Mountain Bike Reviews Forum banner

1 - 14 of 14 Posts

·
Registered
Joined
·
1,949 Posts
Discussion Starter #1
I just got a speed sensor for my edge explore so I can get more accurate speed and distance readings. The directions don’t say anything about calibrating it just how to pare it with the head unit. But when you go in to the sensor menu it has an option to but in the wheel diameter in MM or leave it on auto. I also can’t find any tables to give me an idea of what kind of number to but in it with a 29x2.3 tire.
 

·
Registered
Joined
·
2,343 Posts
If you want to input numbers you'll need to measure the distance travelled.

I believe the sensor will self-calibrate in Auto mode -best suited for longer straight areas.
Find a mile long path you can ride without too many curves and it should calibrate.

I haven't yet purchased but from what I have read here on the forums, that's how you calibrate a speed sensor. I can't recall the distance for best accuracy though.

No internet instruction from Garmin, or YouTube videos explaining how to calibrate?
 

·
since 4/10/2009
Joined
·
30,819 Posts
The magnetless sensors from Garmin don't require calibration. There's an auto-calibrate system in place. I've seen the messages on the screen notifying me that the sensor has been calibrated.

If you want to manually calibrate, you have to measure the rollout of your wheels. Weighted, at your riding weight, using the tire pressures that you ride with. Calculated in mm, this is the number you enter into the manual calibration field. That method is described in so many places, it's not worth typing out here.

For the slight gain in accuracy I might get from doing that, I don't bother. I leave mine on the auto calibration option, just in case my tire pressure is off enough that the computer detects a discrepancy, or if I swap tires, or whatever. I HAVE had the computer occasionally recalibrate the wheel sensor many months after initial setup. Not sure why, but I'm not complaining.
 

·
Registered
Joined
·
9,739 Posts
Yes, it calibrates automatically, which will be best done over a relatively straight segment to minimize the influence of effects the sensor is intended to overcome. There are various circumstances that will trigger it to auto calibrate, but I found that I could trigger it to calibrate by switching from auto, to manual calibration, back to auto. It would then start a calibration when I started riding and alerted me when done...usually in under 1/2 mile. I went into settings and saw what it had come up with for wheel circumference. I did this a number of times over the same segment and found it varied, so took the the average and manually input it. The idea being it would hold that and not try to autocalibrate on it's own in less than ideal circumstances.
 

·
Registered
Joined
·
2,343 Posts
Good tip to average out multiple sensor readings. I wouldn't have considered that.
I will probably take my bikes out and do the same. Curious to see the differences between 27.5+ and 29 as the average compared to a one-time static test.


When I do a wheel rollout I run it over about 60' (sidewalk in front of my house) which is 4 or 5 wheel revolutions on the 29er (been a while since I calibrated).
I performed the 60' rollout a few times and averaged my readings and input the CM into my computer (CatEye wired sensor).
 

·
Registered
Joined
·
9,739 Posts
On the Edge 820, the wheel size (automatic or manual) is displayed under Settings, Sensors, (sensor), Sensor Details. The sensor needs to be awake and connected to the Edge to see it.

When in auto, I don't know all the things that will trigger and auto cal. AFAIK, It doesn't do it for every ride. If you change wheel sizes on the bike, I don't know if it will look for difference in distance between the wheel sensor and GPS and if that will trigger an auto cal to occur.
 

·
Registered
Joined
·
1,949 Posts
Discussion Starter #8
I ended my ride today on a about a half mile of dirt road and it triggered the auto calibration. I haven't looked to see what number it actually came up with. But doing it several times and taking an average sounds like a good idea.
 

·
since 4/10/2009
Joined
·
30,819 Posts
I ended my ride today on a about a half mile of dirt road and it triggered the auto calibration. I haven't looked to see what number it actually came up with. But doing it several times and taking an average sounds like a good idea.
The auto calibrate won't do it that way.

As I understand it, it's looking at discrepancies between what's stored and what it seems to be measuring. So if you do the long, straight section over and over and over again expecting the computer to keep updating, it won't do it that way.

The one time I recall it recalibrating on a ride was when I let it calibrate initially on a not-so-straight section. Some miles later when I got to a much straighter section, it updated the calibration.

It's probably comparing the sensor readings to GPS calculated distance. And GPS calculated distance is notoriously inaccurate the twistier the trail is because it shortcuts the corners. GPS calculated distance is very good over straight, open sections.

I rode a section of trail today that was on an old airstrip on top of a mtn (coincidentally named, Airstrip Trail), before dropping off the side to a fun downhill. If there was ever a spot where my Garmin would recalibrate my wheel sensor, that was the spot. Clear skies, no terrain obstructions, no tree obstructions, totally straight and smooth. It did not recalibrate.
 

·
Registered
Joined
·
9,739 Posts
The auto calibrate won't do it that way....
I believe Pulser is talking about doing it the way I suggested in post #5 above. That is, force an auto cal over a straight segment a number of times, read the value each time, average those values, and input the average as a manual calibration.

I suggest this since I don't know how or when the Edge might decide to do an auto cal, so by putting in a manual value (determined by doing a rollout or using the auto cal avg method) it will prevent it from doing an auto cal when it would be inopportune.
 

·
Registered
Joined
·
1,821 Posts
So my straight-road-roll-out-auto-calibrated sensors were low-balling me on both my road and mountain bikes. I was showing distances about 4-5% shorter than the length of the GPS track.

Remember, it's supposed to be a bit longer than the GPS track.

I did a percent calculation based on the distance difference, and used it to calculate a manual wheel size, and now they agree.

Not sure what the hell was happening, a tire pressure change, or just the 520+ being it's usual self.

....Or maybe I have the wheels locked up way more than I realize. :D
 

·
Rides all the bikes!
Joined
·
3,822 Posts
If it is a straight road, then your GPS track and wheel sensor should line up. You won't notice how your distance is longer than the GPS track until you start getting in to switch backs and such with low GPS signal.
 

·
since 4/10/2009
Joined
·
30,819 Posts
I believe Pulser is talking about doing it the way I suggested in post #5 above. That is, force an auto cal over a straight segment a number of times, read the value each time, average those values, and input the average as a manual calibration.

I suggest this since I don't know how or when the Edge might decide to do an auto cal, so by putting in a manual value (determined by doing a rollout or using the auto cal avg method) it will prevent it from doing an auto cal when it would be inopportune.
Oh, I see. Still, I wouldn't do a manual calibration that way. I'd pull out some paint and put a healthy dab on my tire, ride the bike and use a measuring tape to measure between 3 or so paint marks (basically, as many as you can identify until the paint runs out, or as long as your measuring tape will go, whichever is farthest), convert to metric, and average them out.

If it is a straight road, then your GPS track and wheel sensor should line up. You won't notice how your distance is longer than the GPS track until you start getting in to switch backs and such with low GPS signal.
Yup.

I've been thinking a bit more about doing manual rollouts to calibrate my sensor and my wife's sensor, which are definitely not in absolute agreement even when we do the same ride. Doesn't bother me as much, since it's usually only a couple tenths of a mile different, but she grumbles about it sometimes.
 

·
Registered
Joined
·
9,739 Posts
Doing a rollout measurement is good and I've done it many times over the years. I found that on an MTB with low pressure tires, there is a significant difference between the bike being weighted with the rider and unweighted measurements, weighted yielding a slightly smaller effective circumference, so doing it weighted is somewhat more accurate, if a bit more of a pain to do.

I often ride from home and live on gentle hill, nearly straight for over 1/2 mile. I'd do the autocal over several rides starting from home and then take the avg to put in manually. It was pretty convenient. This, of course, is with me on the bike.

Lately, I haven't bothered and just let the Garmin do its thing as I'm not that concerned about it.
 
1 - 14 of 14 Posts
Top