Grade Anomalies

I rode a 200 mile event on Saturday. My GRADE field seemed to go a great job. Felt better than the native grade. It didn't produce any unreasonable values, seemed to follow the native GRADE closely, but was a bit more stable on what felt like steady gradients (the native grade would sometimes vary +/- 1-2 percent). For grades under 2% I only show an integer, but over 2% I show one decimal point.. since there seems to be more noise in the nearly flat sections of road.

Anyway - two things. One particular user has two observations that seem strange.

1. He uses a RideWithGPS route he loads into his EDGE to simulate a real route using the EDGE to control his trainer. He says the grade field displays grades that are not consistent with the native grade or the expected grade from the loaded file  - sometimes way off. I can't explain why the CIQ field would perform differently. Unless the elevation data in the loaded file isn't properly sent to CIQ's Activity.info.altitude?

2. He swears that my field changes the way the EDGE controls the trainer. LOL. That when my field is running he feels a significant difference in the resistance being sent to the trainer. NO WAY. Right? All my field does is read the distance and elevation and generate a GRADE estimate. Must be a placebo effect of seeing a grade reading. Unless there is another device bug and the compute resources used to run the simple field impacts the priority of sending control data to the trainer?

  • 1. Not sure how this works, but I can imagine that the fit file that he replayed was recorded with whatever algorithm Garmin uses, which itself has some smoothing, etc. So maybe this doesn't matter when you replay a session recorded by the native fields and somehow they display similar things in the real ride and the replay, but your field gets different results because the replayed data is different from what input your app would get during the real ride

    2. Ten years ago we were in some park with the kids and there was some "installation" by the electric company where kids could ride a trainer bike that was connected to a light bulb and they could see how hard it is to pedal fast enough to see any light coming out of the bulb. Of course this was some old 100W bulb, today it would be possible to light up an LED bulb easier. So maybe it's all because of the high battery consumption of your datafield ;)

    Ah, and if there is a bug, that would be this: the trainer gets the "incorrect" grade value from the gut field that your datafield recorded instead of the native grade fut field.

  • Not sure what happens when the edge reproduces a route using the indoor activity profile, but it could be caused by Activity.info.altitude derived from barometric pressure

  • I‘m doing many trainer rides based on activities I did in real life, using my Edge to control my Tacx NEO. I have tested  grade field with this rides and I can tell you: it works as expected - without interfering the resistance (of course!).

    But riding a training based on a real life activity and riding based on a RWGPS prepared course is a big difference: if you prepare a course with any software, for the elevation can only be used a Digital Elevation Model, which is much more unprecise that barometric recorded elevations. That means, that there may be elevation jumps that never happen in a recorded activity. And that leads to huge differences in trainer resistance.

     , you should ask your user to ride a recorded activity and see what will happen.

  • but it could be caused by Activity.info.altitude derived from barometric pressure

    If you play back a recorded fit file in the simulator, the elevations ( Activity.Info.altitude ) are coming from the fit file, and for trainings based on an activity ditto.

    See screenshots of an indoor training - riding a real activity: