I keep my house at 73F and noticed the edge will usually read about 81 inside...this might be due to keeping the backlight on 40%? Maybe it will be more accurate with it off. I haven't had it on a ride yet but after reading some older Edge forums this sounds like a higher reading is normal for Edge units. I was thinking of connecting a Tempe under my saddle - but the Edge models don't support it due to having an internal thermometer. I see that the Garmin store has a Tempe app (not yet on the 1030 but is on older Edge models). Has anyone successfully used this somehow connecting a Tempe as a third party device? I'm assuming as a Connect IQ data field, that it does not write to the FIT file regardless, as that would pull the Internal device temp? Otherwise I figured a Tempe would be a nice feature for when the sun is blazing hot, as I could keep it out of the light.
Apart from the fact that the temperature logging as such is not such of a high value to me, there has been postings on this issue with other Edges before.
Apparently the calibration is done when the Garmin is cooled by the (riding) wind. When stationary, and the display continuously on, the generated heat from the electronics is higher and the heat transfer to the ambient, without riding wind, is worse, so the real and measured temperature of the Edge is indeed higher than ambient.
A solution would be to locate the temperature sensor not on the printed circuit board but close to the outside of the unit, not shield by a thick layer of insulating polymer. But that solution would be more expensive to build so apparently no business case.
Thanks for the details Martin. That helps shed some light on it.
I did take a picture of the 1030 next to the 1000. My room is set to 74. The edge 1000 is sticking at 76. The edge 1030 is sticking at 82. (They were a couple degrees more in this pic as they were in my hands prior).
At that point it would make sense to allow for pairing the Tempe sensor. I realize it's not a data field that many people care for. But if it's there, why not allow it to be more accurate? The 1000 didn't seem to be that far off. ciq.forums.garmin.com/.../1250835.jpg
