When to Replace Gear Cable?

[Ajax Bay]

I don't agree. The bend with the tightest radius is the one where it bends around the pulley on the shifting mechanism. This is much tighter than any of the bends required to run the cable under the bars and it's at this pulley, right next to the nipple, where the cables break.

The cable has to take this bend whether the shifter is 'washing line' or 'under bar tape'.

The shifters that snap cables more often are those with a long lever throw to shift, as these are wrapping the cable around a smaller/tighter pulley. You'll notice that flat bar shifters, which have a really short throw as they are pushed by your thumb (which is comparatively stronger) almost never break cables within the shifter, as the pulley the cable is wrapped around is larger/less tight radius.

My worst shifter by far for snapping cables was Shimano Sora 3400 (really long throw to shift), which was a washing line design.

Washing line vs under bar tape has little to do with snapping cables IMO.

Wrt tight radii and that the cable has to take that repetitive bend on both route options, agree.
Let's leave aside flat bar shifters (think that their radii are less: cable pull is the same (6-9sp) so this is a more likely reason for lower MTBF).
Thanks for your STI-3400 anecdata.
What other STIs have a "really long throw"? Remember at <10sp, the cable pull is the same so "throw" will depend on the length of lever and lever advantage and have the same effect at the little pulley inside the STI.
More accepted (widely shared) anecdata is that cables in 'under the cable' STIs are far more likely to part near the nipple (and a pain to get out to replace).
https://forum.bikeradar.com/discussion/12772533/shimano-sti-5600-gear-cable-snapped-again
@cyberknight has experience.
So have fixed the conclusion below (for me):

Washing line vs under bar tape has a lot to do with parting cables at the nipple IMHO.

 

[cyberknight]

oddly i have sram which are a bit harder to install initially as the cable makes a sharp bend as soon as it enters the shifter but i have never had one break yet whereas i have had it go twice with shimano with no apparant warning .
I know i should probably change them more often but what with work and family i barely get the time where im motivated , have free time or not knackered .Yesterday i needed to put mudgaurds on the bike and after doing a shift at work i had been home 5 hours before i even looked at the box .

 

[faster]

Wrt tight radii and that the cable has to take that repetitive bend on both route options, agree.
Let's leave aside flat bar shifters (think that their radii are less: cable pull is the same (6-9sp) so this is a more likely reason for lower MTBF).

Cable pull is a bit of a moot point here (well it is, and it isn't, as I'll explain later), so I'm not sure why you mentioned it. Cable pull/pull ratio is the amount of cable pull required to move the derailleur a certain distance, and yes, this is fixed for 6-9sp at 1.7mm at the mech for each 1mm of cable. Also fixed (for each speed) is the amount of cable pulled by the shifter for each click. What isn't fixed at all is the amount of cable pulled for each degree the lever sweeps through - shifters with a small pulley will require more degrees of sweep (to get to the next click) and have a longer throw. Shifters with a large pulley will require fewer degrees of sweep and have a shorter throw (like flat bar shifters).

Where cable pull does become relevant is that 11 speed and 'new' 10 speed group sets use a different pull ratio - 1.4mm at the mech for each 1mm of cable. As such, all things being the same, the pulley in the shifter can be made larger, as more cable needs to be pulled. This less tight bend around the pulley reduces the problem of cables snapping in the shifter. It's often said that the change was made to improve shifting, but I'm guessing that Shimano had their eye on cable life too.

Note also that SRAM use either a 1:1 ratio or 1:1.3 ratio - my anecdata, which I'm sure you'll appreciate, says that SRAM suffer less from a cable life perspective for exactly this reason.

Thanks for your STI-3400 anecdata.

That came across as a bit sneery - hopefully I've just taken it the wrong way. We're all just enthusiasts and in most cases, anecdata is the best we have available to back up any theories/guesses we might have, and is always preferable to the 'older stuff is better than newer stuff' assumption which drives much of the opinion on this forum. The forum would be pretty dull without the anecdata.

Fair play though - it's good to see you acknowledge that your side of the argument is all anecdata too!

What other STIs have a "really long throw"? Remember at <10sp, the cable pull is the same so "throw" will depend on the length of lever and lever advantage and have the same effect at the little pulley inside the STI.

I have no idea - why would I? I'm not sure what you're getting at with the next bit tbh. As before cable pull not really relevant, length of lever not really relevant (maybe if I'd have said level sweeps through a wide angle to get to the next click it would have been better than longer throw).

 

[Jenkins]

oddly i have sram which are a bit harder to install initially as the cable makes a sharp bend as soon as it enters the shifter but i have never had one break yet whereas i have had it go twice with shimano with no apparant warning .

This echos my experience. I have had two gear cables break with less than 5 years use on each - one on a 105 shifter and the other on a Claris shifter - while the 3 bikes I have with SRAM are still on the original cables after between 7-9 years of use on each. The bike I had running 105 is now on SRAM Rival so I don't expect any further problems.

 

[cyberknight]

Cable pull is a bit of a moot point here (well it is, and it isn't, as I'll explain later), so I'm not sure why you mentioned it. Cable pull/pull ratio is the amount of cable pull required to move the derailleur a certain distance, and yes, this is fixed for 6-9sp at 1.7mm at the mech for each 1mm of cable. Also fixed (for each speed) is the amount of cable pulled by the shifter for each click. What isn't fixed at all is the amount of cable pulled for each degree the lever sweeps through - shifters with a small pulley will require more degrees of sweep (to get to the next click) and have a longer throw. Shifters with a large pulley will require fewer degrees of sweep and have a shorter throw (like flat bar shifters).

Where cable pull does become relevant is that 11 speed and 'new' 10 speed group sets use a different pull ratio - 1.4mm at the mech for each 1mm of cable. As such, all things being the same, the pulley in the shifter can be made larger, as more cable needs to be pulled. This less tight bend around the pulley reduces the problem of cables snapping in the shifter. It's often said that the change was made to improve shifting, but I'm guessing that Shimano had their eye on cable life too.

Note also that SRAM use either a 1:1 ratio or 1:1.3 ratio - my anecdata, which I'm sure you'll appreciate, says that SRAM suffer less from a cable life perspective for exactly this reason.



That came across as a bit sneery - hopefully I've just taken it the wrong way. We're all just enthusiasts and in most cases, anecdata is the best we have available to back up any theories/guesses we might have, and is always preferable to the 'older stuff is better than newer stuff' assumption which drives much of the opinion on this forum. The forum would be pretty dull without the anecdata.

Fair play though - it's good to see you acknowledge that your side of the argument is all anecdata too!



I have no idea - why would I? I'm not sure what you're getting at with the next bit tbh. As before cable pull not really relevant, length of lever not really relevant (maybe if I'd have said level sweeps through a wide angle to get to the next click it would have been better than longer throw).

its on tiagra 4700 i had the issues that uses 11 speed pull so this goes against your maths

 

[faster]

its on tiagra 4700 i had the issues that uses 11 speed pull so this goes against your maths

Not really.

My post suggests SRAM is easier on cables than new Shimano and new Shimano is easier on cables than old Shimano.

What you're saying is exactly in line with my 'maths'.

 

[Ajax Bay]

Thank you for explaining @faster .
So this is not "really long [lever[] throw" this is the radius of the pulley in the STI the cable is wound (and unwound) round. And the shifters that are failing are those with small pulleys then; the earlier 'post washing line' ones. Ergo (swidt) this is an under bar cable run issue: correlation rather than causation, limited to some of the earlier models 4700 5600 and others? but useful to know.
I called this anecdata but it is an acknowledged issue "accepted (widely shared)" but the quantitative data is unavailable (see also crankset recalls).
2012 thread (with pictures): https://www.cyclechat.net/threads/gear-cable-snapping-in-sti-shifter.101882/
I called your 'worst shifter by far' experience anecdata because it's singular, or maybe the ST-3400 has a reputation for this. But not when I searched therefor. Not using "anecdata" in a sneery sense, rather a mildly humorous and singular sense. Why would the 3400 have a smaller pulley than the 4400 5500 5600 6600 7400 ones (all washing line)?
Short cable pull (newer stuff as in your post above) will require fewer degrees of pulley turn (by lever) for the same pulley radius. So their cable MTBF is less. MTB pulleys are larger so again fewer degrees of pulley turn.
If riders are seeking info on which to base a periodic maintenance regime for RH gear cables, identifying the (cable under bar routing) STIs which suffer from this more would be useful.

 

[T4tomo]

I don't agree. The bend with the tightest radius is the one where it bends around the pulley on the shifting mechanism. This is much tighter than any of the bends required to run the cable under the bars and it's at this pulley, right next to the nipple, where the cables break.

The cable has to take this bend whether the shifter is 'washing line' or 'under bar tape'.

you might not agree, but you're wrong! in the "washing line" yes it bends around the pulley but then comes straight out the side of the shifter

in "under bar tape" it bends around the pulley and the then bends 90 degree in the other plane to exit out the back of the shifter.

 

[Ming the Merciless]

Is this where I say I like my bar ends? Older shifting technology that I’ve never heard break at lever end.

 

[Paulus]

Is this where I say I like my bar ends? Older shifting technology that I’ve never heard break at lever end.

I have never had a cable snap with my down tube shifters either.

 

[roubaixtuesday]

My fixie with a flip flop for two gears is failsafe on the cable front.

 

[Ming the Merciless]

Yep older tech is more reliable when it comes to cables (not) breaking, no doubt about it.

 

[Ajax Bay]

Yep older tech is more reliable when it comes to cables (not) breaking, no doubt about it.

Not sure you're 'on message' with that. The 'older stuff is better than newer stuff' assumption is easily trumped by anecdata.

 

[Mr Celine]

My anecdata shows frequent failures in 5603 Shimano 105. I've had my Roubaix from new in 2009 and quickly learned the necessary get home bodges, either as a three speed or a 10 speed depending on which cable has broken. I can only recall the front cable breaking once, about 2 miles from home so just rode home at 2000 rpm.
I used to replace the cables every winter but gave up as the MTBF seems completely random. Here's a couple of examples -

May 2017

July 2019 -

My winter bike has flat bars and a Shimano 105 9 speed mech. Those cables have never broken at the shifter. The front one frays under the bottom bracket after a few winters at which point I replace them both.

 

[Domus]

Just changed the gear cable on my 11 speed Ultegra as I was changing the bar tape. For info the cable had a definite kink about one inch from the nipple. Been in about 2 years. No idea how much longer it would last or whether the kink has any significance but the cost of a new inner over 2 years is negligible for a certain peace of mind. In my humble opinion that is.