Edge 840 total ascent readings excessively high

I’ve been for a variety of rides now (short, long, hilly and flat) in dry/damp/warm/cold conditions using both Edge 840 and Edge 130 units simultaneously.
The 840 always gives significantly higher readings, in the order of 10% to 20%.
My old cheap 130 is more accurate, i.e. it’s readings more closely match those given if I apply Strava Elevation Correction to the 840 reading and also to those given by the OS mapping tool I use to plan routes. Has anyone found a way to improve the 840 total ascent reading accuracy?

0 7589378 5 months ago

I recently "upgraded" from a 520 Plus to an 840, using them side by side (because of reports of the 840 failing to save rides). I've seen similar results on total ascent. In six rides on various routes, the 840 returns from 8% to 22% higher ascent totals. Over 172 miles altogether, the 840 and the 520 Plus return totals of 5584' and 4965', the 840 total 12.5% higher.

However, I don't use Strava and don't usually load my rides into a mapping tool. When I loaded two GPX files from activities into plotaroute.com, the total ascent corresponded much more closely to the 840's figures than the 520's.

0 Luko79 5 months ago

it is known issue (by me), i posted about it (nobody answered ofc...) . only the flat/false flat and light rolling terrain are the culprit. the filtering/smoothing algorhythm is not so strong as in the previous (or different line) models. if you just climb up and down on a hill, the values are perfect (so the only problem is how much flat or light rolling terrain in a given activity, if a lot, then the difference is high, if minimal, then the difference is not exists or tiny)