Related
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Silly speed readings by Garmin speed sensor on canal towpath

GordonFK
over 5 years ago
Hi!
I'm getting some impossibly high speed readings when cycling on some parts of a canal in my hometown (the canal is Regent's Canal in London). I haven't tried all the canals around, but the readings go off as soon as I enter certain parts of the canal towpath, systematically.

See this activity: https://connect.garmin.com/modern/activity/3145511677
The sensor goes completely bonkers as I enter the canal, and comes back into its senses as soon as I leave it. Some parts of the canal it doesn't happen, but some parts it does it every time.

I've got it with the sensors on both my bikes, plus a replacement one... So three sensors, and as many devices (Garmin Edge 1000, Garmin Edge Explore and Vivoactive 3). Whatever the combination, I'm getting impossibly fast readings (between 90 and 120km/h) on some parts of the canal, systematically.

I thought it could have something to do with being near the water? Like some sort of echo or something? Not sure why some parts of the canal and not some others either...
It seems like a design flaw, is there another type of sensor that wouldn't have that problem? I'm having to disable the sensor on my commute, which is quite annoying. Also, would be nice if connect could fix speeds and distance based on GPS data!
  • 0 over 5 years ago
    Which model of speed sensor? Have you replaced battery? Looks like it isn't even connected to me. Post FIT file.
  • 0 over 5 years ago
    Are you certain the speed is coming from the sensor? Do you see a speed reading when you spin the wheels while stationary?
  • 0 over 5 years ago
    Are you certain the speed is coming from the sensor? Do you see a speed reading when you spin the wheels while stationary?


    Yes I'm pretty certain as the speed reading comes back to normal as soon as I disable the sensor - and goes back mental when I enable it. Very easy reproducing it. Also I forgot to mention, stopping the bike doesn't change anything - as long as I'm on that canal the sensor seems to think I'm driving on a motorway, whatever speed I go at.
  • 0 over 5 years ago
    Which model of speed sensor? Have you replaced battery? Looks like it isn't even connected to me. Post FIT file.


    FIT file attached

    Yeah battery failure was my first suspicion actually, I changed the battery but it didn't solve the issue. To be fair it would have been surprising as this happens with 3 different sensors.
    They're apparently all the same model though, not sure if there's a HW version somewhere?
    One of them is this one, bought May 2, 2018
    https://www.amazon.co.uk/gp/product/...?ie=UTF8&psc=1
    Another is this one, bought Feb 11, 2016:
    https://www.amazon.co.uk/gp/product/...?ie=UTF8&psc=1
    The third one came with the Garmin Edge 1000, bought Nov 22, 2014:
    https://www.amazon.co.uk/gp/product/...?ie=UTF8&psc=1
  • 0 over 5 years ago
    The speed sensors send digital data. The magnet passes the sensor and the sensor transmits out a digital packet. The packet includes the ID number of the sensor (so your head unit can ignore packets sent by other nearby sensors). Keep in mind that these are wheel rotation counters. The usual mode of failure is missing (dropping) counts, which would cause a lower than actual speed.

    In this case something is making it look like there are more wheel rotations than actual.

    It's (probably) not the battery (that losing power would probably produce a lower speed).

    There might be some sort of electromagnetic/magnetic interference causing the sensor to think it's rotating or (unlikly) some sort of weird signal reflection going on.

    I suspect it isn't the canal but something else near it.

    https://support.garmin.com/en-US/?fa...Jd3Tnyif9jRSy6
  • 0 over 5 years ago
    The speed sensors send digital data. The magnet passes the sensor and the sensor transmits out a digital packet. The packet includes the ID number of the sensor (so your head unit can ignore packets sent by other nearby sensors). Keep in mind that these are wheel rotation counters. The usual mode of failure is missing (dropping) counts, which would cause a lower than actual speed.

    In this case something is making it look like there are more wheel rotations than actual.

    It's (probably) not the battery (that losing power would probably produce a lower speed).

    There might be some sort of electromagnetic/magnetic interference causing the sensor to think it's rotating or (unlikly) some sort of weird signal reflection going on.

    I suspect it isn't the canal but something else near it.

    https://support.garmin.com/en-US/?fa...Jd3Tnyif9jRSy6


    Something static it has to be. Echo on the water and maybe buildings the other side of the canal?
    Wouldn't the packet have a unique identifier that would help the head unit exclude duplicate readings?

    It would be nice if there was a way to check rides for other registered users to see if others experience the same issue. Regent's Canal in London is a heavily used towpath there must be others with the same setup.

    Reading your link I'm wondering what could contain a magnet in my daily commuting pack. I guess my smartphone has a magnet in its speaker, like all the other smartphones in the world? There's nothing else that has any magnetism that I carry with me.
  • 0 over 5 years ago
    Something static it has to be. Echo on the water and maybe buildings the other side of the canal?

    If it were something like this, one would expect it would be fairly common. (It doesn't seem it's something people complain about very often.)

    Buildings on the other side are probably too far (the range of the transmission is about 15 feet).

    Wouldn't the packet have a unique identifier that would help the head unit exclude duplicate readings?

    It's not clear that duplicate readings from the same sensor are a real problem.

    The only reasonable "unique identifier" would be some sort of "time stamp" (e.g. seconds since last started). The sensors are designed to be as minimal as possible (that makes them easier to design and use less power).

    It seems Garmin put a lot of effort into making the transmitters reliable and accurate (it's hard to make them perfect).

    Hmm. It's possible that the sensors transmit a rate (wheel rotations per unit of time). That would make duplicate readings not an issue because the rate won't change that much moment-to-moment.
    -
    Reading your link I'm wondering what could contain a magnet in my daily commuting pack. I guess my smartphone has a magnet in its speaker, like all the other smartphones in the world?

    I think it's something in the vicinity of the canal.
  • 0 over 5 years ago
    Checked fit file. You have two (!) speed sensors with low battery! Why do you have two sensors at the same time? This would seem to be a good recipe for trouble.
  • 0 over 5 years ago
    Checked fit file. You have two (!) speed sensors with low memory! Why do you have two sensors at the same time? This would seem to be a good recipe for trouble.


    I definitely don't have two sensors obviously, is it possible you're seing a duplicate one? I've tried to find a tool that allows double checking it, but I can't find one... What are you using?
  • 0 over 5 years ago
    Sorry, you specifically stated that you own three sensors and I said that you have two (!) connected at the same time. Now you write "I don't have two obviously". Well, to me, it is not obvious at all why you don't have two!