header ad

Why You Shouldn't Use KMC Chains


This topic contains 19 replies, has 9 voices, and was last updated by  Matty F 3 years, 10 months ago.

Viewing 15 posts - 1 through 15 (of 20 total)

  • Author
  • #814514

    Matty F

    Yesterday @jcitizen asked me if I had anything written up anywhere I could send him on why KMC chains should not be used with Shimano (specifically XTR) drivetrains. I didn’t have anything in text, so I decided I would just write something up. The resulting e-mail reply took me about an hour and a half and spanned something like 14 paragraphs. So the two of us figured it might be a good idea to do a little proof read and post it up on RFG. So here we go…

    So we can’t really understand this effectively until we understand the properties of chain dynamics in a drivetrain somewhat. So we’ll start with that.

    All chains made my full drivetrain manufacturers (we’ll say Shimano, Sram and Campagnolo are the only companies doing this at the moment) are system engineered to have shift profiles that match and compliment those of the cassette and chainrings most notably, but also the derailleurs. These profiles ideally allow for the smoothest possible shift with their mating components for the given application. That “given application” part is of some importance, as well. Sram has introduced specific road and mountain chains with their 11 speed drivetrains, and Shimano had road and mountain specific for their 10 speed groups. The reasons for the differences (at least in regards to Shimano drivetrains, which are what I I am most proficient with technically) are due somewhat to the mud shedding properties of each, but also because a mountain chain will be design to maintain smooth power transfer while shifting under heavy loads at lower cadence, possibly multiple gears at a time. Road chains are designed to shift optimally at higher cadences with lower torque, and ultimately wait for a shift ramp in the cassette to come around before transferring cogs.

    Obviously when you’re dealing with managing a chain in this regard the profiles of the chain become extremely important. It is even more important for front shifting, where the chain has to maintain power transfer while being picked up/dropped on to a chainring pin under full torque load before it will simultaneously engage both chainrings (which are a huge 10-15t difference), and then fully transition to the final chainring without losing power transfer. The consequences of losing chain engagement due to improper profiling go far beyond just a sloppy shift, however.

    The most noticeable result of lost chain engagement is catastrophic failure of the chain. This is evident from the number of downhillers that snap chains out of the gate. On race day, these guys sprint out of the gate with wattage peaks well over 1,500w, and are often shifting up multiple gears at the same time. This kind of multiple up-shift is the hardest event for any drivetrain to manage a chain on, as the chain is liable to lose sprocket engagement and jump around somewhat, leaving it liable have a side plate land awkwardly on the sprocket teeth, causing the chain to snap. So when you apply the torque of world class sprinters into the mix, this becomes an issue. It also becomes an issue on e-bikes where chain drive torques are enormous.

    Shimano studied their chain shifts and found that they can eliminate these chain breaks by adding a slight bit of material to chains which prevents them from falling awkwardly onto the cassette in a high torque multiple up-shift. The amount of material was so little I think it only added about 4g to the entire chain. The result was the HG901 chain (rather than the original HG900 chain, which is now discontinued) that allowed them to make their existing drivetrains capable of e-bike use. Sram does have anywhere near the engineering or design potential of Shimano, so they had to make an entire new e-bike drivetrain that uses only eight gears and limits you to only one shift in either direction, because their standard drivetrains are not refined enough to manage e-bike loads.

    On top of complete chain failure, the turbulence caused by shifting with mis-matched shift profiles also accelerates drivetrain wear. Any time a shift feels choppy, hesitates or slips, you are not just getting a poor shift, you are also wearing out the components. I like to use the analogy of a Phillips screw. Imagine you have a Phillips screwdriver and you are trying to remove a tightly fixed Phillips screw. If the screwdriver slips in the fastener, it will damage the screw, and the driver somewhat, as well. If the screwdriver slips too many times, enough material will get eaten away from the screw head that the screwdriver is useless to remove it, and the screwdriver will also have suffered some wear. This is akin to the turbulence caused by a poor shift. However, if you manage to use a fresh enough screwdriver and tactfully remove the screw on the first try, little to no wear will accumulate on either of the components. This is comparable to executing a clean, smooth shift.

    So now we’ve established the benefits and importance of using a chain that matches the rest of the drivetrain, and how particular shift profiling is. So now that we have that established, I ask: How the hell could KMC possibly make a chain that performs optimally on a Shimano drivetrain without actually having any engineering or design documentation from Shimano? Furthermore, KMC claims any of their geared chains are compatible with “Shimano, Campagnolo and SRAM derailleur systems” of the respective number of gears. It is simply unfathomable that anyone could make a chain that performs optimally for all of those systems, both road and mountain, especially without any design documentation at all.

    Now let’s look at why KMC chains should be avoided on Shimano drivetrains strictly from a construction standpoint, and ignoring the inevitable incongruities in shift profiles.

    Shimano’s best chains use a surface treatment called Sil-Tec. This is a blend of nickel and PTFE which they actually plate the chain with. I emphasize the world “plate” because it’s important to acknowledge that it is not a coating. It will not wear off within the suggested lifespan of the chain. The plating is thicker than the amount of material that needs to be worn away to consider a chain “worn out”. The PTFE in this plating gives the chain inherent lubrication and added efficiency for its entire usable life. The higher end chain you get, the more components that get the Sil-Tec plating. The highest end CN-HG901 has the inner plates, outer plates and rollers all plated individually before the chain is assembled..

    KMC chains primarily use Ti Nitride coatings on their higher end chains. This is the gold stuff you can get on $3 drill bits at any hardware store. It is a coating and wears off in time. Their absolute highest end chains get a DLC (diamond like carbon) coating, which also wears off. However, no matter what coating KMC uses, they are all effectively useless from a performance standpoint. KMC applies these coatings after the chain is fully assembled, and as such the coatings don’t actually coat the mating surfaces of the chains – the spots where they would actually be practical. They are only applied for aesthetics. Evidence can be seen in the following photos (look under the screw heads in the DLC photo):

    Ti Nitride:

    Diamond Like Carbon:

    In regards to overall drive efficiency, there a few tests out there measuring different factors. Friction Facts has issued a couple tests which are not available on the internet (unless you buy their test package) that shows Shimano chains to have been the most efficient chain in the tests both with the factory lube left on, as well as when being stripped and having a drip lube applied. KMC was about 10% behind Shimano in each test. In the following test by Wipperman, posted on Cozy Beehive (albeit about five years old now), shows that in regards to loaded chain elongation (energy lost in elastically deforming the chain under heavy torque load), Shimano chains populate the upper tier of those tested, while KMCs largely fill the bottom spots…

    Bicycle Chain Stretch Test & Results: http://cozybeehive.blogspot.ca/2011/05/bicycle-chain-stretch-test-results.html

    Finally, KMC’s best 11 speed chain (XL11-SL DLC) is about 2.5x the price of an XTR chain (yeah, two and a half times more expensive than XTR!). Even KMC’s second tier, X11-SL chain is about 1.5x the price of an XTR. You have to go four steps down the ladder to get a KMC chain that is below an XTR chain in price.

    Comments? Questions? Debate? Have at it.



    Epic write up, Matt. Hard to imagine anyone arguing against this beyond “Gold chains look cooler” or “but KMC was cheaper”

    Always blown away by the effort Shimano puts into their products.



    Great explanation @matty-f Thanks for posting this up. Lots of good info here that isn’t always obvious when troubleshooting a drivetrain issue. I know I have posted in reviews and experienced first hand that ditching the spece’d KMC chain is one of the first things I do if I am having shifting issues on a stock build. Before I even try to troubleshoot anything else, that chain goes in the bin.

    Also if anyone hasn’t seen @matty-f ‘s original chain construction article, you should give it a read. You’ll never look at chains the same way again.

    Chain Construction, Wear and Care



    Never had any issues with a Shimano chain, but the last two I had both failed within a week. Never another KMC chain.

    Awsome write up @matty-f.


    Matty F

    Never had any issues with a Shimano chain, but the last two I had both failed within a week. Never another KMC chain

    I hear ya, Owen. When I was reviewing the Kona Process a couple years back it came with a textbook Shimano drivetrain other than the KMC chain. It snapped on the second ride.

    I do at least 95% of my riding on Shimano chains. I only ride other chains if they are on review/loaner/demo bikes. That remaining 5% of riding accounts for 100% of my broken chains in recent memory.



    My LBS did look dumbfounded when I asked if they had one last year.

    The roadie-related reason: a 6700 chain on a 6700 cogset was always noisy, no matter what I did with it. Drove me nuts.

    The X10 chain on the 6700 cogset was absolutely silent throughout the life of the chain. Shifting and longevity weren’t worse enough for me to care. They didn’t cost that much online.

    When I tried the X10 on a Red Cassette, it did indeed exhibit many of the grievances noted above. The 6700 road bike is history now anyways, so it’s a moot point,


    Renegade Hardware

    Interesting write up. I’m curious on the blend and disposition process of their siltech, but it’s worth noting that TiN is significantly harder than Nickle (The stuff you get on cheap drill bits =/= high end TiN you get on carbide cutting tools), and DLC is also harder and has a friction coefficient similar to Teflon (PTFE), and is much more wear resistant than nickel coatings, and based on the Military standards testing we did for sikorsky (I’m an industry expert on PVD and CVD coatings as part of my former job function), wear results from various factors especially the substrate and it’s hardness. An example is you can take a hardened steel (say D2 at 61Rc), rub it against DLC with hard substrate and the steel imparts itself onto the DLC, the DLC does not scratch, this is why Dies and carbide cutters meant for aluminum cutting use DLC (there are other proprietary coatings for steels, usually TiAlN), pistons, cams in race cars, rotating probe element (5” diameter holding 30,000lbs +) in the EH101, Seahawk, and various Navy helicopters. the list goes on. These are also coatings that bond on a molecular level and are only microns thick, but are for specific tasks that aren’t always ideal for the job (common problem with engineering). It’s also worth noting that DLC is a broad spectrum that covers several types of PVD coatings (amorphous carbon), and some are much better than others depending on application (sputter, cathodic arc, etc) The PVD coating isn’t really the issue, it’s what type is picked, how its applied, where it’s applied (where you mention it’s ignored where it matters most), and what its applied on. Still, for a chain, using a coating meant to wear that includes a friction reducing element, is ideal for this application as you point out.


    Renegade Hardware

    Got the details on their coating. this is what Sil-Tec is.


    Rockwell is about 32-34RC, which is very soft, but can be heat treated up to the low 50s depending on substrate (still much softer than PVD coatings which are closer to diamond hardness). Friction is similar to DLC. The thickness varies depending on process, but it’s also very thin. I believe this produces a wear element for reducing friction, which obviously works very well in a chain application. I’ve got a XT chain coming in soon, which has this. I’ll take the links I break off at work to get a reading on the depth of coating using beta backscatter.

    If anyone wants to do it themselves:


    Renegade Hardware

    oops, here’s the kit: http://www.caswellplating.com/electroplating-anodizing/nickel-plating-kits/ptfe-teflonr-electroless-nickel-kit-3-gal.html

    You can literally do this at home. But it’s benefits will be minimal since it’s not going to coat assembled areas


    Matty F


    Good input. Thanks for posting.

    Your data on TiN and DLC seem correct from my experience. However, as you ultimately concluded, the fact that KMC doesn’t apply those treatments to the friction points of the chain means they’re useless no matter what.

    Sil-Tec is somewhat different than typical electroless nickle/PTFE treatments. It is a proprietary technology, and I believe the difference lies within the application process that allows them to get it deeply bedded within the metal and particularly hard. It was explained to me somewhat by a materials engineer (who is also a shop wrench), but the particulars of it are hazy. The real data won’t ever leave Shimano. They have an entire factory for the Sil-Tec treatment (they use it on a handful of components other than chains). It’s not something you can just do at home with a kit from the internet.


    Renegade Hardware

    Yup, of course its correct 😉 and yea, it doesn’t add much value to the chain when it’s on the outside other than some corrosion resistance in the instance of DLC (no porosity whereas TiN has porosity).

    Interesting on the Sil-Tec. I’m going to take that link and inspect it to see what they are doing. As proprietary as it may seem, someone else has probably done a variation of it in defense or aerospace already. Companies always like to hype their technologies. The truth will lie in the part itself, which fortunately we have. I’m interested in stuff like this, so I can’t wait to get that chain in. Been on Sram chains for as long as I remember but switched to shimano for this XT cassette since the sram 1090 wasn’t shifting very well


    Matty F

    I’d be quite interested to see what you come up with on the Sil-Tec chain! Keep me in the loop. Keep in mind that on an XT chain (CN-HG701) only the inner plates and side plates are treated. Not the rollers (or the pins, but they get Sil-Tec. They a chromized instead).

    I know many companies like to toot their own horns about fancy industry technologies which aren’t really that special across other industries. But you have to give Shimano credit in that they don’t “say” they have any of this. I only know about the exclusivity of this process because of my involvement with them. But I don’t think you’ll find any public documentation claiming this is their own technology. They’re very humble. They have developed all sorts of manufacturing processes that you’d never hear about.


    Renegade Hardware

    worth noting that Shimano calls it a surface treatment. You can’t really impregnate coatings into steel substrate that already exists but rather onto them, and in some instances bond them on a molecular level (PVD), so surface treatment is most likely what it is and the name itself implies it’s a silicone. Nikasil (Mhale TM)is another one that’s somewhat similar (nickel plating with occlusions of silicon carbide, sound familiar?), but has other companies doing the same thing under different names. Just a bit of a process or ratio change and presto: proprietary name. Lots of companies do this, esp in Japan since aerospace is big there (our parent company here is Japanese). Here’s another ‘proprietary’ TM of the same idea http://www.microsurfacecorp.com/nicklon_plus.php and http://www.electroless-nickel-plating.co.uk/products-and-services/ni-ptfe-corrosion-prevention.php

    Worth noting that these industries use their equipment significantly longer than we use chains so I see the lubricity value as the prime reason shimano does this. I’m up to 2 a season now. Wonder how many eagle cassette owners go through trying to keep their cassettes from wearing out too fast…. haha.


    Renegade Hardware

    I’d be quite interested to see what you come up with on the Sil-Tec chain! Keep me in the loop. Keep in mind that on an XT chain (CN-HG701) only the inner plates and side plates are treated. Not the rollers (or the pins, but they get Sil-Tec. They a chromized instead).

    I know many companies like to toot their own horns about fancy industry technologies which aren’t really that special across other industries. But you have to give Shimano credit in that they don’t “say” they have any of this. I only know about the exclusivity of this process because of my involvement with them. But I don’t think you’ll find any public documentation claiming this is their own technology. They’re very humble. They have developed all sorts of manufacturing processes that you’d never hear about.

    that true, they do take that approach.

    what chain has it fully coated? the XTR? Damn, I should have gone that route they weren’t too much $$ either.

    now I’m really curious what they are doing. wonder if they licensed the tech from automotive since cylinder liners in a lot of newer cars are using similar technology. would be pretty neat to find out.



    All fascinating info.

    I have the scarring on my calf to prove that KMC chains suck.

Viewing 15 posts - 1 through 15 (of 20 total)

You must be logged in to reply to this topic.

People Who Like Thisx


People Who viewed ThisX