Which is more accurate, generally, smart trainers or crank based power meters?

[Shrike]

Scratching my head over this for a few weeks now and can't seem to find any guidance on it. Basically we have a Neo, which is supposedly really accurate, like 1%. Then we have two Rotor power meters on our two bikes, both Inpower 3D+.

The Rotors both report 10% over the Neo when using Zwift. Zero'd them both loads of times and that doesn't make any difference. Can't find a way to calibrate the Neo at all. Saw some literate that said it can't be calibrated and doesn't need it as it has no moving parts. Okay, fair enough, that probably means something smart in context. I'll let it slide.

So, how do I work out which is closer to real power. 10% is pretty large. Something's wrong somewhere..

[11.4]

That's not an accuracy issue. It's a calibration difference.

[sungod]

bear in mind that they are measuring power at different points

the neo doesn't measure power lost in the drivetrain

whereas a crank based system is measuring close to the point power goes into the drive chain, it should always read higher than the neo

whatever is being lost in the bb, chain, jockey wheels, cannot be detected by the neo, but will be included in the figures from a crank based system

whether the difference should be 10% is another matter

[Stueys]

10% is big, all of those PM's are rated +/-2%, so in theory you can get a relatively broad spread but still not at that big. Drivetrain losses would account for a a minimum amount of power loss (I've a Neo, vector and P2M, they all track very close).

Given you've two PM's tracking close maybe see if the retailer can take a look at the Neo or give you another to try.

[Mep]

I'd rely on the power meter. My Tacx trainers have on occasion displayed up to 20% less power, which I can usually recalibrate.

[Delorre]

I have the a similar "issue", but the other way around. My Stages read 10 à 20w less than my Drivo. I calibrate the Stages before every ride, the Drive can't be calibrated. Given the Stages is left side only, it's possible I put more power on my right leg (I had a hip surgery years ago on th left side). The avg power of my FTP test (220w on the Drivo for the wole thing, 45') had a difference of 18w in favor of the Drivo, the avg for the 20' FTP (325w on the Drivo) was also something like that. So, it's not a fix % deviation in my case. But I don't mind, I have my FTP for the trainer, and an other FTP when I ride outside.

For me, the most accurate device should be the trainer, especially the high end ones. And, as OP has higher reading on the crank vs trainer, it makes some sense, as there is possibly drivetrain loss.

[Shrike]

When I got the first Rotor Inpower, I tested it v the Neo, and it was always within 1 to 3 watts of it. Gave me a lot of confidence in both and that's why I went for another Inpower for the missus a couple months later. Is it possible that somehow the Neo has changed in those months? Or there's some setting I've messed with?

I've been trying both ERG bode and normal mode. Same 10% difference. Really is a solid 10% too at all powers. Weird as.. Come to think of it, I haven't recently tried to put the Neo/Zwift at 100% real gradient, it's about 60% at the moment. Wonder if that could affect it. Will try tomorrow, long shot..

[ms6073]

Can the Rotor Inpower be re-calibrated in a manner similar to an SRM? If the fir Inpower matched up with the Neo when it was new, then I would suspect that it might be time for a new battery in the Inpower or even re-calibrationn. On average, our SRM DA9000's are within 2-3 watts of the Neo until the SRM battery needs replacing.

[Mep]

I'd suggest recalibrating the Neo too. Exact same situation happened with my Tacx Vortex, accurate power numbers in the fall then somehow went off in the winter.

[gewichtweenie]

bear in mind that they are measuring power at different points

the neo doesn't measure power lost in the drivetrain

whereas a crank based system is measuring close to the point power goes into the drive chain, it should always read higher than the neo

whatever is being lost in the bb, chain, jockey wheels, cannot be detected by the neo, but will be included in the figures from a crank based system

whether the difference should be 10% is another matter

"Drivetrain losses" is hugely overstated. The bigger problem, atleast on a non-direct trainer like mine is wheel slip at high power (or, rather, high power changes).

Besides that, I agree with others there is a huge calibration problem

[ms6073]

I'd suggest recalibrating the Neo too.

Unfortunately the Tacx Neo is calibrated at the factory and Tacx does not currently offer any means for the owner to re-calibrate.

[jekyll man]

Q rings by any chance?
There's no saying either the inpower (single-sided?) or the Neo can deal with the micro accelerations like some pm's can, and so over read.

[claus]

Rotor Inpower measures only one side. Maybe that's your problem?

[rcb78]

Doesn't the Tacx have a spin down calibration feature, it should. You should do this periodically, or when things stop matching up closely. This always works for my Hammer vs my Quarq or P2M.

[spdntrxi]

my neo has always reads lower then on bike PM