new 11-26 and 11-32 cassettes (two wheels). New chain also.

The issue now is that if I adjust the derailleur to work smooth on small cogs (12-13-14-15), then it works bad on the bigger cogs (17-19-21-23 require LESS cable tension). If I make larger cogs to work ok, then the smaller cogs require MORE cable tension. The last cogs (11 and 26) generally work ok, due to the limiting screws. It has similar behavior with 11-32 cassette.

A brief description is: when you start shifting from small to large cog, it is like the derailleur moves a little bigger increment than it should, and after around 4 shifts, the differences build up and become visible: the chain moves too much toward the wheel and starts to rub on the next bigger cog.

Grateful for any assistance.

1. Do you have the cable routed on the correct side of the pinch bolt? Wrong side of the pinch bolt changes the leverage ratio and the amount of movement per shift.

2. Do you have internal routing? The rear cable could be crossing the front cable inside the frame and changing the tension as you shift. In essence, the front cable acts like a spring and absorbs some of the tension as you shift to the larger cogs.

3. Are you sure you didn't buy 10 speed cassettes? It sounds silly, but a friend bought a used bike and brought it to me because it wasn't shifting right. It turns out that the previous owner installed a 9 speed cassette on a 10 speed system.

Thanks for the answer.


1. Yes, the cable is o the correct side of the pinch bolt.


2. I can not see how they stay inside the frame, but I can see them out at the bottom bracket - they cross ok a plastic part with parallel routing. I understand that tension might be absorbed here and there making the derailleur move up (toward bigger cogs) less than normal. But what drives me crazy is that here it happens the other way around: the derailleur moves too much toward the bigger cogs - it is like the tension in the cable is too high, but only in the zone with bigger cogs.


3. Everything is 22 speeds: derailleur, chain, cassettes. I think it rather behaves like a 9-10 speeds shifter, that sends too much incremental movement to a 11 cogs system. However, the shifter is also 11 cogs (it is the old one, I used it for years with small cage derailleur).

Not on a SRAM RD; the pinch bolt isn't a fulcrum. It's the half-pulley that forms part of one parallelogram arm that determines its actuation ratio.

This is weird, and no explanation presents itself...

However, since the RD is moving further than it should, you can change its geometry in the direction you need - try putting a couple of layers of tape over the groove in that aforementioned half-pulley bit. This will make it so the RD requires more cable pull to move a given distance.

This isn't one of those shifter-cables-cross-over-each-other-in-the-downtube situations again, is it?

Okay, that was literally the second question. Try disconnecting the ft der cable and seeing how it shifts. How about with the original 11-28 cassette?

In the case of crossed cables, it'll be shifting not far enough if anything, until the FD cable is pulled, which will make the RD move without touching the rear shifter. Pretty easy to diagnose.

Nothing explains the RD moving further than it should from just working the rear shifter, it's weird.

Is it possible the hanger is slightly bent? Not enough to mess things up with the short cage, but noticeable with the long cage? Or maybe the new derailleur didn't thread in quite right? 11 spd shifting is generally very sensitive to problems because the increments are so small.

Hi,
I'm totally lost now.

I opened the shifter, released the cable and cleaned the orange barrel on which the cable wraps (I thought that some dirt might "simulate" increase of the diameter which could lead to bigger movement of the cable). No change.

I switched to the old cassette. No change.

I removed the cable from the derailleur, cleaned the contact surface (although the derailleur is new) and re-attached the cable. No change.

It looks there is no way to force the derailleur to move as desired: it moves too much toward the large cog, whatever I do. I start to think that the derailleur is the issue. What can make the derailleur to move too much? Some bents that are not visible by visual inspection? The "B screw" adjustment? Or anything else?

Thanks

I made a test that seems to suggest an incredible idea: Would it possible that SRAM Force 22 WfLI (medium cage) has a design issue???

I surrounded the cable with a sleeve of around 0.5-0.7 mm thickness, over the cable guiding area - just before the pinch bolt. That changed the geometry, like increasing the "radius" of the circular of contact between the cable and its guidance on derailleur (it is not quite circular, but I have to call it somehow). Surprisingly, after adjusting the index for the new setup, it started to work almost ok.

I removed the sleeve, re-adjusted the index and... again: it consistently throws the chain too far away toward the wheel, step by step, until it starts to rub on the upper cog. So: why would have SRAM design this derailleur with such a strange behavior?

Is there anybody here who successfully uses SRAM Force 22 WfLI (medium cage) with 11-26 and / or 11-32 cassette?

Thanks

Check the hanger alignment.

Possibly longer cage accentuates any out-of-alignment.

(oh, as AeroGut says above)


I've used Force med cage (not 22) exchange with other with no problem.

Thanks for the answer.

How can I check it? Visually, it looks ok.

The hanger is a very thick piece of metal: around 3mm thickness at the level of the frame and then around 7-8mm below the frame, where the derailleur is threaded. It looks much stronger than the carbon frame itself.

But what is strange is that shift by shift, the derailleur throws the chain too much toward the wheel. In other words, too much movement at each shift. It is very constant, like it was designed to do that extra move. Can this be misalignment?

You can't tell enough visually. The only practical way is with an alignment gauge that screws in where the RD does e.g. Park DA-2

A shop can do it, but if you work on bikes much it's worth the ~$70 to buy one, 'tho it may only be used a few times a year.

The hanger is soft aluminum- it's designed to be weak so it breaks before the frame does.

Anyway, it's pretty easy to check & bend the hanger back into alignment, & is the cause of lots of mysterious shifting problems.

You need the proper hanger alignment tool in order to check it.
There is absolutely no way to know if your hanger is aligned correctly in both the vertical and horizontal dimension without the alignment tool.

The thing that is SO counterintuitive for most people is that it doesn’t matter if the hanger itself is brand new and there has not been any sort of accident. In the 10 bikes that I have built up from the frame or set up out of the box, 8 have had hangers that were not exactly in line with the frame.

The key thing to understand is that it’s not about a problem with the hanger per se; the problem is the RELATION between the hanger and the frame, and the connection point between the hanger and the frame is never specced perfectly. Plus, the aluminum on the hanger is really soft (which is why it’s so easy to use the tool to bend it and get it aligned).

Absolutely it can be. The indexing on the shifter and the derailleur is designed to move the cable the right amount if an only if the hanger is perfectly aligned. A misaligned hanger usually only causes problems in certain places. You can get a bike to shift OK through most of the gears even with a really crooked hanger.

I’m a total amateur mechanic, but I’ve adjusted a couple of dozen RDs, and I’ve done a hundred or more such adjustments. Almost always I can dial it in perfectly in just a few minutes. Whenever there is a problem it has always been an out-of-alignment hanger. (Oh, and one time where I had the FD and RD cables crossing over one another inside the frame; in that cases the RD still shifted perfectly, but only on the big ring OR the small ring, but not both.)

Thanks for the answer.
I'll try to fix the hanger alignment, I suppose this is the problem.

Yeah, doh. Hanger.

The hanger, indeed.
Ugly job to force bending the hanger, but now it works ok. And it needs only half rotation barrel adjustment between the 2 wheels with 11-26 and 11-32 cogs.

Thanks all for advices!

What, you tweaked it by eye? I guess it must've been pretty far out of whack.

In fact, now it is easy for me to explain: if the hanger is bent toward the wheel, then, the impact to the smallest cog is low, but it is bigger to the largest cog, because the derailleur extends and amplifies the error - that means too much movement toward the largest cog.
If it is bent opposite to the wheel, then it is the opposite: too little movement toward the largest cog.

I should have taken a minute to think about that, before all the useless tests I done....

Yes, I tweaket it "by eye", but very slowly, step by step, with 4 gradual attempts, each one followed by tests. I deducted the direction of the adjustment to be opposite to the wheel, because the movement toward the wheel was amplified.

So did you use a derailleur alignment gauge? The length of the Park Tools DAG-2 makes the bending part quite easy (sensitive). The 1/4” steel rod that is used to measure runout at 12 o’clock/3/6/9 o’clock on the rim sidewall is a bit time consuming since the rod wants to hit the chainstay and seat stay so you have to remove the rod and return it to the same setting with a tiny O-ring to remember the depth.

Good deduction, and nice work getting 11s sorted by eye.

No, I did not use a gauge. I deducted the direction of the adjustment and made it gradually and with tests. It took around 50 minutes. I suppose it could be a 10 minutes job with a gauge, but I do it only once, so I did not buy a gauge.

Just curious. The “eyeball” method can work, especially if you visualize the pulleys from behind the bike and use a large Allen wrench on the bolt that holds to rear derailleur in to apply force in the direction that zeroes out the misalignment and returns the cage to co-planar with the cassette - but it is not ideal. You might have tweaked it correctly in one axis only to now have a new misalignment that didn’t exist before...

I did not use visual info. I identified the direction of necessary adjustment (opposite to the wheel, because I had too much movement toward the wheel). Then I forced the cage in that direction with small steps, with many tests, until it became noise free. One of the adjustments was to twist the hanger a little, because the last pulley did not follow the chain very accurate.
I used a tougher approach than the Allen wrench, with the derailleur on site: I simultaneously applied half of the force by hand to the fixed corp of the derailleur and the other half on a long hexagonal key inserted in the big hexagonal screw of the derailleur. All with the bike wheel tightly in place. That allowed rapid successive tests.

For sure it is not perfect, but it works all the way noise free, so I think it is very close to optimum.