cybirr 

TL;DR – Garmin tempe is good for hyper local temperature measurement and logging but has some limitations and use challenges.



After reading some mixed reviews, I wanted to test out a Garmin tempe and see for myself. My first test was to measure the accuracy of the sensor against a calibrated basic lab thermometer. The room was air-conditioned and closed off for reasonably consistent ambient temperature. The setup was both the tempe and the lab probe isolated from any heat sink or conductor. Solo cups worked well. Is there anything a Solo Cup Company product can’t be used for? Over an hour with 5 minute intervals, I tracked the temperature of the lab probe as well as the tempe via a 3rd party app on my Vivoactive 4. You can see the excel spreadsheet for the results. In summary, it’s close to a calibrated thermometer that itself has a +/- accuracy of 1 degree Celsius. That should be good for most uses.






Then I mounted it on my bike and went for a ride. The tempe readings were a consistent 4 degrees Fahrenheit warmer than the National Weather Service (NWS) report. The line graph shows its plot via a 3rd party app on my Garmin Edge 530. The grey block graph is the Edge internal temperature. My NWS sites are five and fifteen miles straight line from me so I’m not entirely clear why the discrepancy.









The most common complaint was accuracy and I get that. Radiant body heat or direct sunlight would definitely skew the data. Isolated from those two error sources and the device seems reasonably accurate. Mounting options are not great. The included bracket is marginal for most situations unless you can move it off your body and not in constant direct sunlight.



But my biggest disappointment is Garmin’s current decision to not make it a native sensor for Edge cycling computers (as of post date). It’s an ANT+ sensor so there is no practical reason to not include it on the Edge computers. There are several 3rd party apps that allow work-arounds for the Edge series. Still, it should be an easy fix.

fooferdoggie 

the 1030+ has a temp sensor. I use it all the time. it can read pretty high riding on blacktop.

Iride01 

It's always seemed to me that the biggest thing with temp sensors is how much lag there is from the true ambient temp. When riding a bike or even walking there are plenty of times when going from wide open sun exposed areas to dense tree shaded areas the temps change quite a bit.

On a bike going 20 to 30 feet per second you can be in and out of those areas pretty fast. So will any temp probe respond quick enough for those changes? My instant read food thermometers require 3 seconds to give an accurate reading. So is Tempe able to do better?

Perhaps put it in a cooler temp and then move it to a hotter temp and see how long it takes to match.

Seattle Forrest

​​​​​​For what it's worth. You're probably one of a dozen people who's ever connected a Tempe to an Edge, most cyclists aren't aware Tempe exists. Edge units have thermometers except the lowest price models, their job is to measure the air temp at the barometer. Having two thermometers on a bike isn't something many people desire. Programming, testing, and customer support are expensive; rather than driving up the price of Edge units they let CIQ developers build support for them. 🙂

Tempe is natively supported by their watches. I have mine mounted near the BB.

Seattle Forrest

Tempe is extremely parsimonious with its battery. It broadcasts twice a minute. Your watch finds Tempe's schedule and then turns its own radio on and off in sync to reduce power consumption.

Iride01 

That's something I didn't think of... how often the device is polled. Every thirty seconds seems enormous to me. Is that just one data point or a log of all the data points since the previous transmission?

A lot of the finer changes in temps of my ride will be missed entirely just by that alone.

I'll still be just as happy with sticking my arm out the door before I leave to gauge what to wear and how I might perform.

Though I don't know what the use is the OP has for their Tempe, other than to have something that might be a little more local to them than the TV weather report.

Seattle Forrest

@Iride01 what do you do with your temperature data that losing the transition that happened from 23 to 24 seconds would throw your experience on the bike off?

Iride01 

I don't do anything with it. That's why for the most part my method of temperature gathering for cycling is only to determine what to wear. I know that some have their reasons to be intrigued by the temps, but barring any extreme temperature concerns which would better be known before going on the ride, I've not any idea what use it might be.

If I did want to know temperature to equate what I was doing on a particular segment of my ride, I'd rather know that the resolution of the data can be resolved to allow several data points for just that segment.

900 feet or about 274 meters does not seem adequate enough if I wanted to use air temp variance to evaluate other data on some of my climbs I track. Though I have no idea why I'd even worry about temp other than it's cold, comfortable or terribly hot. Which doesn't require a thermometer.

Seattle Forrest

​​​​​​It occurred to me that I never answered this. Sorry.

Every 30 seconds, a Tempe unit will broadcast 3 things: the current temperature, plus the high and low temp it's seen over the last 24 hours. According to its own internal clock.

Use cases: I have a Tempe clipped to each paper of ski boots I use. I've been night skiing down to -10F. I keep one on my backpack for hiking. When I hike in and camp in the wilderness, I bring 2 Tempes, one stays outside, the other comes in the tent with me. The watch lets me check either of them at will. None of this is for decision making, or correlating performance data. I have one on my bike because my Garmin records temperature but is affected by my body, and this gets rid of bad data for me.

Tempe costs about $25 and weighs maybe 10g. The battery lasts a year or two.