Can Garmins go wrong?

Naomi Gilker

 I have a Garmin Forerunner 305 and the readings from my most recent run are remarkably good: 8.78miles in 1h17 with average pace 8:50 min/mile (5:28min/km) ... I'm a slow runner, so this seems a little unlikely. So unless work really does work, my Garmin probably needs to be calibrated somehow; is this something I need to do?

Bought the device used, so I'm really unaware of what I'm supposed to do with it.

My run-result is even more unlikely because I swam for one hour this morning. This is why I'm suspicious about my Garmin

Any and all advice is most welcome. Thanks!

Naomi

 

Joe Matchette

Naomi - I guess it is possible for the readings to be off....it is an piece of electronics after all.  No calibration is necessary.  Perhaps a software update, but no calibration.

However, what is it that makes you think the distance is off?  Is it a course that you know the distance of?  Or is it just that you don't think you can run that fast?  People are known to surprise themselves around here......

One way to check the distance would be to map out your run using MapMyRun.com.  That will give you a confirmation of the distance.  Did you upload the data to GarminConnect.com?  If so, what does the map show for distance?

Also, the review of the 305 on DCRainmaker.com will tell you all about the device and how to use it.

Hope this helps.

russG

Hi Naomi,
I use the 310xt, and I find that I get more accurate readings if I do a soft-reset every month or so.
I'll write up a short description of what and why ... I have to dash to a 2PM meeting so I'll augment this later, but you'll need a decent test area, ideally a relatively straight relatively open (to the sky, where the satellites supposedly are...) section of road with a known distance of 1/2 mile to 1 mile.

Test to get a baseline - device on, jog it, see what the device says relative to what you know.
If you're off, first try: perform an AutoLocate:

Ensure device is powered off
Press and hold Down button
Press and release Power button
Wait 30 seconds
Release Down

After the AutoLocate is performed on the device, turn it on and put it out and allow 15 to 20 minutes to acquire the current position with a clear view of the sky. The fitness device should now function properly.

IF NOT, then your next try is a soft reset:

Power on then press mode (left side bottom) and lap/reset (front left) simultaneously and hold until the screen goes blank (about 7 seconds)

Let me know how that goes and we can talk about two things:
Do you have the electronic user guide for the 305 (I can help you find it)
Hard reset, if we need to go there.

Best,
Russ

russG

Make sure you re-test after the autolocate and re-test after the soft-reset ... to be able to assess whether the accuracy has improved...

Naomi Gilker

Using mapmyrun.com, the distance was estimated to be 9 miles (versus 8.78 miles with the Garmin - used GarminConnect to get the data). I think that this is fairly close given the limitations of mapping a bike path with an online tool (it's not exactly precise). Yes? No?

Regardless, I think I'll give a soft-reset a try to see if this makes a difference. I have access to a flat 1/2 mile route, and I'd rather ensure that it is accurate now that I'm about three-weeks out from race-day.

Thanks very much!

Al Truscott

A few thoughts:

• You're not wearing a foot pod are you? If you are, then when youre outside, you do NOT want to be using the foot pod for measuring pace/distancve.
• Are you running under a heavy cover of trees or under a tunnel? If the sat. signals are being lost for short intervals, this will give you a shorter/faster total distance/pace .
• My family has 4 different 305s, which I've been using since 2007. I've found it to be a rock solid piece of equipment, better in some ways the its progeny, the 310 and 910. But the battery does wear down, and then it starts to get funky. For $75 or so, Garmin will "replace" your battery (actually, they just send you a refurbished watch).

Best easy test is to run a known distance like four times around a track, or a mile on a bike path which is marked, and see what the watch says. If the known distance and the watch agree, you'll just have to live with being faster than you thought.

rmiller

I found my 305, when I had one before the screen died, was reasonably accurate. Sometimes it was a tad long, sometimes a tad short. And by a tad, I mean like maybe 0.2 to 0.25 miles off in a half marathon or something. Never really amounted to a significant difference - maybe a handful of seconds per mile in pace. Ultimately, I ended up saying that the official time was the official time - and thats what I go by, regardless of my garmin data.

But one thing that will always be a variable is tangents around corners, and that could explain some differences in distance of the GPS vs something like mapmyrun or whatever.

William Jenks

Agree with Al that the 305s are great but I stopped buying them when the price went up and the price of the 310 went down. :-)

I'd differ with Al about using a track to calibrate a Garmin, with the exception of if you are calibrating a footpod, in which case that is a perfect idea. An accurately marked bike path would be much better for GPS. The GPS systems do their best when you go more or less straight and are not so good at dealing with turns of any sort. (In the long run of road racing this doesn't matter, but on a track, it's the worst case scenario for a GPS.) Running around an outdoor track WILL give you an exaggerated sense of how inaccurate the GPS is. I agree that plotting a course by mapmyxyz and comparing to Garmin Connect is probably reasonable as you have suggested.

I assume you weren't running in Ottawa in a place like midtown Manhattan or Chicago where huge skyscrapers can actually interfere with the GPS unit too.

Peter Greagg

http://members.endurancenation.us/Forums/tabid/57/aft/11392/Default.aspx#143328
this link discusses the accuracy limitation of GPS devices — additional issues are those WSM Al mentions to do with interruptions to the satellite transmissions.
There is a more detailed version of this that Steve West posted (that as a data/tech kinda guy just loved) but I couldn't find it.