When to replace a cassette/chainring

[Keith A]

I wanted to get some input on when you should stop throwing new chains at a drive train that has a well used cassette and chainrings. The amount of wear on a chain is easy to measure, but this isn't the case with the cassette and chainrings. I just picked up a Park Chain Checker so I can more accurately measure the chain wear rather than measuring this with a ruler. I know the chain on my CSi is due to be replaced, but am thinking it my be time for a new cassette and chainrings as well.

Thanks in advance for your input.

[Kevan]

relies on sloppy shifting, noisy drivetrain, and slipping off of cogs to be my signal. Rings is an every 3-4 year thing, based on my amount of erosion (sharkfin).

[Keith A]

Only one knucklehead response? (Thanks Kevan) Come on guys/gals let me know how you decide to replace this items.

[vaxn8r]

Noise. Divetrain noise, usually in the cogs that get the most use. For me it's always the 11-13 cogs which go first ;)

I'll usually switch out the chain and adjust everything and if the noise doesn't clear up it's time for a new casette.

[Dave]

None of the common chain wear measuring tools, including the Park chain checker produces an accurate measurement of chain elongation (change in pitch) because they add roller wear to the measurement and most measure over a very short length. The roller wear can be as large as the elongation, so the tool may report twice the actual elongation.

The best way to measure elongation is with a 12" scale. Place an accurate 12” scale on the edge of a pin. The pin at the opposite end will be totally covered when the chain is new. As the chain wears, this pin will begin to “peak out” from under the scale. Change the chain before ½ of this pin is exposed. The maximum allowable wear is 1/16” (.063”) per foot. One half of a pin is slightly more (.070 inch). Ho much easier could it be?

Elongation is only half of the chain wear issue. The rollers also wear - often much more than the pins (which cause elongation). New rollers will measure about .205 inch in between them. When this distance increases to .235-.240, I consider the chain to be shot. It is possible to have this much roller wear and very little elongation. I've had a chain with 6000 miles of use, only 1/4 of the allowable elongation, but with shot rollers.

Changing a chain long before either of these wear criteria is met is not likely to increase cog life. It's entirely possible to wear out at least one or two cogs over the life of a single chain. I've used a single chain for 6000 miles, and worn out one cog, even though the chain showed little elongation. The only practical way to detect a worn cog is by installing a new chain. If the chain skips on a cog, while pedaling under a heavy load, then the cog is too worn to use with a new chain. If you install a new chi and don;t get skipping with any of the cogs, there is no reason to change the cassette.

The best way to maximize cog life is to alternate the use of 2-3 chains, changing every 1-2000 miles, or at leaast before the first chain is half worn. With this method a new chain will never be installed on worn cogs. When both chains are worn out, then most likely the cassette will be too. There might be a valid argument for including a third chain, but it all depends on the cost of the chain relative to the cost of the cassette. For example, if chains cost $40, the cassette cost $120, and chain life is 4000 miles, the cassette could be used for 8000 miles. The cost per mile is 2.5 cents. If a third chain was included in the rotation, the cassette might be useable for 12,000 miles and the cost drops to 2 cents per mile.

[palincss]

Changing a chain long before either of these wear criteria is met is not likely to increase cog life. It's entirely possible to wear out at least one or two cogs over the life of a single chain. I've used a single chain for 6000 miles, and worn out one cog, even though the chain showed little elongation. The only practical way to detect a worn cog is by installing a new chain. If the chain skips on a cog, while pedaling under a heavy load, then the cog is too worn to use with a new chain. If you install a new chi and don;t get skipping with any of the cogs, there is no reason to change the cassette.

The best way to maximize cog life is to alternate the use of 2-3 chains, changing every 1-2000 miles, or at leaast before the first chain is half worn. With this method a new chain will never be installed on worn cogs. When both chains are worn out, then most likely the cassette will be too. There might be a valid argument for including a third chain, but it all depends on the cost of the chain relative to the cost of the cassette. For example, if chains cost $40, the cassette cost $120, and chain life is 4000 miles, the cassette could be used for 8000 miles. The cost per mile is 2.5 cents. If a third chain was included in the rotation, the cassette might be useable for 12,000 miles and the cost drops to 2 cents per mile.

When you talk about "a worn cog" do you mean a tooth, or do you mean a sprocket? A "cog" literally means one tooth, but cyclists typically don't use it that way except on rec.bicycles.tech.

[flydhest]

When you talk about "a worn cog" do you mean a tooth, or do you mean a sprocket? A "cog" literally means one tooth, but cyclists typically don't use it that way except on rec.bicycles.tech.

Everyone remember the Jetsons? George worked for Spacely Sprockets and their main rival was Cogswell Cogs.

[RPS]

Over the years I went from the objective to the subjective camp when it comes to chain/cassette replacements. I use to measure the chain but no longer bother since I replace the cassette and chain together, and I never let it get noisy or to the point of skipping or other problems. For me this is made possible because I mostly ride bikes with Ultegra 9-speed and can buy the chain and cassette as a combo for about $60, and they generally last over 10,000 miles riding in the flats. If I were spending three or four times as much, I’d take another approach.

I probably ride 90% of the time in the three middle sprockets (14-15-16) of an 11-21 cassette, so I wear these the most by far. As the chain stretches, I start to notice a lack of smoothness when riding the like-new sprockets at each end of the cassette. At that point I replace the chain and the cassette. I’ve looked into replacing the middle sprockets only but it’s not worth the hassle. For the chainrings (mainly wear the middle 42T) I use the old lift-the-chain-off-the-chainring test.

[Keith A]

Thanks for all of the good replies so far!For the chainrings (mainly wear the middle 42T) I use the old lift-the-chain-off-the-chainring test.This is a new one time, care to elaborate?

[Ahneida Ride]

Everyone remember the Jetsons? George worked for Spacely Sprockets and their main rival was Cogswell Cogs.

Flyman ...

Leroy was Dogs name right? or was it the son's name ? GEE ? :eek:
Marge was the Robot maid. am I right ?

[gasman]

Astro was the dog and Elroy was the son.
Daughter Judy, Jane his wife... dah, dah,dah dot


I watched the Jetsons' a lot as a kid.

Oh yea-the housemaid robot was Rosie.

[wrench]

though I will add... the rate a chain wears is contigent upon a number of variables, such as brand, terrian ridden, rider weight and also riding style, climbing hills in your 53 for example will certainly wear your chain faster than spinning in a flat. So be vigiant and change your chain accordingly.

[Samster]

i just get a new bike every 6000 miles. it's more fun than rotating chains.

[Dave]

When you talk about "a worn cog" do you mean a tooth, or do you mean a sprocket? A "cog" literally means one tooth, but cyclists typically don't use it that way except on rec.bicycles.tech.

You may be right, but I see the term "12 tooth cog" and many others used all the time in the literature I read and rarely see any reference to sprocket. I say it doesn't matter because a skipping chain on a "sprocket" means it's shot.

Here's Sheldon Browns explanation of how the term cog came into use:

http://sheldonbrown.com/gloss_cn-z.html#cog

http://www.branfordbike.com/cgi-bin/perlshop/perlshop.cgi?ACTION=push&thispage=cassette/cog15.html&ORDER_ID=154463510

[RPS]

Thanks for all of the good replies so far!This is a new one time, care to elaborate?Keith, although it is not recommended, some riders check for chain wear using the quick-and-dirty (literally quick and dirty) test of lifting the chain off the chainring by pulling it forward at the front of the bike. The greater the amount of wear, the more you can pull the chain away from the chainring. With a new chain and a new chainring it won’t lift as much.

As stated above, I don’t use this technique to gauge chain wear, but when replacing the chain and cassette with new ones, I pull on the new chain to see how much it moves on the existing chainring and use it as feedback. If it lifts too much (very subjective) I go ahead and replace the chainring which normally cost me about $20.

You can easily check chainrings visually by inspecting the shape of the teeth, provided you have good sight. As the chainring wears, the teeth will look like waves. This happens because the chain is always being pulled by the same side of the teeth; hence the teeth wear only on one side (some of us who ride tandems switch the two timing chainrings front-to-back to extend their life since the sprockets work on the other side of the teeth when the driven chainring becomes the driver and vise versa).

[BdaGhisallo]

http://cycle.shimano-eu.com/publish/content/cycle/seh/nl/en/technical_service/faq_s/general_faq_s/how_can_check_if_my.html

How about this chain wear tool from Shimano? It looks like they have designed it such that roller wear is eliminated and only chain stretch is measured.

What do you guys think? It's pricey at about $90 but if it is a much better tool it may be worth it.

Also, is it possible that a chain needs to be replaced even when it shows no stretch? Say someone rode on one chain for five years, but never exceeded 50watts of power output, and never rode it in bad weather. Is it likely that it still needs replacing if it shows no measurable stretch.

[Dave]

http://cycle.shimano-eu.com/publish/content/cycle/seh/nl/en/technical_service/faq_s/general_faq_s/how_can_check_if_my.html

How about this chain wear tool from Shimano? It looks like they have designed it such that roller wear is eliminated and only chain stretch is measured.

What do you guys think? It's pricey at about $90 but if it is a much better tool it may be worth it.

Also, is it possible that a chain needs to be replaced even when it shows no stretch? Say someone rode on one chain for five years, but never exceeded 50watts of power output, and never rode it in bad weather. Is it likely that it still needs replacing if it shows no measurable stretch.

As I already noted, its is entirely possible to have extremely worn rollers with 20-30 times the amount of wear as the pins and bushings, yet show a fraction of the allowable elongation. In theory, some of my chains would have gone 20,000 miles or more before reaching 1/16 inch per foot elongation, but the rollers would have disintegrated long before that.

The two tools I use are a precision scale and a plug gage made from a 6mm hex wrench, ground down to about .070 thick, so it fits between the rollers.

http://www.mansontool.com/cart/shopdisplayproducts.asp?id=44&cat=RULES

[Kevan]

http://www.3dmd.net/gallery/albums/textures/walls/brick/BrownBrickWall_tileable.jpg

[Dan Le foot]

I run the chain and cassette until it doesn't run smoothly.. Usually 10-12K miles.
I replace with new chain and cassette. Never had to replace a chain ring yet.
Dan

[GregLR]

Rings is an every 3-4 year thing, based on my amount of erosion (sharkfin).
I recognise when a chainring becomes too worn by an unusual drivetrain noise after fitting a new chain (I use a Rohloff 'Caliber 2' tool to measure chain wear, though I take Dave's points about the shortcomings of such tools). It's very hard to describe but is like a rhythmic 'swooshing' noise. I found what was causing this noise by trial and error - fitting a new chainring eliminated it.

I've just looked at two such worn chainrings (a TA Specialites 38, & a Campagnolo 39) that I have hanging above my workbench, and compared the pattern with a new, unused ring. The worn ones exhibit a similar pattern of erosion, with a wider and deeper 'valley' between each tooth, and hence generally narrower teeth than a new ring.

Greg