Related
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

30s VAM wrong (on some rides, but then consistently)

usrusr
over 3 years ago

I have noticed that on some rides, the 30s VAM field is permanently too low, off by about -20%, perhaps -25%. On other rides it works as expected.

I discovered this comparing to a homebrew cIQ field (30s VAM from an altitude ring buffer when climbing, absolute altitude when not) and of course I was suspecting my cIQ to be wrong on those days (and my legs to be weak). But recently I had an opportunity to cross-check against an 830 (I think) riding side by side and it was quite clear that the cIQ was right that day and not the regular 30s VAM. The cIQ field also appeared noticeably more consistent with my powermeter data. In isolation I wouldn't consider observed VAM per watt as proof, but it's strong supporting evidence. Some days they agree (within the expected margin of different 30s avg implementations), other days the regular version is very noticeably too low.

Key lessons:

  1. it can't be water in the sensor (cIQ would be just as confused)
  2. on/off state of the problem can't be linked to updates (it changed in a week without wifi
  3. neither be linked to reboot (I hardly ever do that, and only attempting to fix sync issues which I did not have)
  4. seemingly random on/off of the problem means that it's difficult to reproduce and to verify any fixes if/when they happen. Bummer. 

Are other altitude related problems related? Perhaps, if they are, and thus sharing the seemingly random on/off state this inconsistency will make them very hard to discover/verify (but native 30s vs cIQ might serve as a valuable canary)

  • 0 over 3 years ago

    Now also some days where the native 30s VAM was noticeably too high. And on one occasion there was a nother 1030 plus riding side by side that was even higher by a factor beyond two, on a perfectly rainless day.

    30s VAM was my preferred performance metric on the 1000, even with a powermeter, but on the 1030+ it's beyond useless. Good hunting, please try.

  • 0 over 2 years ago

    Revisiting this topic, some update in the the last half year seems to have fixed VAM: the regular data field and the cIQ reimplementation mentioned seem to be well enough consistent with each other on 6.40 Thumbsup

  • 0 over 2 years ago

    Unfortunately I have to revoke my "all is well now" for 6.40, had wildly exaggerated 30 sec VAM on the native field on some recent rides. Again roughly the same factor of disagreement between native field and cIQ reimplementation for the entire ride (or for the entire "since last full reboot"?).

    And when one 30 sec VAM is claiming 1000 and the other is claiming 1600, my legs, the way I breathe, my hrm and the powermeter, they all know which of those numbers can't be true (in this case, clearly the 1600 couldn't have been true) and they all agree with each other.