yeah i tracked both today for michael, the drift is pretty substantial, goes from ~28% to 8-10% 45min in.
Appreciate you and the team looking into this.
1 Like
Anything happen with this? Thanks (I’m not sure If I’m hitting this bug or not cause if I am the difference isn’t that much)
I had this with my wheel on tacx vortex trainer - always drifted when it warmed up meaning power match struggled.
New direct drive trainer fixed it though.
Thats not really an issue with the two devices having different power curves though. (my understanding of what he code change will do) Thats a warm up issue that wheel on trainers have.
FYI I have the same model Kickr, and I get the same massive drift as @RONDAL reducing until the trainer is good and hot @25-30mins riding.
Mines against Vector 3’s. I’d suspected the Kickr was the culprit rather than the peddles, this seems to confirm it.
My 4iiii is currently being tested by 4iiii to ensure its not them. I plan on updating this post once i have more definitive technical information.
FWIW I did separate testing with a P2M PM and have had zero drift issues for the last 4 weeks while i have been having my other PM assessed, so i dont immediately agree with your conclusion.
I’ll be interested to see what you find.
I was basing my assumption on RPE. If I rely on the Kickr, then either the opening intervals are too easy, or if they feel right the later intervals are impossible. If I go resistance mode and peddles… then everything hurts. Which unfortunately seems correct.
I posted some graphs about it as linked, but tbh just been riding rollers for the past few months so haven’t thought much about it.
Yes, daily builds and tests are happening. I’m actually trying to get a special build for my ride today but I don’t think it will happen.
As soon as our internal users are happy (and so far there is an improvement but we’re still tweaking) we’ll start opening it up to people outside of the company.
1 Like
Thanks for the update.
I know this is only a data set of four but is this the power matching bug you’re trying to fix? See the critical power curves on these two comparisons:
Zwift is paired to my pioneer (both left and right) over ant+ and the other file is TR paired to a wheel off Kickr2018 over bluetooth and the power meter over Ant+ (with 34154 was only paired over bluetooth to the trainer)