Polar wireless speed transmitter w S-710

Discussion in 'Cycling Equipment' started by Chris, Sep 29, 2004.

  1. Chris

    Chris Guest

    I have a speed sensor with my Polar S710 (and a previous model as well). I
    have always had to correct a few spikes in velocity which is not a problem
    for me. I also use a hard wired computer to avoid toggling for data while
    riding. Both devices are always very close in total distance and ride time.

    Lately, (the past week or so) I have had the number of spikes that need
    correction go way up. I now have to correct 10 or more per hour (out of 180
    samples) which is still not bad but it is annoying. BTW, sampling at 5
    seconds makes it very unlikely that it will ever be off by any significant
    amount. I never have any errors adjacent to one another. I still want to
    look in to improving this. I recall that there is a switch in some of these
    Polar wireless units. I have both old and new styles for cadence and speed.
    I almost never have any problems with cadence. I guess I should note that I
    have the watch mounted on a 90 degree stem stub about 7 cm past the fork.
    This puts it nearly directly above the speed sensor and about 20 cm forward
    or the cadence sensor. I realize that Polar has purposely made these
    transmitter somewhat directional and that the "sweet spot" for transmission
    may not be what seems most obvious.

    I would like to know which transmitters can be adjusted for signal strength
    and how to determine the ideal transmission and reception angle for all of
    them. I did experiment with moving the watch all over and it had almost no
    effect. I think that my next move WRT transmission is changing the angle of
    the transmitter but I would like to know how the antennae is mounted
    internally. If I don't hear from anyone, I suppose I can crack them both
    open and see what I find.

    Thanks for any help.
     
    Tags:


  2. "Chris" <[email protected]> wrote in message
    news:[email protected]
    > I have a speed sensor with my Polar S710 (and a previous model as well). I
    > have always had to correct a few spikes in velocity which is not a problem
    > for me. I also use a hard wired computer to avoid toggling for data while
    > riding. Both devices are always very close in total distance and ride

    time.
    >
    > Lately, (the past week or so) I have had the number of spikes that need
    > correction go way up. I now have to correct 10 or more per hour (out of

    180
    > samples) which is still not bad but it is annoying. BTW, sampling at 5
    > seconds makes it very unlikely that it will ever be off by any significant
    > amount. I never have any errors adjacent to one another. I still want to
    > look in to improving this. I recall that there is a switch in some of

    these
    > Polar wireless units. I have both old and new styles for cadence and

    speed.
    > I almost never have any problems with cadence. I guess I should note that

    I
    > have the watch mounted on a 90 degree stem stub about 7 cm past the fork.
    > This puts it nearly directly above the speed sensor and about 20 cm

    forward
    > or the cadence sensor. I realize that Polar has purposely made these
    > transmitter somewhat directional and that the "sweet spot" for

    transmission
    > may not be what seems most obvious.
    >
    > I would like to know which transmitters can be adjusted for signal

    strength
    > and how to determine the ideal transmission and reception angle for all of
    > them. I did experiment with moving the watch all over and it had almost no
    > effect. I think that my next move WRT transmission is changing the angle

    of
    > the transmitter but I would like to know how the antennae is mounted
    > internally. If I don't hear from anyone, I suppose I can crack them both
    > open and see what I find.
    >
    > Thanks for any help.
    >
    >


    I also have had spikes with my Polar S710, and they are always in the same
    place on one particular part of one of my training ride. My unit spikes
    with the cadence, speed, heart rate, and altitude. The errors can be
    corrected, usually it's one - three sections together. I believe I narrowed
    it down to the signal light transmitter sensor for changing stop lights at
    the turnaround junction of my training ride. At this junction, I'm usually
    in the left had turning lane, waiting for the light to turn green to make a
    u-turn.

    Just curious, do your spikes occur near long stoplight?
    -tom
     
  3. Tom Nakashima wrote:
    > "Chris" <[email protected]> wrote in message
    > news:[email protected]
    > Just curious, do your spikes occur near long stoplight?
    > -tom


    Or Powerlines.
    Yes my spikes too are always at the same place of my ride. As I pass
    underneath some powerlines. Interestingly enough I can pass under other
    powerllines that do not produce any spikes.

    --
    Perre

    You have to be smarter than a robot to reply.
     
  4. Chris

    Chris Guest

    "Tom Nakashima" <[email protected]> wrote in message


    The errors can be
    > corrected, usually it's one - three sections together.


    I know. Luckily my cadence does not spike and neither does heart rate.

    I believe I narrowed
    > it down to the signal light transmitter sensor for changing stop lights at
    > the turnaround junction of my training ride. At this junction, I'm

    usually
    > in the left had turning lane, waiting for the light to turn green to make

    a
    > u-turn.
    >
    > Just curious, do your spikes occur near long stoplight?
    > -tom


    No, it is definitely unrelated to external interference. It even happens on
    country roads miles from any EMF. I tend to think that somehow the speed
    "pulse" gets interpreted by the receiver twice on a single wheel rev. I
    think that if the signal is somehow interrupted and then restored all while
    the transmitter is broadcasting its signal for that rev, that would account
    for this. If this sounds too illogical let me explain it with an example.
    Let's say that each pulse lasts 5/100s of a second. The whole system has to
    have some fault tolerance so a signal that is too short is less likely to
    get through. Again for fault tolerance and to keep costs down the receiver
    does not care how long the signal is. It also is not set to reject any
    signals no matter how close together they are. Now add to that a case where
    the reception is marginal but still persistent. In cases where the spike
    occurs, I am moving along and the transmitter is constantly changing its
    placement relative to the receiver slightly. The transmitter sends its
    signal for 5/100s and the receiver picks it up for the first (or second or
    whatever) and then loses it but catches that last fraction. It counts it
    twice since it recorded to pulses. Also keep in mind that most of the time
    in a case like this, you would probably be persistently getting much less of
    the signal than that ideal 5/100 of a second. If the signal is lost at the
    beginning or the end or even on the middle, as long as it is not restored
    before it completes it broadcast it will count it accurately. I think this
    also explains why there is an interest in digital wireless cyclometers.

    I don't know how long each signal pulse is sent but this is really the only
    logical way that these spikes can occur that I can think of. It never drops
    data, it always comes up with an extra. Also, the amount it is off is always
    equal to one extra rev.

    I think if I do everything I can to ensure each pulse is received as one
    (and not split) then I should be good. That is why I am looking at signal
    strength and optimal direction.
     
  5. Chris

    Chris Guest

    "Per Elmsäter" <[email protected]> wrote in message
    news:p[email protected]
    > Tom Nakashima wrote:
    > > "Chris" <[email protected]> wrote in message
    > > news:[email protected]
    > > Just curious, do your spikes occur near long stoplight?
    > > -tom

    >
    > Or Powerlines.
    > Yes my spikes too are always at the same place of my ride. As I pass
    > underneath some powerlines. Interestingly enough I can pass under other
    > powerllines that do not produce any spikes.
    >


    As I said, I get them even when there is nothing around to give EMF. It
    tends to happen more at speed on rough roads (oscillation?). I think I might
    try to look for other patterns.
     
  6. Chris wrote:
    > "Tom Nakashima" <[email protected]> wrote in message
    >
    >
    > The errors can be
    >> corrected, usually it's one - three sections together.

    >
    > I know. Luckily my cadence does not spike and neither does heart rate.
    >
    > I believe I narrowed
    >> it down to the signal light transmitter sensor for changing stop
    >> lights at the turnaround junction of my training ride. At this
    >> junction, I'm usually in the left had turning lane, waiting for the
    >> light to turn green to make a u-turn.
    >>
    >> Just curious, do your spikes occur near long stoplight?
    >> -tom

    >
    > No, it is definitely unrelated to external interference. It even
    > happens on country roads miles from any EMF. I tend to think that
    > somehow the speed "pulse" gets interpreted by the receiver twice on a
    > single wheel rev. I think that if the signal is somehow interrupted
    > and then restored all while the transmitter is broadcasting its
    > signal for that rev, that would account for this. If this sounds too
    > illogical let me explain it with an example. Let's say that each
    > pulse lasts 5/100s of a second. The whole system has to have some
    > fault tolerance so a signal that is too short is less likely to get
    > through. Again for fault tolerance and to keep costs down the
    > receiver does not care how long the signal is. It also is not set to
    > reject any signals no matter how close together they are. Now add to
    > that a case where the reception is marginal but still persistent. In
    > cases where the spike occurs, I am moving along and the transmitter
    > is constantly changing its placement relative to the receiver
    > slightly. The transmitter sends its signal for 5/100s and the
    > receiver picks it up for the first (or second or whatever) and then
    > loses it but catches that last fraction. It counts it twice since it
    > recorded to pulses. Also keep in mind that most of the time in a case
    > like this, you would probably be persistently getting much less of
    > the signal than that ideal 5/100 of a second. If the signal is lost
    > at the beginning or the end or even on the middle, as long as it is
    > not restored before it completes it broadcast it will count it
    > accurately. I think this also explains why there is an interest in
    > digital wireless cyclometers.
    >
    > I don't know how long each signal pulse is sent but this is really
    > the only logical way that these spikes can occur that I can think of.
    > It never drops data, it always comes up with an extra. Also, the
    > amount it is off is always equal to one extra rev.
    >
    > I think if I do everything I can to ensure each pulse is received as
    > one (and not split) then I should be good. That is why I am looking
    > at signal strength and optimal direction.


    As far as signal strength goes it's easy to fix.

    Open up the speed or cadence sensor with a very small Philips screwdriver.
    Make sure it fits well before you start turning. The screws are located in
    such a way that you have to remove the sensor from the bike to get at them.
    Once it's open you will find a whachamacallit. Hmm. It's like a little
    bridge between two contact points. It can be set in three positions.
    I think if it's connecting the two outside pins you have the strongest
    transmission. Remember that battery will wear out faster.
    If you want to check the exact settings you can probably find them here.
    http://www.pursuit-performance.com.au/cgi-bin/dcforum/dcboard.cgi


    --
    Perre

    You have to be smarter than a robot to reply.
     
  7. "Chris" <[email protected]> wrote in message
    news:[email protected]
    >
    > "Tom Nakashima" <[email protected]> wrote in message
    >
    >
    > The errors can be
    > > corrected, usually it's one - three sections together.

    >
    > I know. Luckily my cadence does not spike and neither does heart rate.
    >
    > I believe I narrowed
    > > it down to the signal light transmitter sensor for changing stop lights

    at
    > > the turnaround junction of my training ride. At this junction, I'm

    > usually
    > > in the left had turning lane, waiting for the light to turn green to

    make
    > a
    > > u-turn.
    > >
    > > Just curious, do your spikes occur near long stoplight?
    > > -tom

    >
    > No, it is definitely unrelated to external interference. It even happens

    on
    > country roads miles from any EMF. I tend to think that somehow the speed
    > "pulse" gets interpreted by the receiver twice on a single wheel rev. I
    > think that if the signal is somehow interrupted and then restored all

    while
    > the transmitter is broadcasting its signal for that rev, that would

    account
    > for this. If this sounds too illogical let me explain it with an example.
    > Let's say that each pulse lasts 5/100s of a second. The whole system has

    to
    > have some fault tolerance so a signal that is too short is less likely to
    > get through. Again for fault tolerance and to keep costs down the receiver
    > does not care how long the signal is. It also is not set to reject any
    > signals no matter how close together they are. Now add to that a case

    where
    > the reception is marginal but still persistent. In cases where the spike
    > occurs, I am moving along and the transmitter is constantly changing its
    > placement relative to the receiver slightly. The transmitter sends its
    > signal for 5/100s and the receiver picks it up for the first (or second or
    > whatever) and then loses it but catches that last fraction. It counts it
    > twice since it recorded to pulses. Also keep in mind that most of the time
    > in a case like this, you would probably be persistently getting much less

    of
    > the signal than that ideal 5/100 of a second. If the signal is lost at the
    > beginning or the end or even on the middle, as long as it is not restored
    > before it completes it broadcast it will count it accurately. I think this
    > also explains why there is an interest in digital wireless cyclometers.
    >
    > I don't know how long each signal pulse is sent but this is really the

    only
    > logical way that these spikes can occur that I can think of. It never

    drops
    > data, it always comes up with an extra. Also, the amount it is off is

    always
    > equal to one extra rev.
    >
    > I think if I do everything I can to ensure each pulse is received as one
    > (and not split) then I should be good. That is why I am looking at signal
    > strength and optimal direction.
    >

    It is a possibility that the speed pulse gets interpreted by the receiver. I
    forward your "Re" to my friend who had that problem a few times on country
    roads, no powerlines, no traffic light signals. Might be due to bumpy
    roads, where the transmitter gets interpreted by violent bounce of the front
    wheel. Says he hasn't confirmed that yet, but told me if it were UFO
    transmissions, I wouldn't believe him.
    We called Polar, they don't have an explanation since it's intermittent, but
    ruled out the flying saucer theory.

    As for me, it's possible my spikes are caused by powerline transmission.
    There are lines at my turnaround junction as well as traffic light
    transmission. I may also look over the shoulder a few times, would hate to
    be "Taken".
    -tom
     
Loading...
Loading...