FBinNY 

It's not that they're somehow defective or inaccurate. It's that they're prone to false highs causing folks to replace chains earlier than they might otherwise.

This won't hurt those selling them, because no actual harm is done. Who's to know how much longer a chain might have lasted, or whether leaving it on longer might have hurt the cassette, or not.

79pmooney

As I understand it, the issue is that checker that only looks a few links is inherently prone to error - unless it independently measures roller wear or can bypass the roller wear. Again, as I understand it, chain elongation comes from "stretch", ie the rivet pins becoming further apart and roller wear, ie the amount the roller can rattle around the pin/supports inside it. We don't care about roller wear because it is the same for all the rollers and the chain simply pulls forward to take up the same slack at all the teeth on the cogs simultaneously. The "stretch" between pins is what matters. The cog teeth wear down; the now narrower teeth being in contact with the further spaced pins. This works but the next new chain won't.

A crude example to demonstrate how this becomes a much greater issue for chain checkers than a long ruler. With a ruler, the rollers never enter into the calcs. It is just pin to pin distance. Your typical chain checker pushes 2 rollers apart and works with that distance. So there is the "stretch" between those X number of pins plus the slop of the two end rollers. A chain could be made that had that slop from day one. It simply wouldn't matter to you, the bike or the shifting. But that chain would flunk from day one.

I believe some chain checkers use methods that do not have this issue. I haven't studied this as it never mattered to me. I just use the ruler. Works. But if I had to get a checker, I'd do that research and fully expect the checkers that do it right to be considerably more expensive. And that's before considering quality. Just, is this designed to get the right answer or just measure the total stretch and calculate what should be acceptable from that?

cyccommute 

The lack of accuracy of chain checkers is highly overrated. I’m not sure how people can screw up a measurement with such a simple tool but, somehow, they do.

The issue of “over-estimating” wear, is something of a red herring as well. Let’s assume that a chain is worn out at 0.75% which is about 3/32” over a 12” chain. Let’s also assume that that happens at 3000 miles as well as assuming that wear is linear (probably isn’t but the math is easier). If the chain checker reads 0.75% at 2500 miles, that’s 500 miles of chain wear you are missing out on.

And how far off are the chain checkers? I’ve done my own tests with several and found them to be spot on, independent of the type of chain checker…short, long, roller isolating or not…when compared to a rule or tape measure. If the checker were off by 10%…which they aren’t…that’s going to be 300 miles too little mileage. If they are off by 1%…which they might…that’s 30 miles. Either one is hardly something to get all hot and bothered about. Chains should be cheap or, more correctly, an expensive chain doesn’t provide for a significant increase in wear mileage. Paying more for a chain is just paying for cosmetics for the most part.

I also have questions about the accuracy of using a rule. I find the photos of people using a 12” rule and then estimating 1/8” and claiming the method to be “more accurate” to measure chain wear highly amusing. You cannot not have an accurate measurement if estimation is involved. If you measure the 1/8” directly, that can be accurate but that depends on other factors as well. Was the chain stretched when the measurement was made or was there slack in the chain? If a tape measure is used what is the variance in the measurement due to the movement of the hook end of the tape measure? The hook can easily move that 1/8” depending on where the hook is in relation to the end of the tape.

_ForceD_

Isn’t a false reading the same as inaccurate?





The one I use, the Park CC-4 is 9 inches long. I.e. it checks the wear over 9 inches(9 links). Using a steel ruler, presumably 12 inches/12 links. Most of the wear check tools I’ve seen are similar, and check the wear of several inches/links, rather than just a couple of links. So is the 3 more inches using a steel ruler going to make that much of a difference?

Dan

JoeTBM 

Thanks, I should have checked and seen that

Camilo

I use a Park CC4 or Pedros Chain Checker Plus for quick on the bike checks. Between my wife and I we have 10 bikes I maintain, including 8, 9, 10 and 12 speed road and MTB. I've switched the 8 and 9 all to 10 speed chains, fwiw, to simplify. I do have one of the smaller, older Park chain tools (3?) but switched to the above versions because the 12 speed AXS flat top chain recommends them. I haven't actually compared the results. Maybe I will.

If the tools show wear, I would take off the chain and hang it from a nail and double check with a steel ruler. But, it's hardly worth the effort - like someone said, the possible over-kill with the chain checkers is trivial in the big picture. SOmetimes I think we over think things.

79pmooney

But - the ruler doesn't even see the rollers. It measures only the pin to pin distance so roller slop never enters in. Only if the check tool has a way of not including roller slop is it anywhere near as good on any chain of roller slop it wasn't designed around.

cyccommute 

The rollers hardly matter. Wear occurs at the pin. Just measuring from pin to pin without taking the elongation of the chain into account will lead to inaccurate results as well. The chain needs to be pulled tight to measure correctly. If the chain is off the bike, hang the chain to measure it or if it is on the bike, it would be best to measure the chain on its bottom run where the derailer spring puts tension on the chain. Or just use a chain checker that isn’t as inaccurate as many would lead us to believe.

_ForceD_

Here’s the thing about comparing a checker tool vs. the steel ruler. On the checker tool (pictured) the difference between OK, 0.5, and 0.75 is just that little notch that I have the nail pointing to. That’s a small measurement. Between making sure the chain is tight, and holding the ruler still, and precisely in the middle (or the edge) of a rivet (and dealing with anything else that could affect the measure like light/shadows, etc)…how can you know that you’re not off by just a smidge, and thereby getting an inaccurate measurement? The chain checker tool takes care of that. — Dan

cyccommute 

Exactly! Any measurement has inaccuracies in the measurement.

mpetry912 

I like this chain indicator. it's longer than the other tools and therefore the span of measured chain is longer, so more accurate, which is why the little short tools don't work. Can't measure chain elongation very well over a span of 2-3 links, the level of precision required is just too great.

Sadly I do not know where I got it or who made it.

/markp

Black wallnut 

The span you measure has zero bearing on the accuracy of your result. The method used to measure has everything to do with accuracy of measurement.

FBinNY 

Sorry, but no.

The amounts involved in chain wear are in the range of 0.001" or so. They wouldn't be measurable with the tools at hand except for cumulative effect. Measuring over greater lengths creates the multiplier needed for the job ie.0.001"/pin x 100 links = 1/10", easily read with the naked eye.

Also, when measuring with a method that conflates roller slop and pin wear, a greater span reduces the effect of roller slop, making the method truer to one for pin wear only.

So, as they say, "Size matters."

Spoonrobot 

Called the Speedtech CW-1089. Pricey for NOS but some cheaper used ones.

https://www.ebay.com/itm/124645390485

Looks neat:

FBinNY 

I sold a decent number of those BITD, for something like $5.00 wholesale.

Maybe I should have hung onto a few.

Black wallnut 

Actually it is .003" or .075mm per link (for aluminum sprockets). If you measure more than one link, or more than 3 links or 10 links or whatever the accuracy of measurement does not change, although if you have uneven wear your chances of finding that increases. The farther the span then naturally the size of the difference between good and replace is greater and perhaps easier to see. I use the Rohloff tool. It is a go/no go gauge that spans 3". The difference between good and replace is quite remarkable.

When I had excellent eyesight I could read a 64th inch scale on a rule. Frank you seem to be claiming that with a rule, which most advocate a 10 inch one, that you are looking for .010" over ten inches. That's a hundredth of an inch! Parallax errors are greater than that. It is the method of measurement that gives one accuracy, not the distance measured. But no one is actually measuring chain wear, tools that act as a go/ no go gauge do not measure, they indicate wear limits. Exceed those limits and you risk damage to your drivetrain such that a new chain no longer fits.

FBinNY 

Mia culpa. I don't know who Frank is, but assume you meant me.

It's a case of typing faster than thinking, compounded by a typo.

I actually meant 0.1" stretch over 10", based 1% stretch. Divided by 20 links that would be 0.005" per pin.

But the point that measuring over more length improves accuracy still holds.

Black wallnut 

A tenth of an inch is much more noticeable. We'll have to agree to disagree on the accuracy statement. 1% is 1% no matter the difference in measuring length. I submit that if you wait until 1% elongation then you will be replacing the rest of your drivetrain much sooner. If you replace the wear item (chain) before it damages your cassette and chain rings then you are better off.

Bill Kapaun

I think the 2 of you are getting accuracy & resolution mixed up.
A longer span will provide greater resolution. That will allow one to READ the results more accurately.

tiger1964 

I need two print that out and tape it to one of the steel rulers I just bought -- better yet, I have an electric engraver somewhere in the workshop. I had bought one of those little measurement tools but looking at a 12" span makes so much more sense.

Recent rides with a friend, his wife came along, she was experiencing a skip/jump under load but "there's nothing wrong with my bike!" Couple of days ago, he and I replaced her chain (I've converted him to waxing)... problem gone.

FBinNY 

No problem agreeing to disagree.

However, I think you might have misinterpreted my point, possibly owing to a poor choice of words.

Yes, 1% is 1%, as 1/2% is 1/2%. However, whatever replace point you prefer, it's easier to measure if you multiply the tiny difference by measuring the accumlated wear over a greater distance.

FWIW I don't buy into the chain gauge arguments. Whatever guideline one chooses, it's just that ----- a guideline. Just as you wouldn't stop and look for service stations if you odometer reads 3,000* miles since your last oil change, I use a broad strokes approach and use measurement to decide that's it's time to consider changing when convenient.

*if that's when you change oil, otherwise pick any number.