Vivoactive 3 HR Broadcast to Edge520 worked! (for one glorious day)

Ah, Garmin's software practises... Must be universally terrible.

What I don't understand it why the broadcast seems to have issues to modern cycle computers yet I use mine regularly with an old Garmin Etrex 30 (ANT+) with no dropouts on mountain bike rides

My last several rides have been plagued by absolutely horrid HR broadcast stability to my Edge520. Too soon to know for sure, but I may have stumbled on a cause: As an experiment I stopped for a minute on today's ride and deleted all third-party watch faces from the VA3. Many of which had HR fields on them... Immediately the broadcast stability was up to stable enough to be usable.

Broadcast failure symptoms:
The VA3 would be in Broadcast mode and displaying accurate-ish HR on my wrist, while the paired Edge520's HR field would indicate – and remain – flatlined until I re-started Broadcast on the VA3. Then after a restart, I'd get maybe a couple minutes but often much less than a minute of broadcast data on the 520 before the HR display failed back to flatline.

This really represented the worst performance I'd experienced since those days earlier this year when the VA3 had to get rebooted to show anything at all. After over an hour of riding with many dozens of manual broadcast-restarts, I remembered that I recently added a new watch face. I tend to like watch faces that have an HR reading on them, and recall reading that faces run in some capacity in the background, I suspect probably also even while broadcast is on. So I thought perhaps watch faces with HR display somehow interrupt or interact with broadcast HR data routing.

For the remainder of the ride I caught a few momentary "––" HR readings on the 520, but they cleared away fairly quickly. IMO this represented more or less a complete change in connection stability. I need to try a few more rides, and then try to pinpoint if any specific face(s) were the problem but this seems a plausible cause.

carbon : When you see OHR reported in the 80-110 bpm range while on a ride - how does that correlate with your cadence? OHR is looking for a periodic change in light condition on the sensor to lock onto which it interprets as heart rate. While cycling your pedal cadence causes a periodic change as well, which I guess can be more pronounced than the change due to your heart beat which causes your cadence to essentially be recorded as heart rate. The same can happen on a run with footfall cadence. That's not specifically a Garmin limitation, but rather a limitation on how OHR works across vendors. Complete dropouts are a different story.

Further experiments haven't really established a clear pattern aside having any third-party face installed runs the risk of killing HR broadcast stability, having several installed seems to guarantee HR broadcast will fail... I thought I'd be able to narrow down what watch face borks broadcast but it's been a little elusive.

Re: WE51's cadence idea, it's plausible but the main issue I'm after is this full failing to broadcast rather than a mis-reading. Looking back at the data for that ride i talked about above, only one HR data point was recorded about every 10 minutes, which was about as often as I had the patience to stop and re-set HR broadcast. Then HR broadcast was unfailingly stable the moment I removed all third-party faces about two-thirds the way through that ride.

For this morning's experiment, I had a couple very simple faces installed - just time, no HR data – and watched the 520's HR field flatline before I'd even started riding. Deleted one watch face, re-set HR broadcast, then it flatlined again. Deleted the second of the two third-party faces and then HR broadcast was fine and remained completely stable for a 2+ hour ride. So, i've disproved specifically "watch faces with HR" as being the cause, but feel pretty confident that it's an issue with having these third-party faces installed at all...

I went for a ride with my 520 this morning, but realised 5k from home I forgot to put my chest strap on. I figured I would try the broadcast feature so turned that on and paired the 520. The 520 reliably displayed what was displayed on the watch face throughout the ride with no noticeable dropouts. Unfortunately what was displayed on the watch face only loosely resembled my actual HR, but this is normal when I use the VA3 to record my commutes (without broadcasting to the 520). The OHR just seems to be useless for riding activities for me. I occasionally have it read what seems to be actual HR for maybe 80% of a ride if I move the watch up my arm and tighten it really tight, but I generally would not use the information from it for any serious ride.

Thankfully the OHR seems to work well when I am not riding... it seems to be good for monitoring HR casually throughout a normal day when I am not doing any intense activity.