Garmin Edge - elevation gain/loss errors
#1
Full Member
Thread Starter
Join Date: Feb 2014
Location: Lehigh Valley
Posts: 258
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 132 Post(s)
Liked 62 Times
in
46 Posts
Garmin Edge - elevation gain/loss errors
On Saturday I rode an event that had ~9700' of elevation gain. My edge 530 recorded it as ~3600 feet. Frequently I would be going uphill, sometimes steep ones (low-mid teens in grade %, I would estimate), and the display would show something like negative 4% grade.
I wonder if anyone else has had something like this occur and whether it was an isolated occurrence? Or whether there was some setting or correction that can be applied? Or if it is a sign that the unit is doomed?
There was a light rain all day, if that matters. But I have never had any issue like this riding in the rain before - though I've never ridden in the rain for so long before either.
I wonder if anyone else has had something like this occur and whether it was an isolated occurrence? Or whether there was some setting or correction that can be applied? Or if it is a sign that the unit is doomed?
There was a light rain all day, if that matters. But I have never had any issue like this riding in the rain before - though I've never ridden in the rain for so long before either.
#2
Perceptual Dullard
Join Date: Sep 2009
Posts: 2,487
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 957 Post(s)
Liked 1,223 Times
in
527 Posts
Edge units use a barometric altimeter to estimate altitude, then there's an algorithm to ignore tiny bumps. But if you live in one spot and travel to another, it's often necessary to re-set where it thinks the starting altitude is. In addition, because the barometer depends on changes in (surprise!) barometric pressure, you can be slightly off if the weather changes dramatically during the course of a ride.
To diagnose whether you have some systematic problem with the unit, look at the actual elevation profile of the ride and compare it to the elevation profile downloaded from your Edge.
To diagnose whether you have some systematic problem with the unit, look at the actual elevation profile of the ride and compare it to the elevation profile downloaded from your Edge.
#3
Senior Member
Join Date: Mar 2010
Posts: 1,715
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 1282 Post(s)
Liked 1,382 Times
in
702 Posts
I have been using Garmin Edge units since the Edge 705 and never experienced something close to this. Chances are something is wrong with your unit. Either the barometric sensor has failed or perhaps the sensor port is plugged. This level of error is an indication something is wrong. Obviously power the unit on and off first perhaps even try a reset but unlikely that will fix it.
Likes For Atlas Shrugged:
#4
Senior Member
Join Date: Dec 2010
Location: northern Deep South
Posts: 8,959
Bikes: Fuji Touring, Novara Randonee
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 2633 Post(s)
Liked 1,985 Times
in
1,242 Posts
It takes a while for Garmins in general to calculate a grade; something on the order of 5 seconds IME. So you hit the bottom of a "valley" and start to climb, then look down to see what the grade is? It's still locked on the that lovely downhill. Keep climbing long enough, and the grade may show 6% when you level off at the top.
Of disputations there is no end, and that goes double for climbing calculations. On a winding road with cuts to reduce the grade, elementary mapping software will often go over the hill and give you an erroneous climbing summation. My Garmin won't match your Garmin, even if they're the same model; individual variations in the barometers will make sure our climbing totals are off, sometimes 20', sometimes 200'. And your barometric Garmin won't match somebody else's non-barometric GPS which uses maps, or even worse the GPS, to calculate climbing totals. (GPS is notoriously inaccurate for altitudes, at least at a small scale. Works OK for airplanes separated by at least 1,000' and flying straight and level, though.)
So relax and try not to obsess about differences.
And BTW, I'd believe the barometer before I'd believe the GPS or a map, unless a major weather front was blowing in.
Of disputations there is no end, and that goes double for climbing calculations. On a winding road with cuts to reduce the grade, elementary mapping software will often go over the hill and give you an erroneous climbing summation. My Garmin won't match your Garmin, even if they're the same model; individual variations in the barometers will make sure our climbing totals are off, sometimes 20', sometimes 200'. And your barometric Garmin won't match somebody else's non-barometric GPS which uses maps, or even worse the GPS, to calculate climbing totals. (GPS is notoriously inaccurate for altitudes, at least at a small scale. Works OK for airplanes separated by at least 1,000' and flying straight and level, though.)
So relax and try not to obsess about differences.
And BTW, I'd believe the barometer before I'd believe the GPS or a map, unless a major weather front was blowing in.
#5
Senior Member
Join Date: Sep 2017
Posts: 8,092
Mentioned: 42 Post(s)
Tagged: 0 Thread(s)
Quoted: 7115 Post(s)
Liked 11,275 Times
in
4,812 Posts
Likes For Koyote:
#6
I'm good to go!
Join Date: Jul 2017
Location: Mississippi
Posts: 15,265
Bikes: Tarmac Disc Comp Di2 - 2020
Mentioned: 52 Post(s)
Tagged: 0 Thread(s)
Quoted: 6344 Post(s)
Liked 4,945 Times
in
3,405 Posts
Sometimes they just screw up elevation gain for a assortment of little reasons that all add up to a big error. Your experience is by far not the first despite others that haven't experienced it yet.
However what does Strava, RideWithGPS or GarminConnect show for the elevation gain/loss when you build that same route on it? Sometimes the routes event planners make have a lot of flaws you wouldn't expect.
As far as the grade field on your device. There will be some lag with it. It takes a few long seconds for my to go from my downhill portion to my uphill if there is no flat between. And even from the flat, it takes a while to show the climb.
Sometimes the static port can get clogged up and make wonky readings. I'm not certain how they do that on the Edge 530. My 500 use to be one little hole that could be plugged with water or gunk. Then both gain/loss and grade will really be messed up.
If your device is less than 2 years old and it continues to be off quite a bit then call Garmin Support. They'll eventually send you another unit if they can't solve it.
However what does Strava, RideWithGPS or GarminConnect show for the elevation gain/loss when you build that same route on it? Sometimes the routes event planners make have a lot of flaws you wouldn't expect.
As far as the grade field on your device. There will be some lag with it. It takes a few long seconds for my to go from my downhill portion to my uphill if there is no flat between. And even from the flat, it takes a while to show the climb.
Sometimes the static port can get clogged up and make wonky readings. I'm not certain how they do that on the Edge 530. My 500 use to be one little hole that could be plugged with water or gunk. Then both gain/loss and grade will really be messed up.
If your device is less than 2 years old and it continues to be off quite a bit then call Garmin Support. They'll eventually send you another unit if they can't solve it.
Last edited by Iride01; 05-20-24 at 10:17 AM.
#7
Senior Member
Join Date: Dec 2010
Location: northern Deep South
Posts: 8,959
Bikes: Fuji Touring, Novara Randonee
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 2633 Post(s)
Liked 1,985 Times
in
1,242 Posts
And?
I've seen climbing divergence of that magnitude before, so I'm not surprised, especially if you're comparing a climb calculated from a map to one derived from barometric pressure. Note OP hasn't told us where the event's climb figure came from.
A couple of additional notes. First, what's the "dead zone" of your device? I have a watch that doesn't add anything less than 60 feet onto a climb (and I think the old Avocet cyclocomputer was similar. If you're comparing that watch to a modern Garmin barometer with a 6' dead zone, guess which one registers more climbing in rolling terrain? Second, while GPS x-y coordinates lock in pretty quickly, the same is not true of z (altitude). Factor in a 4,000-9,000 foot climb is likely not flat, and your GPS may lose satellites near a steep climb or rock fact, and the altitude measurement goes wonky right when you need it. Higher? Lower? Flip a coin. What's the "net climb" calculation worth when you don't know how good the altitude measurement it was based on is? (Hint: not much!)
So, enjoy the vigorous discussion you may have, but don't stress about it.
I've seen climbing divergence of that magnitude before, so I'm not surprised, especially if you're comparing a climb calculated from a map to one derived from barometric pressure. Note OP hasn't told us where the event's climb figure came from.
A couple of additional notes. First, what's the "dead zone" of your device? I have a watch that doesn't add anything less than 60 feet onto a climb (and I think the old Avocet cyclocomputer was similar. If you're comparing that watch to a modern Garmin barometer with a 6' dead zone, guess which one registers more climbing in rolling terrain? Second, while GPS x-y coordinates lock in pretty quickly, the same is not true of z (altitude). Factor in a 4,000-9,000 foot climb is likely not flat, and your GPS may lose satellites near a steep climb or rock fact, and the altitude measurement goes wonky right when you need it. Higher? Lower? Flip a coin. What's the "net climb" calculation worth when you don't know how good the altitude measurement it was based on is? (Hint: not much!)
So, enjoy the vigorous discussion you may have, but don't stress about it.
#8
Senior Member
Join Date: Sep 2017
Posts: 8,092
Mentioned: 42 Post(s)
Tagged: 0 Thread(s)
Quoted: 7115 Post(s)
Liked 11,275 Times
in
4,812 Posts
force10 : Garmin CS is pretty good. Contact them, even though your computer is likely out-of-warranty.
Last edited by Koyote; 05-20-24 at 01:23 PM.
Likes For Koyote:
#9
Full Member
Thread Starter
Join Date: Feb 2014
Location: Lehigh Valley
Posts: 258
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 132 Post(s)
Liked 62 Times
in
46 Posts
Ride with GPS and Strava agree that the route was 9700' - 9900' feet. The ride organizer laid out the route organizer on RWGPS and that file had the elevation gain at ~9700'.
I certainly know the difference between going up a 15% grade and down a 4% grade.
At least for me, the difference between 3700' and 9700' over 94 miles is rather significant. And one that this or any ride computer ought to be able to determine. Historically this unit has been able to do so, granting that datum and measurements will vary within reason. What I'm trying to determine now is whether the degree of error I experienced the other day was the result of a known weakness expereinced by other users of the same/similar unit, or perhaps a software or hardware problem.
I certainly know the difference between going up a 15% grade and down a 4% grade.
At least for me, the difference between 3700' and 9700' over 94 miles is rather significant. And one that this or any ride computer ought to be able to determine. Historically this unit has been able to do so, granting that datum and measurements will vary within reason. What I'm trying to determine now is whether the degree of error I experienced the other day was the result of a known weakness expereinced by other users of the same/similar unit, or perhaps a software or hardware problem.
Likes For force10:
#10
Senior Member
Join Date: Sep 2007
Location: Far beyond the pale horizon.
Posts: 14,369
Mentioned: 32 Post(s)
Tagged: 0 Thread(s)
Quoted: 4317 Post(s)
Liked 1,394 Times
in
973 Posts
Resetting might be useful to see whether the barometer is basically working. Set it to a known altitude. See if it reports the correct altitude somewhere significantly higher/lower.
#11
Senior Member
Join Date: Sep 2007
Location: Far beyond the pale horizon.
Posts: 14,369
Mentioned: 32 Post(s)
Tagged: 0 Thread(s)
Quoted: 4317 Post(s)
Liked 1,394 Times
in
973 Posts
The planners might tend to underestimate the elevation. But tunnels or bridges can cause an over estimate of gain in those places.
The numbers are going to be better for larger gain/mile. That is the result of a 500ft gain over 100 miles might be different by a lot.
I suspect that a +/- 20% difference could be "reasonable".
Your number is so different that something appears not to be working. Maybe, the barometer hole is plugged or wet.
If you can, see what other people got from their Garmins (for this or other rides).
At least for me, the difference between 3700' and 9700' over 94 miles is rather significant. And one that this or any ride computer ought to be able to determine. Historically this unit has been able to do so, granting that datum and measurements will vary within reason. What I'm trying to determine now is whether the degree of error I experienced the other day was the result of a known weakness expereinced by other users of the same/similar unit, or perhaps a software or hardware problem.
It's not likely to be a software problem (assuming you are using the latest firmware).
Last edited by njkayaker; 05-20-24 at 01:49 PM.
#12
I'm good to go!
Join Date: Jul 2017
Location: Mississippi
Posts: 15,265
Bikes: Tarmac Disc Comp Di2 - 2020
Mentioned: 52 Post(s)
Tagged: 0 Thread(s)
Quoted: 6344 Post(s)
Liked 4,945 Times
in
3,405 Posts
Do five more rides and find out. Why worry about one ride and whether that one ride is indicative of a weakness? About the only thing you'd be able to do for it is make sure the static port isn't blocked. Don't blow in it though or use compressed air. And on the Edge 530, I'm not sure where it's at. I have one, but I don't really care about gain/loss or grade. I'm always going up and down on my rides. So I know I get enough of it.
Changing atmospheric pressure during the time of your ride can affect some of this too. And most of these devices don't correct for that during the ride once they've established their reference.. A tenth of a inch of mercury change, 0.10 Hg, is 100 feet of altitude. A full inch will be 1,000 feet of error. I doubt you had that much change of atmospheric pressure, but you should be able to see that any of that along with other things can add up to a big error.
Hopefully it's just this one for time. But I'm sure you'll have another in the future if gain loss is that important to you. As for the grade field, I think it's one of the most useless fields you can put on your screen. It lags to much and your legs will tell you more about what you should be doing.
Changing atmospheric pressure during the time of your ride can affect some of this too. And most of these devices don't correct for that during the ride once they've established their reference.. A tenth of a inch of mercury change, 0.10 Hg, is 100 feet of altitude. A full inch will be 1,000 feet of error. I doubt you had that much change of atmospheric pressure, but you should be able to see that any of that along with other things can add up to a big error.
Hopefully it's just this one for time. But I'm sure you'll have another in the future if gain loss is that important to you. As for the grade field, I think it's one of the most useless fields you can put on your screen. It lags to much and your legs will tell you more about what you should be doing.
#13
Senior Member
Join Date: Sep 2007
Location: Far beyond the pale horizon.
Posts: 14,369
Mentioned: 32 Post(s)
Tagged: 0 Thread(s)
Quoted: 4317 Post(s)
Liked 1,394 Times
in
973 Posts
I've seen climbing divergence of that magnitude before, so I'm not surprised, especially if you're comparing a climb calculated from a map to one derived from barometric pressure. Note OP hasn't told us where the event's climb figure came from.
A couple of additional notes. First, what's the "dead zone" of your device? I have a watch that doesn't add anything less than 60 feet onto a climb (and I think the old Avocet cyclocomputer was similar. If you're comparing that watch to a modern Garmin barometer with a 6' dead zone, guess which one registers more climbing in rolling terrain? Second, while GPS x-y coordinates lock in pretty quickly, the same is not true of z (altitude). Factor in a 4,000-9,000 foot climb is likely not flat, and your GPS may lose satellites near a steep climb or rock fact, and the altitude measurement goes wonky right when you need it. Higher? Lower? Flip a coin. What's the "net climb" calculation worth when you don't know how good the altitude measurement it was based on is? (Hint: not much!)
The Edges don't regularly report elevation gains that are 2.5 less than the reported gain.
Last edited by njkayaker; 05-20-24 at 03:40 PM.
Likes For njkayaker:
#14
Senior Member
Join Date: Sep 2007
Location: Far beyond the pale horizon.
Posts: 14,369
Mentioned: 32 Post(s)
Tagged: 0 Thread(s)
Quoted: 4317 Post(s)
Liked 1,394 Times
in
973 Posts
Changing atmospheric pressure during the time of your ride can affect some of this too. And most of these devices don't correct for that during the ride once they've established their reference.. A tenth of a inch of mercury change, 0.10 Hg, is 100 feet of altitude. A full inch will be 1,000 feet of error. I doubt you had that much change of atmospheric pressure, but you should be able to see that any of that along with other things can add up to a big error.
A 0.4 inHg is the difference between calm/clear and warm-air/rainstorms. If there were a lot of changes going on there, one would hope the OP would have noticed (reported it). Whatever weather could cause a 6000 foot error would certainly be notable!
https://bassforecast.com/barometric-...d%20rainstorms.
A barometric reading over 30.20 inHg is generally considered high, and high pressure is associated with clear skies and calm weather.
A barometric reading below 29.80 inHg is generally considered low, and low pressure is associated with warm air and rainstorms.
A barometric reading below 29.80 inHg is generally considered low, and low pressure is associated with warm air and rainstorms.
Last edited by njkayaker; 05-20-24 at 01:51 PM.
#15
Perceptual Dullard
Join Date: Sep 2009
Posts: 2,487
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 957 Post(s)
Liked 1,223 Times
in
527 Posts
At least for me, the difference between 3700' and 9700' over 94 miles is rather significant. And one that this or any ride computer ought to be able to determine. Historically this unit has been able to do so, granting that datum and measurements will vary within reason. What I'm trying to determine now is whether the degree of error I experienced the other day was the result of a known weakness expereinced by other users of the same/similar unit, or perhaps a software or hardware problem.
#16
I'm good to go!
Join Date: Jul 2017
Location: Mississippi
Posts: 15,265
Bikes: Tarmac Disc Comp Di2 - 2020
Mentioned: 52 Post(s)
Tagged: 0 Thread(s)
Quoted: 6344 Post(s)
Liked 4,945 Times
in
3,405 Posts
Have you looked at the elevation graph for your ride. There was one time my elevation graph had a big glitch mid-ride and all of a sudden it started showing everything as being quit a bit higher. Even though the ride was loops over the same route.
I just chalked that up to stuffs going to mess up occasionally. I never noted it again. For the most part, it's correct enough for the use.
I just chalked that up to stuffs going to mess up occasionally. I never noted it again. For the most part, it's correct enough for the use.
Likes For Iride01:
#17
Full Member
Thread Starter
Join Date: Feb 2014
Location: Lehigh Valley
Posts: 258
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 132 Post(s)
Liked 62 Times
in
46 Posts
Then I'll repeat my suggestion: compare the downloaded elevation profile with the actual elevation profile. What you're looking for is 1) if it was recording properly for a while and then suddenly diverged; 2) if it was always off; 3) if there was a starting discrepancy and then things evened up; or 4) if it was "compressed" in the sense that the it had the peaks and troughs in the right places but the gains and losses were "muted." This will give you a clue as to whether this is a transitory or a systematic issue.
There has been a suggestion that a barometer port has been clogged or affected. If that ever to happen, it would have been on this ride - steady rain throughout and 80% gravel. 90 minutes cleaning the bike yesterday and I’m still finding fine grit.
Any suggestions on where to look or how to check?
#18
Perceptual Dullard
Join Date: Sep 2009
Posts: 2,487
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 957 Post(s)
Liked 1,223 Times
in
527 Posts
https://support.garmin.com/en-US/?fa...nZ99DcunfAue66
Likes For RChung:
#19
Senior Member
Join Date: Sep 2007
Location: Far beyond the pale horizon.
Posts: 14,369
Mentioned: 32 Post(s)
Tagged: 0 Thread(s)
Quoted: 4317 Post(s)
Liked 1,394 Times
in
973 Posts
Keep in mind that Garmin believes using a barometer (properly functioning) produces the best results.
#20
don't try this at home.
Join Date: Jan 2006
Location: N. KY
Posts: 5,979
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 988 Post(s)
Liked 530 Times
in
365 Posts
After riding in a rainstorm, a local rider's Garmin now reads approx 15% to 20% lower than other Garmins on the same ride. It hasn't improved weeks later. No more wet riding, and he did some days of putting the device in a bag of dry raw rice, which is supposed to help dry out electronics.
Your "flatline" elevation chart does sound like a failed barometer inside the Garmin.
Your "flatline" elevation chart does sound like a failed barometer inside the Garmin.
Likes For rm -rf:
#21
Senior Member
Join Date: May 2021
Posts: 8,808
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 4598 Post(s)
Liked 5,142 Times
in
3,177 Posts
I’ve been using an Edge 530 for several years and it is pretty consistent with reported elevation gain over countless rides on known courses. The only time I’ve known it to be significantly off is when the weather changes dramatically during a ride. But even then I have never seen an error of anything like this magnitude. In this case it is clearly malfunctioning!
#22
Senior Member
Join Date: Dec 2010
Location: northern Deep South
Posts: 8,959
Bikes: Fuji Touring, Novara Randonee
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 2633 Post(s)
Liked 1,985 Times
in
1,242 Posts
FTR, do you know when the course was laid out and the elevation planned? I ask because the old Topo and similar baselines were used up to about 15 years ago, and they showed the problems I discussed earlier. Modern RWGPS or Strava (say, last 5 years) seem to incorporate lots of riders' tracks using new (barometric sensor) GPS devices to refine some of the earlier mapping bugs. Of course, if few riders have ridden these roads, the refining data will be slim, so the mapping base may not have been updated very well.