I was think the entire cassette. If you look at Shimano, they add in a carrier that isn’t steel to lighten things up.
So probably more what I’m looking for would be a steel single block cassette that would be super rigid and prioritized shifting & longevity over any weight saving “gimmicks” / techniques
I would like this also for softer aluminum freehubs. I have a DT Swiss freehub and all the cassette cogs dig into it and make it super annoying to swap cassettes.
I said 11-23. 
Thanks for the link……that price, though. 🫨🫨
Yeah just the closest I could think of straight away.
Haha yeah definitely pricey. The main reason I haven’t tried it out TBH.
On Shimano DuraAce and XTR cassettes, the middle cogs are made from titanium, the smallest are steel, giving them a distinct three-color look.
Sure, I’m well aware those exist. My reply was in the context of trainer use… and I’d avoid using peak level gear for trainer duty since it is a crazy waste to me. I pick the cheapest cassette that matches the speeds & range I need and save the spendy gear for actual bike use.
Yup, 105/Rival gear works just fine and weight is irrelevant.
In fact, I would hope that manufacturers implement virtual gearing so that we can just have a single, cheap cog in the rear.
Sorry, I didn’t want to come across as nit picky. I know that you know 
Rhetorical question out of some degree of ignorance: Does it really matter what cassette is used if the trainer is in ERG mode?
I mean, okay, I can see the rpm of the flywheel changing with the gear that the physical system is set for, it seemed that lower gears also adjusted to the ERG ‘speed’ quicker, and didn’t appear to ‘float’ over the shorter/quicker changing interval demands. I was always told to use the small ring, and if I/people used TBR (the big ring) it made the simulated watt demand drive the flywheel faster (taller gear?) and the noise of my H2 seemed louder, but to be honest, it was damn loud no matter what I did.
I would imagine that with a virtual flywheel, there isn’t as much of a change as it’s all simulated. (Why would a ‘smarter smart trainer’ insist on simulating TBR when it doesn’t have to)
Years ago when I was riding fer more outside, I was using a 12-25 cassette as there was no ‘hills’ here, just freeway overpasses. Then I moved and they had hills. I’ve tended to leave cassette spacing the same (11-34 on the Roubaix, and 11-40-ish on my Fatboy) and it works well. I adjust the gearing on the ‘demand’, and having some more gears does help, but running out of gears does suck.
I’m just curious on a mechanical level and haven’t really had a good definition/explanation.
Thanks…
Yes it does matter @GPLama made a great video about it
I have old school gearing on my dedicated trainer bike 52:39 12-25. Even though 39-25 is not that low compared to a modern drivetrain, low cadence <80rpm efforts above 250w in ERG mode feel nothing like riding up a steady steep incline outdoors. The lack of inertia is really noticeable, ERG mode spiral of doom just kills me though the deadspots even though I have really smooth application of power from winching my MTB up mountains for decades.
They do. 
Here’s some eye candy: my custom 9-sp 12-21 that I used for a TT race simulation:
What I’ve never seen discussed about using this sort of stuff is, the RD matters, not so much because of the cage length, but also because of the parallelogram angles, i.e how much the guide pulley drops when shifting to easier cogs. The MTB-intended Acera RD worked well down to about a 28t cassette, but with smaller cassettes the chain would not derail or mount properly when shifting to the larger… err, less-small cogs. For this setup I needed to reinstall the bike’s stock 105 RD (intended for a cassette up to 25t), with a shorter cage, which, without chopping the chain, made the granny ring unusable & therefore even more difficult to climb the hill to get home.
Also when shifting chainrings, unless you have a tight spacing like 52/42, you’re dumping 5 or 6 gears to find the same cadence on the next ring.
Longevity? Are people really wearing out cassettes inside? I am not experiencing that at all. Been running the same cassette since 2018 and its still totally fine (and it had plenty of outdoor mile on it prior to that)
First, the most commonly used gear went bust, then the second-most used gear, at which point I got a new cassette.
I swapped freehub bodies once because the splines on that hub were scarred from the cassette. I was shocked that I did that much damage to hub splines. I also have a Hunt wheel set and the hub splines are somewhat damaged too. It has a thin metal strip bolted on to a spline and the rest are aluminum. I guess weight reasons, but if I can chew up the splines, I can’t imagine what more advanced riders could do.
Not exactly cassette erosion/damage, but possibly more expensive.
This is by no means absolute (as FH damage is rather common, unfortunately), but a common cause to that damage is insufficient torque applied to the locking ring when installing the cassette. The system partly relies on proper compression (and resulting friction between the cogs & spacers) to distribute load across more than the single cog. So using a torque wrench to the proper torque (40Nm for Shimano IIRC) is worthwhile. Proper torque will help limit this but even with that step taken, some brute loading can still result in FH scaring.
Yes, here are a few reasons that come to my mind:
And tolerances might have something to do with it, too. I have Shimano XTR hubs on my mountain bike from 2018. I recently took the cassette off (a new one) and the hub was pristine. Just wow. On my previous road bike the freehub was quite marred from the cassette.
Regardless of mode, coasting performance between physical and virtual flywheel trainers is pretty straight forward. Pick a gear of choice (50x17), spin up to a known cadence (90rpm) on a desired virtual road or workout, coast and time the spindown to zero revolution of the trainer axle. Would be interesting to compare at high & low gearing, along with the various modes (ERG, RES, STD, SIM) to see where any differences might exist.
Personally, I generally always reach for a torque wrench when installing a cassette, or based on history, count the number of ‘bumps’ that the ring has as it is tightened. In all the varied hubs that I have, and knowing that the thickness of the metal surrounding the lock ring threads, I’m careful to not over-tighten the ring. But the king of damage I had was in one direction, but it makes sense as coasting isn’t an active use of muscular power. I’ll have to pay more attention when installing cassettes I guess. Thanks…
I admit to not always using a torque wrench either. I definitely have and it is great. But I also like my dedicated cassette wrenches for their basic ease of use vs the torque. Doing it the right way enough times gives me relative confidence I am close, and the fact that I don’t seem to get FH damage is likely part of that along with my relatively middling power figures 
And just for reference, I think better in old units: 40N-m is 30 ft-lb
With my dumb tools, they are right around a foot long so aiming for that 30 lb force is a decent target that I think I hit within reason.
This whole thread is why we need a standard “virtual shifting” trainer protocol. So that whether you are using a dedicated “smart trainer frame” like the Elite Square or your own bike / frame, you use something like the Zwift cog + shifters that would do “virtual shifting” via the trainer.
