Re: Garmin Edge 305 Cyclometer

[Eric]

Garmin seems to keep a premium price on thier products, but it is
mostly deserved. I had a Magellan GPS years ago and they weren't even
close to the equivalant Gamin unit. Since then I haven't considered
anything else.

 

[[email protected]]

the unit does not track speed and speed derivitives via satellite data
but regular old ground data-is that correct?

 

[Richard Bollar]

[email protected] wrote:
> the unit does not track speed and speed derivitives via satellite data
> but regular old ground data-is that correct?

No, that's not completely correct. The Edge uses GPS for speed data
where possible, but it will revert to the GSC-10 speed sensor when a
satellite signal is not available (tunnels, heavy tree cover, etc.)

 

[Sandy]

Dans le message de
news:[email protected],
Richard Bollar <[email protected]> a réfléchi, et puis a déclaré :
> [email protected] wrote:
>> the unit does not track speed and speed derivitives via satellite
>> data but regular old ground data-is that correct?
>
> No, that's not completely correct. The Edge uses GPS for speed data
> where possible, but it will revert to the GSC-10 speed sensor when a
> satellite signal is not available (tunnels, heavy tree cover, etc.)

This is a great subject about which I know nothing. It's nice to learn all
this stuff.
But, why would you want a GPS to measure ground covered by your wheels at
the relaxed accuracies of a satellite, when you down-to-the-millimetric
wheel kit will do better ? Or maybe there is an option to choose one or the
other on this equipment ??

--
Sandy
--
Ce message, issu de l'agriculture biologique, a été rédigé avec des
électrons recyclés. En conséquence, il est possible que des fautes
d'orthographes s'y soient malencontreusement glissées.

 

[Neil Brooks]

"Sandy" <[email protected]> wrote:

>Dans le message de
>news:[email protected],
>Richard Bollar <[email protected]> a réfléchi, et puis a déclaré :
>> [email protected] wrote:
>>> the unit does not track speed and speed derivitives via satellite
>>> data but regular old ground data-is that correct?
>>
>> No, that's not completely correct. The Edge uses GPS for speed data
>> where possible, but it will revert to the GSC-10 speed sensor when a
>> satellite signal is not available (tunnels, heavy tree cover, etc.)
>
>This is a great subject about which I know nothing. It's nice to learn all
>this stuff.
>But, why would you want a GPS to measure ground covered by your wheels at
>the relaxed accuracies of a satellite, when you down-to-the-millimetric
>wheel kit will do better ? Or maybe there is an option to choose one or the
>other on this equipment ??

This is definitely but /one/ of the benefits of using a GPS for
cycling.

I enjoy using it to create an exact record of the route that I rode,
then uploading the data to software like TopoZone [1] to create a 3-D
version of the ride, allowing me to view grades climbed, calculate
power output, and generally log interesting rides.

Another fun advantage is the ability to 'get yourself hopelessly
lost,' then--taking the GPS out of your backpack--find your way back
with ease.

[1] http://www.topozone.com/
--
Live simply so that others may simply live

 

[Richard Bollar]

Sandy wrote:


> This is a great subject about which I know nothing. It's nice to learn all
> this stuff.
> But, why would you want a GPS to measure ground covered by your wheels at
> the relaxed accuracies of a satellite, when you down-to-the-millimetric
> wheel kit will do better ? Or maybe there is an option to choose one or the
> other on this equipment ??

Well, first off, second-to-second GPS distance is very accurate as it's
calculated from the GPS frequency doppler shift and isn't dependent on
the accuracy of the position solution. The Edge then uses these
calculations over roughly 2km to determine the circumference of the
wheel. Unless you turn the GPS off, it will always use GPS signal
first and speed sensor second.

Now, why would one want to pay so much for a bike computer... I like
being able to map out my rides, and I also like to be able to compete
against myself. The Edge has a feature that allows you to show your
progress against a prior workout in real time. A very cool feature.

 

[peter]

Richard Bollar wrote:
> Sandy wrote:
> > But, why would you want a GPS to measure ground covered by your wheels at
> > the relaxed accuracies of a satellite, when you down-to-the-millimetric
> > wheel kit will do better ? Or maybe there is an option to choose one or the
> > other on this equipment ??
>
> Well, first off, second-to-second GPS distance is very accurate as it's
> calculated from the GPS frequency doppler shift and isn't dependent on
> the accuracy of the position solution.

The Doppler shift is the primary input in determining the current
speed. But using only speed data and integrating over time to get
distance traveled would result in significantly larger errors than
considering the direct measurements of successive positions. In
practice the Kalman filter computational techniques used in GPS
receivers tend to incorporate all the available input data (current
pseudo ranges and pseudo range-rates as well as recent history) to
produce the best estimate of the outputs (current position, speed,
heading, and distance moved)

> The Edge then uses these
> calculations over roughly 2km to determine the circumference of the
> wheel. Unless you turn the GPS off, it will always use GPS signal
> first and speed sensor second.

Seems like a poor design decision. I've been using Garmin GPS
receivers in lieu of a cyclometer for quite a few years now and find
that the odometer functions are adequate for my purposes but are
clearly inferior to a well-calibrated cyclometer that counts wheel
revolutions to determine the distance. Responsiveness of the speed
indication to changes is also limited compared to a regular cyclometer
due to the once per second update rate of most consumer GPS units
(incl. the 205/305 according to the specs I've seen).

However, the magnet sensor that lets the unit determine speed &
distance by wheel revolutions is an optional extra in some versions of
this product so they may have wanted to keep the operation of them
consistent by using GPS measurements as much as possible, even when
less accurate.

 

[Lou Holtman]

"Richard Bollar" <[email protected]> wrote in message
news:[email protected]...
>
> Sandy wrote:
>
>
> > This is a great subject about which I know nothing. It's nice to learn
all
> > this stuff.
> > But, why would you want a GPS to measure ground covered by your wheels
at
> > the relaxed accuracies of a satellite, when you down-to-the-millimetric
> > wheel kit will do better ? Or maybe there is an option to choose one or
the
> > other on this equipment ??
>
> Well, first off, second-to-second GPS distance is very accurate as it's
> calculated from the GPS frequency doppler shift and isn't dependent on
> the accuracy of the position solution. The Edge then uses these
> calculations over roughly 2km to determine the circumference of the
> wheel. Unless you turn the GPS off, it will always use GPS signal
> first and speed sensor second.
>
> Now, why would one want to pay so much for a bike computer... I like
> being able to map out my rides, and I also like to be able to compete
> against myself. The Edge has a feature that allows you to show your
> progress against a prior workout in real time. A very cool feature.
>

Can you use the Garmin Edge 305 without the speedsensor mounted?

Lou

 

[michel]

>
> Can you use the Garmin Edge 305 without the speedsensor mounted?
>

I have the Edge 205, and it comes without a speed censor, and as far as
I know the Edge 305 comes without one as well, here in the UK.

As stated before: the speed the Edge gives, seems less stable than with
a standard bike computer - possibly more so if going slow: e.g. when I
was climbing the speed the Edge gave was continuously varying between
13 and 16 km/h. I am quite sure with a normal bike computer I would see
differences of about +/- 0.5 km/h.

With the Edge 205 (without barometric altitude meter) the height
fluctuates rapidly by something like +/- 10m. Consequently the grade it
gives does not make much sense.

At the moment I have to say I am a bit dissapointed with the Edge. The
unit itself (hardware) seems okay (though I would have prefered
replaceable batteries), but the whole unit seems to be set up more as
an expensive trip-computer than a navigation tool. The Trainingscenter
software works okay, but is extremely limited in functionality and only
exports data in an uncommon .hst file (and all data in one file). The
user manual is very very limited. That being said, I am still trying to
get things to work the way I want, so maybe it will be okay after all.

Anyone succeeded in uploading "Routes" from affordable software to the
Edge? If I understand correctly the Garmin "Trip and Waypoint manager"
is not compatible with the Edge. Although I am not willing to spend
another £150 for the mapsource UK topo, would this give you this
functionality? Is it at all possible? I managed to upload waypoints
with 3rd party software, but these still need manual editing (on the
Edge itself) to get them into routes, which will involve lots of
button-pushing.

Also: I thought this would be completely obvious, but where can you see
the actual position where you are "at the moment"? Is it possible to
set the datum such that the position it gives corresponds to the
"Ordnance Survey" map coordinates?

Finally: I believe the unit does not have WAAS, but have heared some
conficting info - anyone know more about this?

Cheers!
Michel

 

[michel]

Richard Bollar wrote:
> [...] The Edge then uses these calculations over roughly 2km to determine the circumference of the wheel. [...]

How do you know these things? I am quite sure it is not in the user
manual?

Are there any other sources of info, except the user manual?

Cheers!
Michel

 

[Richard Bollar]

michel wrote:
> Richard Bollar wrote:
> > [...] The Edge then uses these calculations over roughly 2km to determine the circumference of the wheel. [...]
>
> How do you know these things? I am quite sure it is not in the user
> manual?
>
> Are there any other sources of info, except the user manual?

Well, I own an Edge 305, so that's how I know. When you ride, you get
a message about 2km into the ride that says something like "wheel size
calibrated." The size of the wheel does vary by a couple of mm on each
ride. Seems like changes in air pressure or temperature would easily
account for that variance.

 

[Richard Bollar]

Lou Holtman wrote:
>
> Can you use the Garmin Edge 305 without the speedsensor mounted?
>

Definitely. In fact there's a version sold without the speed sensor.

 

[Richard Bollar]

michel wrote:

> Anyone succeeded in uploading "Routes" from affordable software to the
> Edge? If I understand correctly the Garmin "Trip and Waypoint manager"
> is not compatible with the Edge. Although I am not willing to spend
> another £150 for the mapsource UK topo, would this give you this
> functionality? Is it at all possible? I managed to upload waypoints
> with 3rd party software, but these still need manual editing (on the
> Edge itself) to get them into routes, which will involve lots of
> button-pushing.

I am told that if you have Training Center, that you can download
MapSource directly from Garmin. Since I already own MapSource, I can't
test. http://www.garmin.com/support/mappingsw.jsp

> Also: I thought this would be completely obvious, but where can you see
> the actual position where you are "at the moment"? Is it possible to
> set the datum such that the position it gives corresponds to the
> "Ordnance Survey" map coordinates?

No on both counts. I assume you're locked into WGS84.

> Finally: I believe the unit does not have WAAS, but have heared some
> conficting info - anyone know more about this?

I have seen no evidence that WAAS is supported on the Edge. None of
the WAAS signals are shown on the satellite screen.

You might also be interested in this page I created which shows some
comparison data between the Edge and various other Garmin GPS
receivers: http://www.bollar.org/edge305.htm

 

[David L. Johnson]

On Wed, 01 Feb 2006 19:19:06 -0800, peter wrote:

> The Doppler shift is the primary input in determining the current
> speed. But using only speed data and integrating over time to get
> distance traveled would result in significantly larger errors than
> considering the direct measurements of successive positions.

I don't see that at all. The path taken is not going to be a straight,
level line between marked positions along the road. Hills will cause one
sort of inaccuracy in using GPS data to determine distance, but the
actual path taken will be a larger error. Now, it may be debatable which
measurement you want, but certainly integrating speed over time gives a
better measurement of distance actually traveled. Of course, a standard
computer should find distance by simply counting wheel revolutions, which
is more likely to be accurate than either, if it is set up properly.

--

David L. Johnson

__o | The motor car reflects our standard of living and gauges the
_`\(,_ | speed of our present life. It long ago ran down Simple Living,
(_)/ (_) | and never halted to inquire about the prostrate figure which
fell as its victim. -- Warren G. Harding

 

[peter]

David L. Johnson wrote:
> On Wed, 01 Feb 2006 19:19:06 -0800, peter wrote:
>
> > The Doppler shift is the primary input in determining the current
> > speed. But using only speed data and integrating over time to get
> > distance traveled would result in significantly larger errors than
> > considering the direct measurements of successive positions.
>
> I don't see that at all. The path taken is not going to be a straight,
> level line between marked positions along the road. Hills will cause one
> sort of inaccuracy in using GPS data to determine distance, but the
> actual path taken will be a larger error. Now, it may be debatable which
> measurement you want, but certainly integrating speed over time gives a
> better measurement of distance actually traveled.

I think you missed the word "only" in my sentence above. The Kalman
filter calculations of GPS units take into account both the velocity
(Doppler) data and the position (range) data in trying to minimize the
errors. Using *only* the Doppler data will result in larger errors.

> Of course, a standard
> computer should find distance by simply counting wheel revolutions, which
> is more likely to be accurate than either, if it is set up properly.

Agreed, and that was the main point of my post. If the unit has a
wheel revolution counter then it seems silly not to use that data as
the primary basis for both distance and speed measurements.

 

[Richard Bollar]

peter wrote:

> > Of course, a standard
> > computer should find distance by simply counting wheel revolutions, which
> > is more likely to be accurate than either, if it is set up properly.
>
> Agreed, and that was the main point of my post. If the unit has a
> wheel revolution counter then it seems silly not to use that data as
> the primary basis for both distance and speed measurements.

What's the best way to do this? I've never been able to get better
than 2-3% accuracy with my cycle computers. Then again, this was with
various models of Polar HRMs which I have found to be really touchy.

 

[Lou Holtman]

Richard Bollar wrote:
> peter wrote:
>
>
>>> Of course, a standard
>>>computer should find distance by simply counting wheel revolutions, which
>>>is more likely to be accurate than either, if it is set up properly.
>>
>>Agreed, and that was the main point of my post. If the unit has a
>>wheel revolution counter then it seems silly not to use that data as
>>the primary basis for both distance and speed measurements.
>
>
> What's the best way to do this? I've never been able to get better
> than 2-3% accuracy with my cycle computers.

A typical value of a wheelcircumference is 2100 mm. 2-3% of that is
42-63 mm. That's a lot. You must be doing something wrong measuring the
wheelcircumference.

> Then again, this was with
> various models of Polar HRMs which I have found to be really touchy.

What do you mean by touchy?

Lou

--
Posted by news://news.nb.nu

 

[peter]

Richard Bollar wrote:
> peter wrote:
>
> > > Of course, a standard
> > > computer should find distance by simply counting wheel revolutions, which
> > > is more likely to be accurate than either, if it is set up properly.
> >
> > Agreed, and that was the main point of my post. If the unit has a
> > wheel revolution counter then it seems silly not to use that data as
> > the primary basis for both distance and speed measurements.
>
> What's the best way to do this? I've never been able to get better
> than 2-3% accuracy with my cycle computers.

I put a dab of ink or paint on the tire of the wheel with the magnet,
sit on the bike in a normal riding position and roll forward for a few
tire revolutions. Then I use a tape measure to determine the distance
of 3 or 4 circumferences and divide.
Seems to be accurate to 0.5% or better based on comparisons with
mileage markers on country roads. But I haven't used a regular
cyclometer in quite a few years since the GPS measurements are close
enough and I like to see where I am on the moving map.

 

[michel]

Richard Bollar wrote:
> I am told that if you have Training Center, that you can download
> MapSource directly from Garmin. Since I already own MapSource, I can't
> test. http://www.garmin.com/support/mappingsw.jsp
>
Thanks, will give it a try.

> > Also: I thought this would be completely obvious, but where can you see
> > the actual position where you are "at the moment"? Is it possible to
> > set the datum such that the position it gives corresponds to the
> > "Ordnance Survey" map coordinates?
>
> No on both counts. I assume you're locked into WGS84.

So I bought a GPS that cannot tell me where I am?!?

> You might also be interested in this page I created which shows some
> comparison data between the Edge and various other Garmin GPS
> receivers: http://www.bollar.org/edge305.htm

Nice page!

 

[Richard Bollar]

Lou Holtman wrote:
> Richard Bollar wrote:

> > What's the best way to do this? I've never been able to get better
> > than 2-3% accuracy with my cycle computers.
>
> A typical value of a wheelcircumference is 2100 mm. 2-3% of that is
> 42-63 mm. That's a lot. You must be doing something wrong measuring the
> wheelcircumference.

Possibly, but it's +/- 2-3 percent. I ride the same 26.1 mile loop day
after day and the readings from the Polar are usually between 25.9 to
26.7 with a fair number of readings outside that range. This is on two
different bikes each of whcih I have calibrated separately. I am aware
that I could be impacting the values by not making sure my tires are
inflated consistently.

Before the Edge, I didn't often take a GPS with me, so I only have 14
samples, but the readings are all between 26.07 and 26.16

>
> > Then again, this was with
> > various models of Polar HRMs which I have found to be really touchy.
>
> What do you mean by touchy?

I find that it's difficult to get the sensor placement right so that
the Polar will read the speed and cadence sensor at the same time.
Also, since it's a wireless system, I always assumed that interference
could be messing with the signal. I gave up trying to improve the
value, as it was good enough for my purposes.