Ooooh, I see, that makes total sense.
That didn’t even occur to me, I’m just doing TR with no interest in Zwift: you really don’t want to waste any watts in Zwift.
1 Like
Yeah until I started Zwift racing I would have been happy as a clam with my low Neo 2T and scaling my PMs down to match it (or more likely using power match as you suggested above since most of my other sources match the pedals). Zwift is literally the only place I can think of where accurate (rather than precise) power measurement matters. Or maybe if you’re sending your ramp test results to a pro team hoping to get a contract!
@Dede , I agree with those suggesting that you use power match. That will solve all your issues (until you enter a Zwift premier league race and have to use your trainer as the power source)
1 Like
Correct, the ‘order’ should be things closest to your foot are highest, and things closest to the ground are lowest. So a pedal based power meter is higher than a trainer (or PowerTap wheel hub). Depending on how clean our drivetrain is (or mine is), this usually accounts for between 1-3% drivetrain loss, pretty consistently.
I haven’t seen any indication that trainer companies are trying to align to a slightly higher value. There’s been plenty of watercooler chat with these companies about this idea, but I’m not aware of any doing it. Similarly, in talking to Zwift’s anti-cheating team scientists, they’re keenly aware of this too, and are actually doing some pretty interesting ‘big-data’ tracking of it across riders (which is used as another data point to catch cheaters for pro-riders uploading files).
For better or worse, all the evidence points to this not being the case, for either myself or @GPLama .
How do we know this? Because frankly, they suck at it. Literally, out of ALL of the trainers sent to me over the past 2-3 years, only a single unit, the Elite Direto XR and JetBlack VOLT, have managed to be correct on the first attempt. Now in some cases, they’re shipping pre-prod units, but these are pre-prod units intended to be reviewed, and usually on final firmware. So by time I finish the review, it’s final firmware. Just off the top of my head, the Tacx NEO 2T took 2-3 units, the Wahoo KICKR V5 took months of post-launch firmware issues until accurate, the 4iiii Fliiiiight was similar. And we don’t even want to get started on the fiasco that’s been smart bikes. All four brands in this space have required numerous attempts to get firmware functional after initial delivery (Stages, Tacx, Wahoo, and Wattbike).
I outlined all of those scenarios in each of my reviews, but when you step back, it’s really kinda astounding. So either they aren’t doing special QC/QA on it, or they really suck at it. Even more so since what myself and GPLAMA do is so astoundingly repeatable. We both have our own standardized ERG mode tests (that we publish and make available), and we both ride the Zwift Titans Grove course as at least one of our SIM mode rides, as it’s astoundingly good at breaking trainers/bikes. And then we both usually go off and other rides beyond that. But virtually every failure we see is from either our respective ERG mode tests, or Titan’s Grove.
In any case…
Oh, wait, minor update I forgot: The new trainer I’m testing yesterday, complete with the CEO of this company on the e-mail strings? Broke. SIM mode not working with Zwift BLE. Sigh…
2 Likes
Updating on the issue: yesterday i did a clean run again on mounting the outside bike to the trainer; made sure everything looked ok, drivetrain clean etc. It’s a 148 boost, 1x12 eagle; not as easy to mount as my regular old 3x8 that sits in the Direto.
It must have been an issue on how i mounted it the first round, since now they both read consistently well within 1% (actually less). Almost exact match. On the outside ride i did afterwards (right at threshold) it did feel right too and on 10min. intervals at that power even HR matched what i would expect.
So in my case, in the end, it works as good as i might hope for; moreover outside workouts on trails in the mtb are not so easy to nail to the exact W 
BTW, i did have a blast on the workout; found a regular section with fairly even grade and long enough, and did the intervals there. Works great.
As for using power match, the spider PM is on the outside bike and an old one lives on the trainer, so a no go.
Still regardless of the issue of position in the drivetrain (pedal, spider, hub), it would have seemed a bit too much to have close to 6% difference.
In any case, the Direto has been very reliable since day 1, still calibrates to spec (for whatever it’s worth); the Sigeyi looks to be quite good this early - hope it keeps going like this.
1 Like
Great news, glad it worked out. @dcrainmaker thanks for the enlightening reply. I guess that brings me back to the idea that trainer companies just plain suck at power. Which is a bummer since it matters a lot for Zwift racing, and the only way to do it “right” is to go through what I have these last couple months, owning multiple PMs, doing lots of tests, and potentially having to buy another trainer until you have one that’s verified accurate. Clearly the vast majority of people aren’t going to do that.
As long as you and @GPLama keep doing your good work and pushing for them to be better I think they’ll get there. That said, I think your reviews could stand to have a paragraph in there about how trainer accuracy often varies wildly between units and just because you have managed to find an accurate trainer (once the firmware issues are resolved and you’re on a production version) doesn’t mean every unit that people buy of that model will be accurate.