Park Tools chain checker and checking chain wear generally

[Cooljase]

Hi all.
Any experts out there on chain wear?
I was under the impression that chains should be changed when they'd 'stretched' by 1%. So I was surprised to notice that when I bought a Park Tools chain checker yesterday it measured not 1% of stretch but 0.75% of stretch.
I just checked up on this, and noticed that the one I've bought is a CC-3.2. It seems to be an updated version of the CC-3. The CC-3 measured 0.75% and 1% of wear. But get this - the CC-3.2 measures 0.5% and 0.75% of wear and NOT 1%.
There's no information about this change on the Park Tools website but it would seem that they've changed the point at which we should chuck out our chains.
It would seem that, effectively, they've been saying chains were ok at 0.75% stretch and now suddenly they're not. Huh? Why not?

My only interest is that I keep my bikes healthy, so I want to chuck out a chain before it begins to wear the cassette and chainrings too fast, but equally I don't want to throw money at new chains unnecessarily.
Can anyone advise on whether chucking out at 0.75% is sensible or not. By the way, the chain I'm worrying about now on my winter bike has nearly 0.75% of stretch and has done 3,368km (2092 miles) in all weathers (but it's been a mild winter here in the UK). That's only 3 months of riding, and a new chain every 3 months is more expensive than I'd like.

Any thoughts appreciated.

Jason in London

 

[tafi]

I do own a park gauge like this but stopped using it when I realised that this method of chain checking (by pushing two rollers apart) is inaccurate. It is innacurate becasue it mistakenly assumes that the play in a roller can be included in the wear (which it should not).

There's a good explanation of this here:

http://pardo.net/bike/pic/fail-004/000.html

And if you search the treads you will find a further dicussion of this issue.

Now I just use a ruler (with imperial inch scale) and measure between pins. 10 pitches = 10 inches on a new chain. If it measures 10.1" then the chain is increased in length by 1% and fully worn. The ruler is not only more accurate, but is far cheaper.

 

[alienator]

I do own a park gauge like this but stopped using it when I realised that this method of chain checking (by pushing two rollers apart) is inaccurate. It is innacurate becasue it mistakenly assumes that the play in a roller can be included in the wear (which it should not).
There's a good explanation of this here:
http://pardo.net/bike/pic/fail-004/000.html
And if you search the treads you will find a further dicussion of this issue.
Now I just use a ruler (with imperial inch scale) and measure between pins. 10 pitches = 10 inches on a new chain. If it measures 10.1" then the chain is increased in length by 1% and fully worn. The ruler is not only more accurate, but is far cheaper.

+1. I measure over 12 full links. If the measurement over those 12 full links is greater than 12 1/8", I put on a new chain. As a bonus, the machinist rule that I use can also be used to measure all sorts of other things, whereas the chain checker is the proverbial one trick pony.

 

[CAMPYBOB]

I usually use the .75-1.0% Park tool as a quick check after I clean the chain. Every once in a great while I will measure the chain with calipers for wear, but that's been a rarity.

Just going from experience with their 10-speed setup, the chains get replaced in the 1200-1500 mile range. Despite tossing chains before the 1% mark was hit, I still burned thru cassette stacks and shorted my chain replacement mileage mark up a bit. Maybe tossing earlier and getting the new Park chain tool would be a good idea now that I've went to 11-speed. Less surface area is going to lead to even faster wear (product advertising shlock be damned).

My older model Park stretch tool measures over four links IIRC. The newer version looks like it measures over six links. I'll spring for it.

FWIW, I averaged a cassette stack about every five chains or so and after 25K-30K I replaced a 53T chainring. The chainring was running along just fine and shifting perfectly, but was worn very sharp and starting ti 'hook' the teeth profile. The cassettes, particularly the 16, 17 and 19 tooth cogs would skip when that fouth, fifth or sixth new chain hit the worn teeth.

 

[Cooljase]

Hey, thanks to everyone who's answered - really helpful.
Yes, I get that the tool is inherently inaccurate (I kind of realised this already but thanks Alienator for the link to the extremely detailed explanation). I just want something that's very quick and easy to use and should give me an approximate view on when to replace the chain for longest possible cassette and chainring life.
I just emailed Park Tool and the guy there said that the change from 1% to 0.75% (ie from the CC-3 to the CC-3.2) reflects today's chains - 9- and 10-speed, he said, could go to 0.75% but 11-speed need to be replaced at 0.5%.
I conclude that, inherent inaccuracy notwithstanding, the CC-3.2 (and similar tools, of which there are many) is still a useful enough addition to the toolbox because it's so quick and easy to use. And my cassette and chainrings will probably thank me for replacing the chain every 3,000km or so, depending on conditions.

 

[CAMPYBOB]

I picked up the 'new' Park CC-3.2 this afternoon.

Comparing it with the older CC-3 tool, it is 2" longer. The old tool is more accurately manufactured, being laser cut judging by the edge pattern. The newer version appears to simply blanked out (stamping die cut). With the sheared edge, there is at least a .003" tear ridge on one edge. Neither version exhibit any sign of the gage surfaces having been benched into tolerance. Good enough as a comparitive tool, I suppose and still probably more accurate than trying to read a scale to .010" while holding the chain strtched and trying to keep the 0-end aligned to a pin.

The marking on the old tool was silk screened ink, which I'm told would rub off...if you just let the tool bounce around in your toolbox. Mine still looks like new after five or six years of use. The new version has all markings stamped into the tool.

The old tool has a satin nickel finish, matte and even. The new CC-3.2 has a brighter, polished electroless nickel finish.

Here's to tossing chains in the trash even sooner! /img/vbsmilies/smilies/smile.gif

 

[davereo]

I use the Nashbar chain checker (I needed to save cash to buy new chains). The Nashbar checker is also blanked with sheared edges. The tool itself is accurate enough. When .75 wear is indicated I purchase my replacement chain so I have it on hand when 1 % is reached.

I have checked my results against a Brown and Sharpe 12" machinist scale using 10 links and measuring 10.1. The 12 links method is only good if you have a scale greater than 12" in length so you don't have to estimate.

Never include the connector link in your measurement and check in more than one place.

 

[swampy1970]

It'll be a cold day in hell before I buy a chain checker. /img/vbsmilies/smilies/biggrin.gif It'll also be overcast with a chance of showers down there too before I do much more than pull the chain up off the chainring a little...

I put about 4,000miles on a chain before it replace it and over the course of more than a couple of years I still have no noticable wear on chainrings and cassette. Maybe I need to push harder than 300 watts on a regular basis and take to riding more in mud and dirt roads to get the type of wear that people report after only 2,000...

... then again I don't really use much in the way of new fangled chain lube goop. The stuff that shimano packs in their chains stays on until mother nature gets rid of it - after that it's oil good ol' oil.

 

[alienator]

The 12 links method is only good if you have a scale greater than 12" in length so you don't have to estimate.

It's weird how that works.

Never include the connector link in your measurement and check in more than one place.

Good points.

 

[tafi]

In reality how you measure your chain wear doesn't matter very much to what the game is about, riding your bike. So if you're happy with how you do it, just ignore my rambling.

But I'm a pedant and I can't resist the temptation to correct a few misconceptions inherent in some of the above posts.

Firstly, talking about whether the body of your chain checker is stamped, forged or CNCed or made from the beak of a Peruvian wood duck has no bearing whatsoever on how useful, how precise or how accurate the tool is. It is interesting that no-one has mentioned how straight the two pins are, or how much slop is in the pivot since these two factors are of far greater importance to the precision.

Secondly, some of us are getting wraped up in the idea that a precision made tool is all that is needed to make accurate measurements. The Park tool could be made from platinum by the Beaureau of Weights and Measures Lab in France for all I care; the measurements made by it will still be innaccurate because the method for which the tool is designed is inherently inaccurate.

Precision is the measure of repeatability or reproducibility (they are subtly different) of measurements. Make multiple measurements of the same thing with precision and there should be very little variability between them. The analogy in target shooting is how well clustered all your shots are on the target. If we use a precise chain checker and make repeat measurements we can get the same reading (or close to it) over an over. This is the definitiion of a precise tool and it has nothing to do with how good it looks. Clever manufacturers will make it by as cheap a method as possible, whilst retaining the principle dimensions.

Accuracy is the ability of the tool and method to measure the real value. In target shooting it is the ability to hit the bullseye. Note that it is quite different from precision. You could be a precise shooter without actually hitting the bullseye. If the tool or method has an inbuilt systematic error in it, then you will have the wrong reading. If the innacurate tool is precise then you'll just keep getting the wrong reading repeated over and over.

The ideal is a tool which is both accurate and precise, these sorts of tools cost a lot in any currency (if you have to ask then you can't afford). But if you have a choice you'd be better off trading a little precision for higher accuracy (the sweet spot where the $2 ruler sits - and NEWSFLASH - you can get ones longer than 12").

In the case of all available chain checker tools (except Shimano TL-CN40 or TL-CN41), the method used has an inherent systematic error, due to the fact that they rely on pushing two rollers apart. If this describes the tool you use, then there is a systematic error in every chain wear measurement you make, no matter how "nice" the tool is. How big this error is depends on the chain and the tool. But the way it works means that every measurement will overestimate the degree of wear. It won't lead to you changing a chain too late, though it could be incresing your maintennace costs (and waste) unnecessarily. In some cases the error can be small, but I have seen enough cases where it is big enough not to trust such a tool (like new Shimano chains which the Park tool thinks is already at 0.5% or new Campag chains where the tool will not even fit).

As I said above, the method you use to measure chain wear will probably not affect the level of enjoyment you feel when riding your bike. The benefit of heeding the above may be small in the long run (though if you really heed my advice you'd already save on the cost of a chain checker).

 

[davereo]

Originally Posted by tafi .

In reality how you measure your chain wear doesn't matter very much to what the game is about, riding your bike. So if you're happy with how you do it, just ignore my rambling.

But I'm a pedant and I can't resist the temptation to correct a few misconceptions inherent in some of the above posts.

Firstly, talking about whether the body of your chain checker is stamped, forged or CNCed or made from the beak of a Peruvian wood duck has no bearing whatsoever on how useful, how precise or how accurate the tool is. It is interesting that no-one has mentioned how straight the two pins are, or how much slop is in the pivot since these two factors are of far greater importance to the precision.

Secondly, some of us are getting wraped up in the idea that a precision made tool is all that is needed to make accurate measurements. The Park tool could be made from platinum by the Beaureau of Weights and Measures Lab in France for all I care; the measurements made by it will still be innaccurate because the method for which the tool is designed is inherently inaccurate.

Precision is the measure of repeatability or reproducibility (they are subtly different) of measurements. Make multiple measurements of the same thing with precision and there should be very little variability between them. The analogy in target shooting is how well clustered all your shots are on the target. If we use a precise chain checker and make repeat measurements we can get the same reading (or close to it) over an over. This is the definitiion of a precise tool and it has nothing to do with how good it looks. Clever manufacturers will make it by as cheap a method as possible, whilst retaining the principle dimensions.

Accuracy is the ability of the tool and method to measure the real value. In target shooting it is the ability to hit the bullseye. Note that it is quite different from precision. You could be a precise shooter without actually hitting the bullseye. If the tool or method has an inbuilt systematic error in it, then you will have the wrong reading. If the innacurate tool is precise then you'll just keep getting the wrong reading repeated over and over.

The ideal is a tool which is both accurate and precise, these sorts of tools cost a lot in any currency (if you have to ask then you can't afford). But if you have a choice you'd be better off trading a little precision for higher accuracy (the sweet spot where the $2 ruler sits - and NEWSFLASH - you can get ones longer than 12").

In the case of all available chain checker tools (except Shimano TL-CN40 or TL-CN41), the method used has an inherent systematic error, due to the fact that they rely on pushing two rollers apart. If this describes the tool you use, then there is a systematic error in every chain wear measurement you make, no matter how "nice" the tool is. How big this error is depends on the chain and the tool. But the way it works means that every measurement will overestimate the degree of wear. It won't lead to you changing a chain too late, though it could be incresing your maintennace costs (and waste) unnecessarily. In some cases the error can be small, but I have seen enough cases where it is big enough not to trust such a tool (like new Shimano chains which the Park tool thinks is already at 0.5% or new Campag chains where the tool will not even fit).

As I said above, the method you use to measure chain wear will probably not affect the level of enjoyment you feel when riding your bike. The benefit of heeding the above may be small in the long run (though if you really heed my advice you'd already save on the cost of a chain checker).

Just for ***** and giggles this afternoon I measured my chain checker tool with my Enco dial caliper. What I found is that the .75 side measured 4.730" and the 1.0 side measured 4.750". I measured the diameter of the rollers on three seperate chain segments 2 Sram 1 nine and 1 ten speed and 1 KMC 10 speed. The rollers measured .300". So by adding .300 to the length of the chain checker the measurement is consitantly the same when the gage is allowed to fit between the rollers 5.03" for the .75 wear and 5.050 for the 1.0 wear. This is an accurate tool that is consistant.

BTW it is flashed chrome and has a hole shaped like a link plate stamped in it. Very cool and it has gone up in price by two dollars since I got mine. That a 35% increase in value. One of my better investments./img/vbsmilies/smilies/wink.gif

 

[alienator]

Just for ***** and giggles this afternoon I measured my chain checker tool with my Enco dial caliper. What I found is that the .75 side measured 4.730" and the 1.0 side measured 4.750". I measured the diameter of the rollers on three seperate chain segments 2 Sram 1 nine and 1 ten speed and 1 KMC 10 speed. The rollers measured .300".  So by adding .300 to the length of the chain checker the measurement is consitantly the same when the gage is allowed to fit between the rollers 5.03" for the .75 wear and 5.050 for the 1.0 wear. This is an accurate tool that is consistant.
 
BTW it is flashed chrome and has a hole shaped like a link plate stamped in it. Very cool and it has gone up in price by two dollars since I got mine. That a 35% increase in value. One of my better investments./img/vbsmilies/smilies/wink.gif

Except it's measuring changes in the pin-to-pin distance and roller wear. It's changes in the pin-to-pin distance that are the problem in chain wear.

 

[CAMPYBOB]

This is an accurate tool that is consistant.

Agreed, Dave. It's plenty good enough to find a ferefnce point to pitch chains (pun? <groan!>) in the trash. The issue for me is forcing myself to trash them even earlier. I believed the 1% stretch limit was acceptable for 10-speed chains. I've come to the conclusion that it is too far to prevent premature cassette tooth wear.




Firstly, talking about whether the body of your chain checker is stamped, forged or CNCed or made from the beak of a Peruvian wood duck has no bearing whatsoever on how useful, how precise or how accurate the tool is.

Actually, it does. The die-punched edge is neither flat, regular or conducive to accurate measurement. Further, it is more prone to wear than a flatter, wider surface. A ground surface is a better gauging surface than a W.E.D.M. surface, which is a better surface than a laser cut surface is a better gauging surface than a milled surface is a better gauging surface than a stamped-coined gauging surface is a better gauging surface than a plain stamped edge. Lastly, as the punch and die wear over a production run and between sharpenings the size of the part produced will vary. Only thou or two...but we are being pedantic, aren't we? Edge form into the die clearance (10% of material thickness or theae abouts) will also vary.

BTW, where can I buy some Peruvian wood duck beak? I want to send some to Campy and have them make me an 11-speed downtube index shifter set of it.

But I'm a pedant and I can't resist the temptation to correct a few misconceptions inherent in some of the above posts.

Familiarize yourself with comparative readings. Pedantic or not, it works just fine. If you can precisely locate the zero end of your steel scale withing .005" to the centerline or edge of your chain rivet while holding it stretched and getting an accurate parallax-free reading and your eyes can actually see or interpolate .005" on a steel scale with .010" divisions...you're a better man than me Charlie Brown. I rate it a three-handed job. It's highly inaccurate, hard to repeat and the precision level sucks. But, if you are concerned about not spending more than $2 (heaven forbid you spent $5,000 on your bike!) it is plenty close enough for government work.

Inside-roller or same-side roller measurment methodology be damned, the commercial Park gauge offers me a more accurate statement of the wear state of the chain than my steel scale. God help me if I stimulated the economy by spending $11.

As an aside, a decent machinist scale from Starrett or Brown & Sharpe or Fowler costs MORE than a chain stretch tool. And friends don't let friends ride...er...buy Mitutoyo!

But the way it works means that every measurement will overestimate the degree of wear.

I only wish that were the case.

Only a few of my chains went the full 1% on my highly inaccurate, low precision, systematically incorrect gauge and I still trashed cassette gears every three, four or five chains. Maybe that's just their service life, but I don't think so. Those damned old clunky Regina Oro 5-speeds used to last multiple seasons with no skipping!

That's another point to tossing them just a few hundred miles earlier...maybe. The new gauge has a .25% reduction in wear span, so maybe it will help save the now even more narrow, more quickly worn (despite Campagnolo's ad copy) 11-speed setup. This season I'll pay close attention to the chain tool and scale readings.

(like new Shimano chains which the Park tool thinks is already at 0.5% or new Campag chains where the tool will not even fit).

WTF? Both of my Park tools fit my 11-speed Campy chain just fine.
Hey, no one can be held responsible for the fact that shimaNO can not manufacture a chain in tolerance! /img/vbsmilies/smilies/duck.gif Both the .75% on my Park CC-3 and .50% on my CC-3.2 tools are a no-go on my (slightly in the case of the 11-speed and 1/2-way to the trash bin in the case of the 10-speed) used Campy 10 and 11 speed chains.

I'll borrow a shimaNO tool and see how well it stacks up against the Park tools.

Anyone know where I can get a discount on buying bulk chain by-the-spool?

 

[dhk2]

When we last had this discussion a year or two ago, I understood and accepted the expressed view of alienator and tafi, that pin-to-pin distance was the real indicator of chain wear, not the inter-roller (floating) dimension as measured by my trusty Park gauge. However, then I argued that the CC-3 was "good enough", and that in my experience the gauge reading correlated with pin-to-pin length as measured by a steel rule or vernier caliper. I knew this because I was used to checking with a steel rule (over 10") before replacing the chain (after the Park gauge dropped in).

All good until my last HG-93 chain, where I was surprised to find the CC-3 would drop into every other roller spacing after only around 2K miles, or half the usual life. After measuring with the steel rule, finding the chain was indeed only about half worn, I continued to use it until the checker dropped into every space, At that time, the rule showed almost 0.1" of stretch, which was back to my typical 4K mile life. The chain obviously had some kind of manufacturing tolerance issue causing the spacing to open up on every other roller. I wouldn't really call it "defective" as it didn't seem to affect the function at all.

After this experience, I consider the CC-3 useful only as a quick indicator of wear, certainly not the definitive measure I once thought it was. I may continue to use it as a quick check, but will always rely on the pin-to-pin measurement before actually replacing the chain.

 

[CAMPYBOB]

All good until my last [COLOR= #ff0000]HG-93 [/COLOR]chain...

Ahah! I've spotted your problem!

If you are getting 4K miles out of your chains...you're doing something right! I've read a couple of articles by Pro Tour mechanics that stated they pitched chains around the 2K kilometer mark. Fear of breaking? Wear? Just because they have a warehouse full of free chains? All three? All I know is that is about the point I hit the limit on the gage. Who knows...maybe they use Park gauges?

I've found the correlation between the Park gage (.75 & 1.0%) and scale measurement to be pretty close on my Campy chains. As tafi pointed out, the Park tool, if anything, indicates more wear than there actually is. Not by much.

When the gauge 'goes', the measurement is saying pretty much the same. And...if anything it would seem I need to pitch my chains sooner rather than later.

 

[dhk2]

CB, I use Prolink oil on the chain every 100-200 miles, wipe it down occaisionally, and rarely if ever clean it. Actually, they go up to 6K when the new chain is paired with a new cassette. I change cassettes every second chain, about 9-10K miles. A well-oiled, glistening black chain is a thing of beauty.

 

[CAMPYBOB]

I'm old school...real old!

I clean weekly or every other week with one the Park plastic snap-on chain cleaners and Simple Green. Lather. Rinse. Repeat.

Now comes the old school...a dose of straight 30-weight non-detergent to start soaking into the pins/bushings followed by a heavy coat of 120 or 145-weight gear oil.

Wipe it down and then wipe the rear wheel off for the next two or three rides! Slings a bit, but parafin was never my bag.

Between cleanings I add a little more gear oil 'maybe' if there's grit on the roads or I've been thru a bunch of dust/dirt. A few swipes of bearing grease if it's going to be a wet race.

Like I've said in other threads, I run 'em wet!

I guess I'm not doing too bad if you're changing cassettes at 9-10K miles. I figure I wasted that Record titanium in 4K miles, but most of the nickel plated Chorus stacks run for 6K to maybe 7K miles. It's only the main running gears that have noticable wear to the point of skipping. Then it's bang the CC for a couple of more chains and bite the bullet for a Chorus or Record stack of cogs...and 10-speed cassettes are already getting tougher to find (another excuse to go to 11-speed!).

I need to go measure the tooth width of an old Regina Oro 5-speed in the 'museum'. Those old hunks of steel seemed to run forever.

I wonder if Rohloff is traded on the DAX? I could always use more stocks.

 

[alienator]

CB, I use Prolink oil on the chain every 100-200 miles, wipe it down occaisionally, and rarely if ever clean it. Actually, they go up to 6K when the new chain is paired with a new cassette. I change cassettes every second chain, about 9-10K miles. A well-oiled, glistening black chain is a thing of beauty.

Yes it is. Today was a chain cleaning.....I'm going back to ProLink from Chain Lube N.5. The chain is a Campy Ultra Narrow 10spd, and with 2950 miles on the chain, it's showing 0.26% wear.

 

[dondon1]

What is all the fu..! compliant sh!t about which chain checker is more accurate than another chain checker and/or ruler. Trust your chain checker tool and just get one and use it. Before you use your chain checker buy your self a new chain, take some measurements with your used chain and new chain and compare the results and make up your mind whether to replace it or not....The point is trust your chain checker maker...stop BS about which method is more accurate....peace out

 

[CAMPYBOB]

Saturday was chain replacement day. I ran a Chorus 11-speed chain to 1850 miles. Despite weekly/bi-weekly cleaning and lubrication it showed 3/16" of stretch when hung alongside the new chain. Damned near half a link.

The new chain bedded into the the casstte gears with no skipping and no wear could be felt thru the pedals.A new cassette stack will be installed this winter when the next chain swap takes place.

The OEM Chorus chain went just under 2K miles with right at 1/8" stretch. The second chain was changed out at 1500 miles PM interval.

I've got the 11-speed chain cleaning/R&R down to about 45 minutes and Campy be damned, I stone the burr resulting from snapping off the guide arbor on the pin after using the rivet setting tool.