msu2001la

I'm running a Stages crank-arm power meter and there is some lag in Zwift registering power changes. I'd say it's in the neighborhood of 2 or 3 seconds delay, but I've not really found it to be a problem. I mostly notice it when I'm just starting to pedal, and it takes a few rotations of the crank before the avatar starts pedaling. Same with coasting, where I'll stop and the avatar will continue pedaling for a few seconds before coasting. When I'm out riding and start to hammer on a sprint or something, it's responding as quickly as I'd expect.

When the grade changes, the trainer instantly responds by adding/reducing resistance. It also seems to work fine in workout/ERG mode, but also I'm not sure I'd even notice a few seconds difference. The group rides I've done have all been pretty large packs of riders, so I've not noticed that there's any issue in staying with the pack. I do notice that riding with smaller groups requires a bit more attention to my pedaling effort to make sure I'm not surging ahead or dropping back, but I've assumed this is related to other factors (like differences in trainer difficulty setting, where the uphills hit some riders harder than others, therefore they instinctively pedal harder).

I haven't tried running the power directly from the Kickr Snap yet, but I'll play around with that at some point and see if I notice any difference.

The price difference between the Snap and Core is still pretty huge, IMO. $500 vs $900 in the US market. That $400 cost difference is less than what I spent on my crank power meter.

Het Volk

Seems to me, that one could, through a combination of cheap spin bike, some pedal base power meters (Garmin Vector), and a speed and cadence sensor, you could make an at-home smart trainer that does not require a bike. However, that seems like a real hassle, but, if one already has a spin bike they like, it is an option. I think the speed would be the toughest sensor to add to a spin bike unless spin bike already has it.

msu2001la

I'm confused as to why you're telling me this, but Zwift doesn't need a speed sensor if you have a power meter, and I'm pretty sure those Garmin Vector pedals will act as a cadence sensor. Your spin bike would still be "dumb" though... you'd have to manually control the resistance.

caloso

I think the confusion is over what people mean when they say a trainer is "dumb" or "smart." I have a Kinetic Road Machine, which Kinetic calls "smart" in that it has a little pod that transmits BT power data. But I think most people (myself included) would call it a dumb trainer because it just reports power data similar to when I ride my PM-equipped bike on my old Nashbar fluid trainer. I think when people say "smart trainer," they mean one where the resistance is controllable in response to a game or power training software.

Het Volk

Good point on the definition between dumb and smart. I was thinking that for purposes of using Zwift or other apps in a fashion where you could at the very least, control your avatar, there is a way to kind of bolt on some form of tool to allow you to train aside from listening to music or watching TV. Since I prefer mostly the ERG mode during the winter, my mind was focused mostly on that approach with Zwift.

What is going to be amazing to watch, is that in the next 5 years, almost ALL trainers or spin bikes will be smart outside of Peloton or others which are so class dependent. I do not see a home market for spin bikes or trainers without smart capabilities. I suppose, for race warm-up, there will always be some basic wheel-on or direct-drive trainer that is purely just for warm-up and for nothing else as a cheap way to warm-up for races.

caloso

Probably. When I was shopping for a trainer last winter, I really wanted a smart trainer but I expected that most of the time I'd be using it to warm up for race. Didn't really expect to spend so much time on it in my garage.

DaveLeeNC

I have a decent spinner bike (Lemond Revmaster Sport) that is paired with Garmin Vector pedals. I have done some Zwifting with this combo and it works just fine as far as it goes. It takes power input from my pedals and calculates speed from that data (plus my profile info plus grade/etc). But it is a dumb trainer in the sense that if I am going up a 12% grade and I want it to feel like a 12% grade I have to tweak the resistance knob to do that. Like most spin bikes the resistance is a oiled leather strap against a great/big flywheel. It provides a relatively true 'road feel' that is smooth and noiseless, but even small tweaks to resistance kind of take a while to settle out. It is workable, way better than having Zwift guess your power based on modeling of a traditional dumb wheel-on trainer, but is not particularly satisfying compared to what I THINK would be the experience of a true smart trainer (particularly if it had good power measuring capability).

In particular you really cannot do shorter HIIT type stuff on my setup. You give the resistance a big twist, even God doesn't know what resistance level you are going to end up with, so you start sprinting with one hand on the knob and twist again, it takes a few seconds for the resistance to kind of settle out, maybe you are where you want to be and maybe not - but the sprint is close to over before you ever get there.

It works pretty well for my purposes but I have dropped all subscription stuff and either stream entertainment on my tablet computer while putting in 'some miles' (and in that mode this setup is quite effective) or for more intense work sessions (intervals of a minute or more) it is typically just me and the spinner bike (and bike computer).

It is interesting that (assuming that the Covid status does not preclude this), riding in this mode is going to be crucial to my training for the Six Gap Century this year. IMHO the cramping issues that I fought in the 2019 Six Gap were simply because I had never trained doing 30+ minute hard/constant efforts. And that is somewhere between hard and impossible to do here in the rolling terrain of south-central NC (USA). So for this year I am going to simulate some long climbs on my spinner bike (probably at the end of some outdoor training rides). So for me (in this case) indoor riding is 'superior' to outdoor riding.

dave

msu2001la

I see what you're saying now. Yes, there are ways you can get on Zwift now without a smart trainer or power meter. Zwift has a list of compatible "dumb trainers" where you can use a speed and cadence sensor and allow Zwift to estimate power based on those readings. Theoretically, Zwift knows the power curve for the specific trainer. I've tried this in the past and it wasn't very accurate. It might be possible to do this with a spin/stationary bike (even if Zwift doesn't include it as a list of approved trainers) and I've seen some videos where people attach speed sensors to the flywheels of spin bikes to get speed/distance info, but again... the accuracy would likely be way off. I suppose if your goal is just to have an app that tracks rides and provides a visual, it might be enough. It was pretty worthless for trying to do structured workouts based on power output though.

I agree that we will start seeing "smart" features trickling down to even cheaper trainers. Already the Tacx Flow ($370) does everything needed to fully enjoy Zwift and other apps. It provides power output and the apps can control the resistance. It's just a matter of time before there's a sub $300 smart trainer that does the same. That said, I think there will still be a market for ultra-cheap bare-bones indoor trainers for people who have no interest in Zwift/Rouvy and just want to ride, manually control resistance, etc.

Het Volk

As for a cheap, pre-race / event warm-up machine, it would be nice if they could put out a cheap, direct-drive trainer that one could just throw the bike on and spin away. I suppose rollers offer that now....but getting onto rollers in parking lot is not a lot of fun.....

Hoping we see one so that one can have something that if stolen at an event, you would not be completely out-of-pocket.

jadocs

Smart rollers are where it is at. Not very many people use them because they are only familiar with the old tech. They have come a long way (InsideRide and TruTrainer - I would not consider anything else...ie. Elite rollers pale in comparison).

Got to ask though....why would you want to use "smart" features in a parking lot?

If you want to run zwift off your phone outside and are worried with powering up the rollers....TruTrainer smart load drums do not require a power source. It generates it's own power.

Het Volk

i don’t want to run smart features in a parking lot. I just want a cheap, dumb direct-drive trainer that I would not worry about getting stolen. I just loathe wheel-on trainers.

jadocs

Why don't you just use rollers? If you are worried about getting on, just use them at the side of your vehicle/wall....or even place a bike stand next to it.... for support until you get going. Bike stand actually works really well, just spread the legs....don't raise it, put it next to the rollers and use it for support. As you are aware, you can get cheap rollers that fold and take up even less space. A lot more convenient IMO than having to take your wheel off and on.... and getting grease all over your hands.

msu2001la

Smart trainers need power, which is usually at a premium at most race venues. I've never seen anyone warming up on a direct drive trainer at an amateur race, though pro teams use them.

Most teams around here prioritize light, portable and cheap... so they are just using dumb wheel-on fluid trainers or rollers in their tents because it's a lot easier for multiple bikes to be swapped in/out. I also see people warming up on these:

GeHu

I know very old thread but being a retired electronic engineer and cyclist including indoor thought I would drop my 2 cents. Tried 2 wheel off direct drive from Elite (Suito) a couple of them and a Wahoo Kickr Core (way better customer support by miles. Stay far way from Italian Brand Elite). Then a wheel on much cheaper Saris M2 smart wheel on trainer. The only one of all that got it right (or at least even close.) From what I understand of the firmware designs a direct drive trainer's firmware is way more complicated and thus way more capable of screwing up. The wheel on trainer at least is providing the centrifuge statistics naturally from a real rear wheel moving into the algorithms required. Ultimately then of course, I went with a good ole wheel on trainer and it has been problem free ever since (other than the classic things from back in the day with a wheel on trainer such as wearing tire, tension etc.). I will take that in a new york minute compared to the unrealistic effects a wheel off trainer ends up generating and getting it wrong most of the time on spin algorithms. End of story been happy ever since with no more frustrations and just enjoying the ride on Rouvy and Zwift.

PeteHski

What issues are you talking about? I had a good experience with an Elite Direto X for a couple of years and the ride feel was quite realistic. I now use a Wahoo Kickr smart bike which is very nice, but another category again.

I think wheel-on trainers can work well if your bike has a power meter, but I don’t have much confidence in their own power measurement and they are often pretty limited in steep slope simulation - which may or may not be an issue depending on what you are trying to simulate.

spelger

had a kicker, have an elite currently. have ridden wheel a on trainer. your experience is different from mine. i'm a FW engineer too and i am not getting the more complex part.

himespau 

I've had no trouble after 2 years on the Elite XR-T. Matches pretty close to my Favero Assioma Duo pedals. On the other hand, everyone that I know who has gone from wheel on to direct drive has seen massive power decreases (and, if they have an on-bike power meter, their trainer power meter suddenly matches their bike power meter when the bike power meter was "always under-reporting" in the past).

spelger

when i first started out using an on wheel trainer at my LBS one of the guys would calibrate his setup then get off and loosen the drive wheel making it easier to pedal.

msu2001la

My Kickr Snap has been trouble free since I purchased it (right after starting this thread). I have no regrets - it's been a very solid trainer and the "wheel on" aspect hasn't bothered me a bit. I've done at least 3000 miles on it and it works great. The max resistance (1500 watts and 12% grade) seems to not be an issue for me - maybe y'all are laying down way more watts than I am. As mentioned above, I have a crank-arm power meter so I don't have to go through the hassle of doing a spin down or worry about calibrating the Snap. Tire wear has been a non-issue for me as well - I've just been running old road tires that I've retired from outdoor use and they seem to go forever on the Snap.

OK, all that said - I would not buy a Kickr Snap now. In Dec 2020 the indoor trainer market was going crazy. Snaps were $500 and Cores were $900 - and both were sold out everywhere and hard to find. Current pricing on a Snap is $300, but you can find Kickr Cores for $400-$500 at the moment - even less if you're willing to search a bit. The Core has better gradient simulation, and is a smoother operation all around. Bonus points for never having to check tire pressure or crank down the roller, and the onboard power meter seems reasonably accurate (unlike the Snap). Beyond that, the differences between these aren't as big as some people make it seem.

Aidoneus

I got a Wahoo Snap On in February and for the first two months was happy on Zift. Then I was told you need to calibrate it--all I did before was tighten the knob till I couldn't force the tire to slip by hand. After doing the basic Wahoo calibration, I could hardly even turn the pedals. I finally found how to do a factory calibration. Now my power readings are down from where I started but not ridiculously so. I have heard people say that you either do no calibration and show too much power or you do a calibration and show too little power. And, yeah, those with a power meter on their bike seem to avoid these issues. Now I wish I got a direct drive model or really splurged and got a smart bike.

RChung

With the cost of on-bike power meters nowadays, I believe I could reasonably argue it makes sense to spend *less* and get a used dumb trainer and a used on-bike PM. That's less than the price of a Snap and you'd have consistent power both when riding indoors and out. Even if you bought new, it'd be close to the price of a Snap. I'm not sure what the value propsition for a Snap is anymore.

himespau 

The only issue I could see would be that indieVelo (if you go there) won't let you race that way. I think I heard MyWhoosh was also moving to only letting trainer power be used in races. Zwift only just this coming month is going to start banning dumb trainers and estimated power (via speed sensor) from their races (not all races, just the ones they coordinate), so they're a bit behind and you should be able to race there with a dumb trainer and a power meter for quite some time. In general though, at current prices, $300 for a Snap vs $500 for a core (if you value the 1 year of Zwift at $150), would come down to whether or not you could afford the core. I hadn't realized the Snap prices had fallen so much, but that's the list on Wahoo's website.

RChung

Just to clarify, Zwift will be banning dumb trainers with only a speed sensor and without an on-bike power meter from their races? I don't think I've seen anyone in Zwift with that setup in a while -- you can see them in game because their avatar has (or used to have) a dumb trainer attached to the rear wheel -- so I'm not sure that's going to have a huge effect.

At the top levels of e-sport, Zwift requires dual-recording so you need both an approved direct-drive trainer and also an on-bike PM.

I haven't looked at either IndieVelo or MyWhoosh but I would think it a mistake if they were banning setups that weren't ANT+ FE-C compliant (which is what a dumb trainer + on-bike PM is); but companies often make decisions without checking with me beforehand so there's that.

himespau 

IndieVelo has always required smart trainers as the source of power for ranked races (power meters are allowed as primary for unranked races) because it's so easy to cheat with a powermeter (just change the listed crank length or mess with the gain among other things for pedal-based for example). indieVelo has dual record on by default (I just connect both my trainer and my Assioma Duos and they both show up with all the extra data from the Duos displayed). MyWhoosh, after getting the UCI esports championships is getting a bit more serious about cheating, so that's why I've heard they're moving that way too.

In Zwift, not all people using Zpower have the semi-transparent trainer attached to their wheel. They just lack the lightning bolt next to their name in the nearby riders list, but there are fewer of them (why Zwift finally decided it wouldn't hurt their bottom line to remove them from the Zwift-run races - community races can have hardware requirements or not).

PeteHski

Well you could also get a used Snap, which is a more fair price comparison against other used gear. It comes down to whether you want more consistent power data or slope simulation. The value proposition for a Snap is still as a relatively low cost Smart trainer, if you want the resistance control.