Ultegra CS-R800 11 Speed Cassette - Runs Rough

[oldpotatoe]

Originally Posted by NHAero

Of course not. However, you are such a valued contributor to this forum, and yet the constant sniping at anything not made by Campagnolo is frankly tiresome and a misuse of perfectly good electrons.

SORRY, yup, Campagnolo fan-boy thru and thru-Guilty!! But geez, bike stuff, toys...and lotsa electrons..don't think we're gonna run out..

But bikes, not covid or politics or whatever..small beans...

[Krenovian]

I have an offending 27 tooth cog on a 11-34 cassette. Those of you of you that have taken a file to the teeth, what aspect of the teeth or tooth are you filing? Not sure this will work on my cassette as it sounds more like the chain is rubbing on multiple teeth as if the derailleur is out of adjustment. No cable tension to adjust as it is a DI2 system which I’ve attempted to adjust with no luck. Shifting is good, no rubbing on any other cogs.


Sent from my iPad using Tapatalk

[NHAero]

Put the bike in a stand so you can place the chain on the 27T cog. Then slowly advance the chain. What happens is that two of the "tallest" teeth, roughly diametrically apart on the 25T cog, catch the outer edge of the side plates and temporarily lift the chain, then it snaps back into place. Look at the Youtube link in the 5th post in this thread.

If you advance the chain slowly enough you'll see the chain catch. Mark those two teeth on the 25T cog with a Sharpie. Pull the cassette, and gently file an angle on the leading edge of those two teeth. On my cassette, I had enough miles that those teeth actually showed a shiny spot from the rubbing.

You can do a bit of trial and error, you shouldn't need to remove much - I made like half dozen easy passes with the file.

Originally Posted by Krenovian

I have an offending 27 tooth cog on a 11-34 cassette. Those of you of you that have taken a file to the teeth, what aspect of the teeth or tooth are you filing? Not sure this will work on my cassette as it sounds more like the chain is rubbing on multiple teeth as if the derailleur is out of adjustment. No cable tension to adjust as it is a DI2 system which I’ve attempted to adjust with no luck. Shifting is good, no rubbing on any other cogs.


Sent from my iPad using Tapatalk

[steveadore]

Originally Posted by NHAero

Put the bike in a stand so you can place the chain on the 27T cog. Then slowly advance the chain. What happens is that two of the "tallest" teeth, roughly diametrically apart on the 25T cog, catch the outer edge of the side plates and temporarily lift the chain, then it snaps back into place. Look at the Youtube link in the 5th post in this thread.

If you advance the chain slowly enough you'll see the chain catch. Mark those two teeth on the 25T cog with a Sharpie. Pull the cassette, and gently file an angle on the leading edge of those two teeth. On my cassette, I had enough miles that those teeth actually showed a shiny spot from the rubbing.

You can do a bit of trial and error, you shouldn't need to remove much - I made like half dozen easy passes with the file.

Still struggling with the same issue (27T cog clicking due to tall teeth on 25T cog lifting the outer chain plates). Is the filing a foolproof and efficient solution with no tradeoffs (e.g. causing skipping or affecting cassette longevity)?

Btw, when you say that the leading edge of the offending teeth needs to be filed at an angle, do you mean just at the top of the teeth (top 2-3mm)?

[BdaGhisallo]

Shimano have quietly updated the Ultegra cassette and released a CS-R8101 iteration.

Hopefully, this one addresses the issues that have been encountered with the original R8100 cassette.

The part numbers for the 8101 cassette are:

11-30 ICSR810112130

11-34 ICSR810112134

[steveadore]

Originally Posted by BdaGhisallo

Shimano have quietly updated the Ultegra cassette and released a CS-R8101 iteration.

Hopefully, this one addresses the issues that have been encountered with the original R8100 cassette.

The part numbers for the 8101 cassette are:

11-30 ICSR810112130

11-34 ICSR810112134

That's true, but this thread is primarily about the Ultegra level 11 speed cassette (in my case, the HG800), whereas the 8101 is a 12 speed cassette

[BdaGhisallo]

Originally Posted by steveadore

That's true, but this thread is primarily about the Ultegra level 11 speed cassette (in my case, the HG800), whereas the 8101 is a 12 speed cassette

My mistake. I'd never actually heard of any issues with the 11sp cassettes, only the newer 12sp ones.

I should read a little better!

[steveadore]

Originally Posted by BdaGhisallo

My mistake. I'd never actually heard of any issues with the 11sp cassettes, only the newer 12sp ones.

I should read a little better!

I have no experience with 12sp but read about them when I was looking for info about my 11sp cassette. It's the HG800 11-34 cassette and the issue is with the 27T cog large chainring combination (the adjacent 25T cog has a couple of taller teeth that are too close to the 27T and lift the chain at certain positions of a revolution). From what I've read, it's quite a widespread issue, but unfortunately it hasn't been officially acknowledged, so there's no updated cassette

[dddd]

I have a sneaky suspicion that these sort of problems are more related to indexing error, caused by the sharp sort of bends that the cable endures going into the shifter-area routing, especially during installation.

Where many assemblers force the tight loop of cable into the cable housing port only after the entire length of the cable has been fed through the shifter's spool area, the more-correct installation feeds the tip end of the cable into the cable housing initially, so that the tightly-bent loop that gets forcefully stuffed into the cable housing/port is the part of the cable that will later be cut off aft of the rear derailer.
And the cable is always better fed into the shifter than pulled through it, reducing the chance of the cable getting bent/curled (this having to do with the higher tension force caused by "cable-wrap gripping forces" of the sort used to secure ropes around posts in shipping yards).

Curled cable performs poorly in terms of transmitting the shifter's motion to the derailer, as it acts like a spring (adding elasticity to the cabling).

Having a bent cable running through the shifter is a recipe for imprecise indexing, as any permanent bends will not allow the cable to follow the exact same path in all of the gear positions.

I have over the years fixed many mystery indexing faults by using thinner 1.1mm cable, since it's thinner diameter is significantly more easily pulled straight along or adjacent to a curved path, and because it's thinner diameter is significantly less likely to become bent or curled following any given bend radius that it's installation forces it to conform to.

It's also possible that forceful pre-stretching of the installed cable might incur some bend local to one of the shifter's tight-radius guide features, which then flows less than gracefully into the spool area of the shifter as the larger cogs are engaged (again, a thinner 1.1mm cable is less vulnerable to becoming bent following any given radius at any given tension level).

It's possible to modify a cog's teeth to accommodate any slight indexing error, but perhaps better to go to the source of the problem.

[steveadore]

Originally Posted by dddd

I have a sneaky suspicion that these sort of problems are more related to indexing error, caused by the sharp sort of bends that the cable endures going into the shifter-area routing, especially during installation.

Where many assemblers force the tight loop of cable into the cable housing port only after the entire length of the cable has been fed through the shifter's spool area, the more-correct installation feeds the tip end of the cable into the cable housing initially, so that the tightly-bent loop that gets forcefully stuffed into the cable housing/port is the part of the cable that will later be cut off aft of the rear derailer.
And the cable is always better fed into the shifter than pulled through it, reducing the chance of the cable getting bent/curled (this having to do with the higher tension force caused by "cable-wrap gripping forces" of the sort used to secure ropes around posts in shipping yards).

Curled cable performs poorly in terms of transmitting the shifter's motion to the derailer, as it acts like a spring (adding elasticity to the cabling).

Having a bent cable running through the shifter is a recipe for imprecise indexing, as any permanent bends will not allow the cable to follow the exact same path in all of the gear positions.

I have over the years fixed many mystery indexing faults by using thinner 1.1mm cable, since it's thinner diameter is significantly more easily pulled straight along or adjacent to a curved path, and because it's thinner diameter is significantly less likely to become bent or curled following any given bend radius that it's installation forces it to conform to.

It's also possible that forceful pre-stretching of the installed cable might incur some bend local to one of the shifter's tight-radius guide features, which then flows less than gracefully into the spool area of the shifter as the larger cogs are engaged (again, a thinner 1.1mm cable is less vulnerable to becoming bent following any given radius at any given tension level).

It's possible to modify a cog's teeth to accommodate any slight indexing error, but perhaps better to go to the source of the problem.

But how do you explain that indexing is perfect with 10 cogs out of 11, including both the next smaller and bigger cogs? And most people have experienced issues with the one 27T cog. Sure, chain angle has sth. to do with, as the angle of the chain makes it easier for it to catch on some of the taller teeth of the 25T cog, which are too close (both physically in terms of space and also in diameter/tooth count) to the 27T when the chain is on the latter

[charliedid]

Originally Posted by dddd

I have a sneaky suspicion that these sort of problems are more related to indexing error, caused by the sharp sort of bends that the cable endures going into the shifter-area routing, especially during installation.

Where many assemblers force the tight loop of cable into the cable housing port only after the entire length of the cable has been fed through the shifter's spool area, the more-correct installation feeds the tip end of the cable into the cable housing initially, so that the tightly-bent loop that gets forcefully stuffed into the cable housing/port is the part of the cable that will later be cut off aft of the rear derailer.
And the cable is always better fed into the shifter than pulled through it, reducing the chance of the cable getting bent/curled (this having to do with the higher tension force caused by "cable-wrap gripping forces" of the sort used to secure ropes around posts in shipping yards).

Curled cable performs poorly in terms of transmitting the shifter's motion to the derailer, as it acts like a spring (adding elasticity to the cabling).

Having a bent cable running through the shifter is a recipe for imprecise indexing, as any permanent bends will not allow the cable to follow the exact same path in all of the gear positions.

I have over the years fixed many mystery indexing faults by using thinner 1.1mm cable, since it's thinner diameter is significantly more easily pulled straight along or adjacent to a curved path, and because it's thinner diameter is significantly less likely to become bent or curled following any given bend radius that it's installation forces it to conform to.

It's also possible that forceful pre-stretching of the installed cable might incur some bend local to one of the shifter's tight-radius guide features, which then flows less than gracefully into the spool area of the shifter as the larger cogs are engaged (again, a thinner 1.1mm cable is less vulnerable to becoming bent following any given radius at any given tension level).

It's possible to modify a cog's teeth to accommodate any slight indexing error, but perhaps better to go to the source of the problem.

You must be correct, based on the sheer number of words you wrote.

[oldpotatoe]

Originally Posted by charliedid

You must be correct, based on the sheer number of words you wrote.

Yup, good inner wire routing is certainly key but with one cog that's angry on this cassette, probably not inner wire install.

[steveadore]

It has been suggested to me that I should experiment with slightly higher B-screw tension. Theoretically, that might be helpful in keeping the chain further away at the bottom (6 o'clock position) from the cassette and thus prevent the chain from catching on the adjacent cog's tall teeth at pickup.
I'm not convinced, but it seems to have worked for some folks with the same cassette, so I'm willing to give it a try

[NHAero]

This has been asked and answered upthread. I have several of the HG800 cassette in service on three bikes and multiple wheelsets. Look carefully at which teeth of the offending cog lift the chain, and use a file to ease that top leading edge. I have about 10,000 miles on these cassettes and no clicking or fiddling with cable tension. It is IMO a design flaw in the product but easily resolved.

[steveadore]

Originally Posted by NHAero

This has been asked and answered upthread. I have several of the HG800 cassette in service on three bikes and multiple wheelsets. Look carefully at which teeth of the offending cog lift the chain, and use a file to ease that top leading edge. I have about 10,000 miles on these cassettes and no clicking or fiddling with cable tension. It is IMO a design flaw in the product but easily resolved.

That will be my next step probably but I need to get a thin needle file first