AHare wrote:
I've actually seen the "erg doesn't adjust" bug when coding in Tour de Giro. We'll send the thing a "adjust erg setting to ___W" ANT+ packet and it will acknowledge successful receipt, but then won't change the physical resistance. I've got a couple interval workouts I've built, and it's a workout-killer if the thing doesn't adjust going into or coming out of a hard interval.
The best "fix" I've got is to send the change requests intermittently, not just when the workout changes, but that can cause issues too. The kickr is a bit of a diva.
Your solution of having a ramp is good too: Assuming TR sends the same "set erg mode level" packet once per workout-level change, then it forces it to send a pile of them going into and out of the hard interval, increasing the likelihood one of them will stick (90% of them usually do, but if the one that misses is the start or end of an interval, :sadface:)
I've been lurking here for a while now, it's an interesting thread on the 'inner workings' of the KICKR. There's a KICKR app in the App Store I wrote a while back as a bit of a side project. Would be nice if Wahoo setup a dev forum themselves...
I've seen the same "erg doesn't adjust" connecting via BTLE. My initial approach was to set the Erg power once at the start of each interval, excluding ramps where the power was set continuously (say 5-10 times per second). This worked well for the constant intervals, but resulted in the ramps being delayed (slope of set power not matching actual power), the KICKR was not keeping up with the erg power values I was sending it. Once a ramp finished, it would take some time (say 30sec) for the KICKR to catchup. Current solution is to simply send a set erg request at a lower frequency (1-2 times per second); this seems to work ok, although has the potential to introduce some small delay to the start of intervals.
I imagine TR similarly limits the number of set erg requests being sent to the KICKR, so adding a ramp before an interval may not fix the issue you are seeing. But based on my observations it's the KICKR doesn't lose requests, it simply struggles to keep up with them, so sending a "pile" may not help things.
Power match type capability is something I was thinking of adding a while back, but this thread has kind of convinced me it may be a can-of-worms and I'd assume Wahoo (via firmware) would be better able to support this.
FWIW, I found the Wahoo API (iOS and Android) really easy to work with. Appreciate this isn't available on desktop environments, and if dealing with ANT+ is anything like writing custom *.fit files, well you have my sympathies.