Garmin GPS vs. Speed/Cadence sensor

Garmin Speed/Cadence SensorMy trusty Garmin GSC 10 Speed/Cadence sensor gave up on me the other day. I think this may be the original sensor that came with my Garmin Edge 500, and I know for a fact that my Edge 500 is the single oldest piece of cycling gear that I still own. I have data on Garmin Connect going back to August 21, 1999. Therefore I am not entirely sad nor upset that my sensor went out. It has served me well for a very long time. It has survived more bikes then I can remember, inadvertently tracked a couple of Amtrak Capitol Corridor trips, managed to NOT be on any of the bikes that got stolen, and held it together through more than a few crashes.

What did surprise me however was what seems to be a change in my overall average speeds since the GSC went out of service. That change, of course, was down.

It took me awhile to convince myself that I really was seeing consistently slower averages. (Which also means, obviously, that it took me awhile to finally replace it. In fact, I still haven’t at the time of this writing.) One I convinced myself that I wasn’t just making excuses for a period of laziness, I started thinking about the why.

First, let’s set up the rides I’m talking about. I was comparing my normal commute to work rides. These are rides that I know very well, do all the time, and know what speeds I can expect on average over certain parts of the route. I know about what speed I’ll be going at my comfortable cadence in a given gear. These are the things that made me confident that I wasn’t just making excuses up to explain away my “lazy” numbers.

All that being said there is a bit of subjectivity in this test. I didn’t have some other, third party, that I was riding along with both with and without the sensor to verify my findings (although I’m thinking about doing precisely that.) I feel that even that subjectivity is fairly well mitigated however.

Incidentally, I happened to mention this on Twitter in a very off-hand sort of way, asking if anyone else had witness such a thing. I wasn’t complaining – I wasn’t even upset – just curious:

My ride data is very different since my Garmin speed/cadence battery died. Anyone else notice this? @GarminFitness #garmin

JustAnotherCyclist on Twitter

To my very pleasant response, I quickly saw this response:

@JAnotherCyclist Sorry to hear that, reach out to our product support team in the following link for help:

Twitter response from Garmin

Garmin Edge 500
Garmin Edge 500 on Barfly mount.

OK. So they didn’t give me any new information. But they did acknowledge that I was talking about them. And let’s face it, that is a hell of a lot better than many other companies can manage.

So that gets us back to the why.

When I notice things like this, I’m the type that constantly tries to think of reasons and causes. I put my somewhat basic knowledge of how GPS works to try and figure out why the averages might be different. It is my understanding that Garmin combines the data from both the sensor and GPS. This is especially relevant due to the location I did most of my riding in after the sensor failed: downtown San Francisco.

Why does the fact that I was riding mostly in downtown San Francisco have any impact on the Garmin? Honestly I can not say for certain that it did. But I do have some theories.

First is stopping. On my commute to work I have to stop a lot. Stop lights, stop signs, pedestrians stepping out in front of me… Lots of stopping. And that means lots of auto-pausing that is impacting my average. I’m fairly certain that, with only GPS data, my Garmin is not detecting stops as quickly as it did when it had the input from the speed sensor too. That could theoretically drive the average down compared to the same ride where the speed sensor data is included.

The other thing is the fact that I’m riding through a bunch of tall buildings. I know for a fact that my Garmin can not acquire the satellites when I am standing inside my office building. That will also limit the amount of satellite data available to me when I’m riding outside but between the buildings. It takes a bare minimum of 3 satellites to get a position, but any accuracy requires 4. If buildings are blocking communications with satellites, then my numbers will not be as good. This was very obvious to me as I would see the speed display jump from 6 MPH to 18 MPH in one switch as I took off from a stoplight. I never saw a change like that with the speed sensor on. Instead I would see a series of smaller incremental changes.

These things together seem like really good explanations of why my averages might be lower without the speed sensor. I have yet to be able to talk to anyone that understands the calculation algorithms used in the Garmin so I could be completely wrong.

Or…. I could be making up excuses about why I’ve been slower lately and trying to blame it on my equipment. It is possible I suppose.

Naw… I’m certain it is the Garmin.