Bicycle Mechanics - Failure of mixed-brand cassette - is that a rule?
Bikeforums.net is a forum about nothing but bikes. Our community can help you find information about hard-to-find and localized information like bicycle tours, specialties like where in your area to have your recumbent bike serviced, or what are the best bicycle tires and seats for the activities you use your bike for.
05-12-11, 03:47 AM
I wanted an 8-cog 12-32 cassette, which you can't find any more. Sheldon Brown said, "There is no problem mixing 7-speed or 8-speed flat sprockets into a 9-speed cassette, or vice versa." So I got a new 8-speed Sunrace 11-32, and started looking for sources of cogs to swap in. Found suitable Shimano clusters (an 8-speed for the 12T with its integral spacer, and a 9-speed for the 14T & 16T) in a used parts bin - but looking pretty new. Got the hybrid cassette together, and tried it out. The shift from 18T to 16T - across the boundary - just didn't work. The derailleur balked, and then jumped two cogs on the next shift.
So I tried taking all the cogs, except the 12T, from the 9-speed donor (using 8-speed spacers). And that worked. The puzzle then is that both source sets are viable and shift smoothly - but the mix doesn't work at all.
9-speed cogs are only .02 mm thinner than 8-speed, adding up to less than 1/100 of an inch for the whole set; pretty sure that, as Sheldon says, that's not an issue. For lack of any other possibility I can see, my conclusion is that the Hyperglide ramps are more important than Sheldon says, and the two brands don't synchronize them. If so, worth knowing for the would-be cassette mixer. Thoughts?
05-12-11, 04:05 AM
Cogsets are phased: the rotational offset of the cogs is finely tuned to the horizontal distance and the difference in tooth count so that the chain fits across the gap.
The phase has to relative to other cogs but also relative to the mounting splines.
If you swap out one cog for another of the same tooth count but a different phase (relative to the splines) it wont work.
Is the 9 spd integral spacers too? I would think you need the same spacer as the 7-8 spd, but I have never tried it.
05-12-11, 10:28 AM
Jeepr - cassettes normally use separate spacers (aluminum or plastic) for most cogs, but machine the spacer into the steel of the smallest one or two. That's why I needed to find an 8-speed source for my 12T. So all the spacers I used were indeed 8-speed spacers.
Michael - whether it's phase as you say or Hyperglide shaping as I'd surmised, there certainly seems to be something which makes amateur attempts to assemble cassettes dicier than Sheldon suggests. This is pretty unequivocal: "Building Custom Cassettes: Shimano wants you to use one of its standard combinations... but you don't have to if you don't want to! It is not difficult to customize Shimano cassettes. If you substitute an un-approved cog, Uniglide or Hyperglide, it will still work, but the shift to/from that cog will probably not be as smooth as a Hyperglide shift normally is. Since people managed without Hyperglide for several decades, this shouldn't scare you off." ( http://sheldonbrown.com/k7.html) Also puzzling, if phase is so critical, that Uniglide cassettes - where there were 18 different ways to mount any cog, 9 rotational positions and then another 9 after you flipped the whole cog around ("sprockets were reversible, so that if you wore out one side, you could flip them over and the other side was just like new!") - worked. One more quote, to verify that Hyperglide cogs, according to Sheldon, will work despite the removal of phase constraints: "you can fit Hyperglide sprockets onto a Uniglide body by grinding or filing off the one wide spline." Thus restoring 9-position freedom.
Looking carefully at Sheldon's words, he does say "Shimano cassettes". So I see two possibilities: either brand purity is critical, due perhaps to brand-wide phase conventions (and Sheldon meant this, but doesn't make it very clear), or Sheldon is wrong when he says - categorically - "it will work." I don't have a position, or enough experience to even have a hunch. (I did end up with a working cassette with cogs from two source cassettes - and it's all-Shimano.) I would find either answer surprising, interesting - and useful to know.
Hyperglide ramps are more important than Sheldon says
This. The late great Sheldon was indeed a guru, but he failed to get Hyperglide - it makes more difference than he figured, and I'd emphatically recommend against ignoring which cogs belong next to which.
"It will still work" seems pretty optimistic from my experience - Sheldon seems to have forgotten that pre-Hyperglide shifting sucked.
05-13-11, 04:18 AM
Done mixed cassette before. Initial results were as yours, with bad shifting; then redid the derailier cabling and recalibrated shifter, etc. Worked OK after that, not as fast shifting as normal cassette, but didnt skip cogs at least.
So my advice is to check the derailer/shifter and make sure its absolutely perfect cause even the most minor flaw is going to combine with the non-hyperglide cassete and mess things up. check the derailer hanger alignment, make sure the cable is perfectly lubricated, is the last 'u-turn' loop of housing the right size, etc...
If all else fails, switch to a non-indexed friction lever as your shifter. manual trimming solves everything, and really it was hyperglide that made indexed shifters possible so...implication is that if you mix your cassette and turn it non-hyperglide, you'd likely need to also revert to non-indexed shifting.
05-19-11, 09:46 PM
Thanks, Kimmo & Xenologer.
To report the final outcome: maybe I'm just lucky, but putting cogs with neighbours they were not meant for (but of the same brand) has worked out for me.
My Frankenstein cluster of 7 used Shimano 9-speed cogs (all from one cassette and in their original order), 6 8-speed spacers, and a new Sunrace 8-speed 12T was working fine. But getting a new chain prompted me to take a closer look at the used cogs. They proved to be visibly worn, and I had replacements available in two Sunrace cassettes, so I decided to try out an all-new, all-Sunrace mix.
It is working well, even though the three smallest cogs do not have their original neighbours. Here are the source clusters, with the cogs I'm using in bold:
Weirdly, the only shift which is noticeably balky is from 18 to 16 - although those cogs were neighbours in their source cassette. The 16-14 and 14-12 are fine.
The moral, I guess, being that it may be worth trying out mixed cassettes, with the question of whether sticking to one brand helps left open.
05-20-11, 12:33 PM
There's no hard rules about which combinations will work well. I've just learned to try it out and see what happens. I've now got a 12-28 10-speed cassette made from 3 cassettes of two brands which shifts extremely well, but other more subtle changes I've made to some cassettes have caused shifting problems.
Hmm, so it's more a case of 'suck it and see'...
Don in Austin
05-21-11, 08:33 AM
Whu don't you just buy this: http://compare.ebay.com/like/230531938848?var=lv<yp=AllFixedPriceItemTypes&var=sbar&_lwgsi=y ?
Don in Austin
05-21-11, 09:04 AM
I built a frankencassette years ago out of two 7 speed hyperglide cassettes, and I had similar results - barely acceptable shifting, but the ability to hit all cogs when needed.
I have ridden many turn overs of cassettes put together on my own. As wear progresses, I replace only individual cogs. I found it beneficial to stock spacers of different thickness that allow to compensate for variations in characteristics of cogs. There is a guy on Ebay who makes cheaply custom spacers, of any thickness you desire. Otherwise, I found than an Sram ESP rear derailleur is more tolerant to variations in the cog spacing, than a derailleur following the Shimano standard.
05-21-11, 11:35 AM
Anti f- worry not. There is a solution.
Ive been doing this for the last year and a half. Partly because there are such limited options if you want an 8 speed cogset these days.
But 12-32, no problem. Buy a sram or shimano 11-32 nine speed cassette. Remove the screw (sram) or drill out the three rivets (shimano) that hold the cassette together. Chuck the spacers and replace them with your eight speed spacers. Discard or save the 11t first position cog.
The shimano 11-32 (now 12-32) is probably the better to work with. The
second (now first) position twelve tooth cog also has the lockring ratchet marks around its inner circumfrence and all you have to do is put on a twelve tooth lockring ( which is slightly larger than the one that comes with the cassette- that one wont work) and go. The second position (now first) 12t cog has a built in spacer that is about .010 inch narrower than an authentic 8 speed 12 t cog, but it is not likely to make any problems. Just adjust the lower limit screw a bit.
If im feeling really picky I will make an extra .010 spacer from a layer or two of waterproof sandpaper, superglue it to the back of the 12t, and trim it with a razor, but its more trouble than its worth, and not really necessary.
A shimano hg 50, or a sram pg 950 or 970. The srams are done the same way, but though its easier to get the cassette apart, you have to do a little engineering if you want the lockring to work correctly. Shimano 12t second position nine speed cogs are multi-use, if you will: they are also made to be used as first position cogs, so all you have to do there is make sure you have the correct size lockring.
05-22-11, 11:15 AM
Don: I tried to get that cassette through my LBS - distributor said no longer available. (SRAM also makes a 8-speed 12-32, but it weirdly jumps: 18-20-24.) As for Ebay - some bad experiences with cross-border shipping have left me unwilling to try again.
Gerry - bull'e eye; that's what I wanted to hear. The 9-speed 11-32 looks like a new standard, so I was already planning to rely on respacing it - with a foggy patch in my plans about the 12T. Hearing what you've done is great. (Thanks for the PM - can't reply in that channel because I don't have 50 posts.)
Powered by vBulletin® Version 4.1.12 Copyright © 2014 vBulletin Solutions, Inc. All rights reserved.