4

I have a Shimano hollowtech crankset mounted in a road bike, and it has horizontal play through the bottom bracket. There is no "radial" play (that is, no "wobble" but the cranks can shift sideways, parallel to the axle.

  1. Is this horizontal play normal?
  2. If its not normal, what am I missing to take up the gap?

By measuring from chainring to seat tube, the difference of 1.15mm is measurable. enter image description here

Video at https://imgur.com/a/iExI5Z2 or https://youtube.com/shorts/BblXEGbxIRE

enter image description here
BB is a Tiagra BB-RS500

enter image description here
Crankset is a FC-6603-A

enter image description here
NDS crankarm is branded Ultegra, so its very likely not MTB.

enter image description here
Drive side.

6
  • 1
    The horizontal play is absolutely not normal. Just to check, was the crankset assembled correctly (as in the NDS arm is fully seated) and the crankset cap is fully preloaded?
    – Weiwen Ng
    Commented Apr 23 at 15:42
  • @WeiwenNg yep - the BB was brand new, and has the tube between them installed. The crankset was used, but the left-side crankarm seats firmly amd preload nut is done up with a 3d printed tool to fit its cap. I suspect there's supposed to be a spacer, but between which surfaces?
    – Criggie
    Commented Apr 23 at 19:27
  • 1
    No spacers for stock Shimano BBs. Chris King may have spacers, but I don’t think so. There is no chance this is an MTB crank, right?
    – Weiwen Ng
    Commented Apr 23 at 19:56
  • @WeiwenNg it was bought used, and is branded as an Ultegra triple and is labelled FC-6603-A The BB was new, and is a BB-RS500 and is Tiagra groupset though isn't labelled as such. More photos to come, and I realised on the ride into work that feeler gauges would have been a better tool to measure the offset.
    – Criggie
    Commented Apr 23 at 23:11
  • @WeiwenNg the frame dates from 1994, so its just in the area of weird sizings before standards prevailed. EG finding a headset lower race that would fit was a mission.
    – Criggie
    Commented Apr 23 at 23:12

3 Answers 3

3

For Shimano FC-6603 crankset 3mm spacer (CR6774) required on non drive side marked as "10" in shimano documentation.

2
  • 2
    Additional context: this is a triple crankset, it's extra long on the DS, and I'd forgotten that triples existed. So an extra NDS spacer makes sense.
    – Weiwen Ng
    Commented Apr 24 at 15:47
  • @WeiwenNg my (excessive mass) needs a triple because going up hill is slow, but I also like a really high gear for tailwinds. Also my old bikes tend to top out at 28 or 32 tooth cassettes, so running a compact chainset with a dinner-plate cassette isn't feasible. Triple 4 lyfe, yo!
    – Criggie
    Commented Apr 24 at 22:52
1

Further to other answer, Road triple shimano has (should have) a plastic spacer on drive side (i don't have one here to measure), i think it's 3 to 5mm thick, and further, has a non drive side spacer similar thickness but same diameter as bottom bracket shell and slightly conical.

The reasons for this are that the casting for the drive side triple is the same as the double of the same version in the range. The drive side spacer is added to put the double inner position to the correct chainline for triple middle. The spindle is longer and the non drive side is spaced to even up the l/r spacing. Obviously for this reason, triple q factor is bigger.

If you have lost your spacers, and can't get shimano or aftermarket replacements, try using an MTB bottom bracket, putting in the shell spacers where necessary to get a good fit. The spindle won't mind and the mating faces will be correct.

4
  • From your photos it's already obvious you're missing the non-drive-side spacer. But we can't see the other side.
    – Noise
    Commented Apr 24 at 18:33
  • Looking at your measurements, i think you are only missing the ND spacer. There is a range of adjustment in the position of the LH arm that accounts for the "missing" mms
    – Noise
    Commented Apr 24 at 19:05
  • The symptom while riding is that its really easy to drop the chain off the big ring inboard, while riding. And its really hard to shift up back to the big ring. I've discovered that sliding both feet to the left can assist the shift, and that cross-chaining makes it even more likely to drop. Based on the two answers I'm going to make some spacer/shim for the NDS, just not sure if it will be between crank and BB, or BB and frame.
    – Criggie
    Commented Apr 24 at 22:49
  • Aside, I'm almost 50 - a wider Q factor is a good thing.
    – Criggie
    Commented Apr 25 at 3:21
1

Well - thanks to the two excellent answers, I've fixed my problems.

  1. Play was reduced to almost nothing by adding a 1.2mm thick spacer with an OD of 35mm and an ID of 24.5mm You can see it here in Grey.

enter image description here

enter image description here

  1. Additionally, there's a strange little hook in the pinch area of the NDS crank. On removal, it looked like this, with the pin mushed into the plastic as shown.

enter image description here

There's a divot on the axle for that tiny pin to rest into, but it doesn't seem to provide anything of importance. Redoing this likely achieved nothing.

  1. The upshift problem was definitely exacerbated by the horizontal play, but there was also no pin/hook/ramp to help guide the chain onto the big ring. So I cheated and added a low-profile M4 bolt into a pre-existing hole. This really helps haul the chain up, even if it takes an entire crank rotation to find.

enter image description here


Final update after ~60 km of riding the problems are resolved. Unexpected downshifts on the front have stopped, and intentional shifts up from the middle to the big ring work perfectly, though one crank rotation is a bit longer it would take a chainring with multiple pins to shift. I'm okay with that.

4
  • 1
    I see you are not very familiar with this system. The pin lets you know the arm is sufficiently on the spindle to clamp safely. If your pin won't engage, it is likely * in your specific case* that your new spacer is not thick enough to match the original spec .
    – Noise
    Commented Apr 25 at 17:54
  • @Noise correct - I never owned a hollowtech style before. The pin seems relatively useless and when it was set wrong, still didn't do anything. However it was slightly gunked up and the low-mass was insufficient to make it move against resistance. Perhaps it would work more obviously on brand new parts.
    – Criggie
    Commented Apr 25 at 19:17
  • 1
    The pin and it's plastic piece are to prevent the crankarm from falling off if the pinch bolts lose their torque. The pin moving into place after winding in the preloader is also a good indicator that the left arms splines are sufficiently engaged. Regarding the pinch bolts, they should be brought up to 12-14 N-m torque together. I probably switch back and forth about 6-8 times before final torque is reached. Toward the end, if the bolt moves at all prior to the click, I go back to the other bolt to check that the torque stayed stable. If that bolt turns prior to click, I go back
    – Jeff
    Commented Apr 27 at 3:23
  • 1
    Point is, moving one bolt can affect the torque of the other. I use the upper 14N-m as final torque. After a ride or two and with every drivetrain clean, I check the torque on the pinch bolts. In a few instances, I've had to tighten the pinch bolts back to 14 N-m at the first post-ride check. Thereafter, (& usually the case from the beginning) there's been no problems. There's more than a few complaints of people losing the left crankarm of HT2 cranks, but I love the ease of install. I've probably done over two dozen installs just on my bikes since about 2015 when I got my first HT2 cranks
    – Jeff
    Commented Apr 27 at 3:37

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.