5

I've got 2 new fat Bikes (Framed Minnesota) with Shimano RD-M6000 rear derailleurs on them. (And another with an RD-M4120.) They're all using Shimano Deore triggers from the same line.

I am having a lot of trouble getting reliable (let also crisp) shifting on the M6000s across the entire 11-42, 10-speed cassette.

Specifically, I'm having trouble on both of them getting the chain to drop reliably from the 18 tooth cog (7th gear) to the 15 tooth (8th gear) (4th cog to 3rd cog, counting from the smallest). This is the main trouble spot when I get the cable tension adjusted for the crispest shifting up the cassette to the bigger cogs. (By crisp, I mean the shift engages quickly right after the shifter clicks, and an immediate release of that paddle keeps the gear.)

With this level of tension adjustment, Shifting up and down works fine on every other move, but it always hangs on that 18-15. When I use the shifter tension adjuster (which only has fairly large detents, like 1/4 of a turn maybe?) to reduce tension to the point where that 18-15 step drops reliably, it means shifting back up the entire cassette gets more difficult.

Specifically, at that looser tension, in order to get an actual shift (and not just a noisy shove against the next larger cog) I must push the lower paddle a tiny bit past the click and hold a bit, waiting for the chain to move up, before I can then relax, and have it hold the cog.

This is not the end of the world: It's not difficult to get the hang of this shift, and it only takes a split second, but it's not really what I expect from a mid-level Deore drive train.

Here's a video I shot showing the performance when it was (mostly) behaving.

(I get better performance with a lot less messing around with the above-mentioned 4120, and also with lower-end Microshift, Deore-compatibles on other bikes.)

I wondered if I were missing something obvious, so I re-installed the mech using the Shimano shop guide, and following that procedure was an unmitigated disaster. It has you adjust SIS tension by shifting into the 2nd cog, then increasing tension until the chain just begins to run against the 3rd cog and make noise. When I do that, the tension is so tight that it won't move down the cassette cogs reliably anywhere. I went back to the method I usually use, which is to mount, check stroke settings inside and out, turn the tension adjuster so I have about 10 clicks to play with, then attach the cable and try shifting from small to big, noting where shifts hesitate, and increasing tension only a click at a time until those shifts work, then testing the other direction -- which is usually just fine at that point.

I installed and adjusted many rear derailleurs that way, and these are the first that I haven't been able to tame.

The performance seems so randomly odd (usually, IME, if tension is too much, the derailleur will hesitate on several cog-drops, not just one.) that I pulled the cable on one bike. One of the pieces of housing was really frictiony, so I replaced it with a smooth piece of brand new PFTE-lined housing, and thought that would be the end of it, but it still wouldn't tune to my expectations.

I've also checked the derailleur hangers to make sure they're straight enough (check) and the small-cog limit screws seem set so that the cage is aligned well. Finally, the cage seems straight enough as checked against a straight edge.

EDIT, video of second (identical) bike exhibiting similar problems even after hanger was measured and adjusted and slicker cable housing was installed.

0

2 Answers 2

5

At around 1:21 when you shift down onto the second-largest cog, that brief grinding heard is a noise that usually goes along with the b-gap not being enough. That is probably not the whole problem here but it could be contributing. (See page 21 of the dealer manual document). I would start by screwing the b-tension screw in until that noise goes away.

It's possible the hanger alignment isn't good enough. I understand you've checked but I presume only by eye. Finickiness getting the cable tension dialed such that there's no lag in either direction even with a new cable system, all new parts etc is almost always the hanger alignment. There's not much else it can be. If it's not that, the next thing to look critically at on a new bike is whether there's a kink in the cable hiding somewhere invisible, a piece of housing routed funny or too long/short, etc, both of which can cause friction that can make it impossible to get it shifting well in both directions.

Modern RD systems should be seen as having no tolerance for RD hanger misalignment at all. Hanger alignment used to be something that could reasonably be seen as a "shop-level" problem, but it is not anymore. I won't promise that it's a hanger problem you're up against here, but it might be and I have little hesitation to recommend adding a hanger alignment tool to your kit for anyone who works on bikes. Most of the time when I build a new bike, they need some amount of alignment to really make the RD behave right.

Notably, when we look for acquired hanger misalignment on a bike in use, we usually mostly look in the vertical direction because that's the kind of damage that can happen easiest when the RD gets bonked against something. On a new bike, hanger misalignment in the horizontal "twist" direction comes into play more, which can be a little harder to see and gauge without a tool.

It can be seen in the video that there's quite a bit of cable slack between the housing stop and pinch bolt in the small cog position. That would usually go along with needing to push and hold to get it to shift into bigger cogs, as you're needing to do. But, the lag you're getting shifting onto some smaller cogs would presumably be much worse if you tried upping the tension. One possible cause of that lag is the cable not moving freely enough in the housing when you release tension, and the most common cause of that on a new bike is some kind of cable issue as above, either a kink hiding inside the housing, or a bad housing length choice somewhere causing excess friction. When everything is right you shouldn't be needing to click and hold as much, or have quite as much initial slack. It also shouldn't feel like the 1/4 turn movements of the barrel adjuster is too large for the margin of error you're working with - the current systems do have low margin of error, but when everything else is right it's still 2-3 of those 1/4 turn detents.

Engaging the clutch will tend to cause some increased lag but should not make the shifts unable to work when everything is right. It's best not to overthink the clutch aspect of setting up and adjusting these derailleurs. Troubleshoot and adjust it with the clutch off to the point that it's shifting smooth and more or less equal in both directions, and for the most part it will just go with the clutch on.

5
  • You were right on the alignment. It looked close, but when I double-checked it again with my DIY alignment tool, it was off 6mm horizontally 10mm vertically. I had tweaked it back, but I either didn't get it right or I did vertical first, then horizontal -- unfixing the vertical. I got those to within a mm of each other and the behavior is much better. I may shorten the housing run from the shifter itself up front too. It feels smooth through there, but it's probably 2-4 cm longer than I'd have cut it myself. All that little stuff can add up.
    – user36575
    Commented Jan 21, 2022 at 0:11
  • 2
    The last several bikes I've built up were on frames I manufactured myself, with dropouts and hangers I cut from very flat aluminum stock, so I guess I was spoiled and foiled by regularity on that front. I'll check the other recalcitrant one with the hanger tool tomorrow.
    – user36575
    Commented Jan 21, 2022 at 0:14
  • 2
    Glad that helped. The overlong housing notion wouldn't usually be if the run going from the shifter is a little too long, but more like if it being too long or too short somewhere was causing an overly abrupt angle at one of the stops or going around a corner. Commented Jan 21, 2022 at 2:09
  • Well, it's not the whole thing, unfortunately. I put the OTHER one up on the stand (identical to the first) and it had hesitant shifting. I straightened the hanger carefully (was out 10mm vertical, 5mm horiz) then adjusted. Same problem at 18-15 shift, but also hesitant going up 33-36 IIRC. I actually cut nice new housing (shimano & jagwire) and put dry PFTE on cable and reset. Still really wonky behavior, with hesitant shifts both up and down the cassette. I'm about ready to steal the RD-M4120 off my bike and try it on these to compare.
    – user36575
    Commented Jan 21, 2022 at 16:12
  • Here's a video of the similar problems on the 2nd bike (as described above). youtu.be/J3b3FlDlnBE
    – user36575
    Commented Jan 21, 2022 at 16:22
3

The real answer may be: The Cassette is lousy.

Final Edit: The real answer is: Cassettes were lousy. Replacing these with SunRace 11-42 solved the problem. Had them shifting trouble free within a minute or 2.

I have found it impossible to get the shifting reliable on those bikes. Decent, with frequent problems, is the best I can get them. So, I took a look at the cassette, which, per Framed bikes, was supposed to be a Sunrace 10 speed 11-42. That's what's on my Framed Minnesota. However, both these Framed Minnesota women's fatbikes have a "Sunshine" cassette.

Sunrace is budget, but their parts have always worked OK when I've encountered them, so, I took the wheel out of my bike (with the Sunrace) and I put it into the women's bike with the RD-M6000.

It took me only a couple minutes to get it shifting perfectly.

I will see if I can get Framed to send me the componentry they promised, but they do not answer their customer service email, even after multiple tries. I may have to just buy better cassettes myself -- if I can find any available.

EDIT 1: I did some filing on the teeth that take up (or drop) the chain on the 2 offending cogs (32 on the way bigger, 18 on the way smaller) and this improved shifting a bit. However, it's still not acceptable. After peering at the cassette long enough I noticed the spacing looked wonky. I grabbed a length of 2mm spoke to use as a feeler gauge, and it was sticking between some cogs and was much looser between others. IIRC, the nominal spacing between shimano 10 speed is 2.35mm, so a 2mm spoke should just slide in nicely. It did for many spaces, but hung up in several (due to burrs on the tooth cuts in some, and just narrower spacing in others) and it was signficantly looser between others, including the spacing between the shifts that are hesitating. Since, slow-mo observation of the shifting action shows that this is a matter of fractions of a mm between catching and shifting or catching, falling off, and staying put, I think this is further evidence that the Sunshine cassette is of unacceptably low quality.

When I put the 2mm spoke between the cogs on the Sunrace cassette (that shifts reliably) the spacing was much more consistent.

EDIT 2: I have received a pair of sunrace 11-42s I bought from a reliable seller on ebay, and the first thing I notice is that the cassettes separate into pieces at the 15t-18t gap and the 32-28 gap. In the sunshine cassettes, these are the problem cog jumps where the shifting sucks and the spacing feels off. Assuming they come apart the same way, this may be the source of the problem -- bad spacers? These would be very simple to replace with a 3-printed spacer, which would be a nice solution if the problem fits.

I'll report back after replacement with the sunrace and investigation.

5
  • 1
    seems plausible, sunshine is very low end. Commented Jan 21, 2022 at 22:22
  • 1
    You will probably find somewhere a disclaimer that component specs are subject to change without notice. To be honest, if they had to make this swap due to availability reasons, and they had such a disclaimer, that's pretty much fair game. A lot of stuff just hasn't been available, and with the way prices have been jumping it's also easily possible they wound up paying as much for the substitute part as the one they wanted. If it really doesn't function then that's one thing, but otherwise there's a lot to be said for roling with it in this era. Commented Jan 22, 2022 at 2:15
  • yeah, that's fair, but only if the components work. I buy cheap components (lotsa microshift) all the time for the bikes I build. Most have decent performance these days, honestly. Given that these two cassettes have the exact same fault -- and the fact that I could find multiple levels of the sunrace 11-42 from reputable bike dealers, I'm not as willing to feel generous about the swap there. I ended up ordering a pair of the sunrace b/c framed doesn't answer email. Today I may pull one and take a file to the teeth on the 18t cog since that's the biggest sticking point (literally)
    – user36575
    Commented Jan 22, 2022 at 13:31
  • edited, see above. filing teeth helped some, but measuring cog spacing with a feeler showed significant variation.
    – user36575
    Commented Jan 22, 2022 at 16:00
  • Only solution was replacing the SunSHINE with actual Sunrace. The sunshine cassettes can be disassembled via small torx. I may do this and measure the spacers to see if they're consistent and/or the 2.35mm they should be. May fabricate replacements for any out of spec and test shifting thereafter.
    – user36575
    Commented Feb 2, 2022 at 17:51