| Author |
Message |
john
Guest
|
 Posted:
Sun Dec 11, 2005 5:08 pm Post subject:
Antenna cable delay affect on position |
|
|
Hello
A question to the group
Let's suppose I have a GPS at point A it is connected via a 5metre cable to
the external antenna at point B 5metres away.
What position does the gps report A or B?
Further A and B still 5metres apart but now connected via 20metres of
cable - what position?
What difference if cable is screened or not screened?
If accuracy is a question we can increase the distances from 5 and 20 to 50
and 200 metres.
--
John |
|
| Back to top |
|
 |
Rob Kimberley
Guest
|
| Posted:
Sun Dec 11, 2005 5:08 pm Post subject:
Re: Antenna cable delay affect on position |
|
|
The position reported will always be that of the antenna. When using GPS for
timing purposes, delay compensations can be made for the cable, and a
nanosecond per foot is a pretty good estimate.
Rob
"john" <john.davies@rocketmail.com> wrote in message
news:kjmt63-gjg.ln1@news.infowest.com...
| Quote: |
Hello
A question to the group
Let's suppose I have a GPS at point A it is connected via a 5metre cable
to
the external antenna at point B 5metres away.
What position does the gps report A or B?
Further A and B still 5metres apart but now connected via 20metres of
cable - what position?
What difference if cable is screened or not screened?
If accuracy is a question we can increase the distances from 5 and 20 to
50
and 200 metres.
--
John
|
|
|
| Back to top |
|
|
Sam Wormley
Guest
|
| Posted:
Sun Dec 11, 2005 5:08 pm Post subject:
Re: Antenna cable delay affect on position |
|
|
john wrote:
| Quote: |
Let's suppose I have a GPS at point A it is connected via a 5metre cable to
the external antenna at point B 5metres away.
|
GPS receivers report the positions of their antennae. |
|
| Back to top |
|
|
john
Guest
|
| Posted:
Sun Dec 11, 2005 5:08 pm Post subject:
Re: Antenna cable delay affect on position |
|
|
Hej
What about cable delay - is it automatically compensated for? What would its
affect be on position if it were not compensated for?
Thanks for the help
"Sam Wormley" <swormley1@mchsi.com> wrote in message
news:xrWmf.620177$_o.155087@attbi_s71...
| Quote: | john wrote:
Let's suppose I have a GPS at point A it is connected via a 5metre
cable to
the external antenna at point B 5metres away.
GPS receivers report the positions of their antennae.
|
|
|
| Back to top |
|
|
Woody
Guest
|
| Posted:
Sun Dec 11, 2005 5:08 pm Post subject:
Re: Antenna cable delay affect on position |
|
|
Since the download speed of the gps is about 1k and the signal on the cable
runs close to the speed of light the delay is negligible...
"john" <john.davies@rocketmail.com> wrote in message
news:9uot63-h3h.ln1@news.infowest.com...
| Quote: |
Hej
What about cable delay - is it automatically compensated for? What would
its
affect be on position if it were not compensated for?
Thanks for the help
"Sam Wormley" <swormley1@mchsi.com> wrote in message
news:xrWmf.620177$_o.155087@attbi_s71...
john wrote:
Let's suppose I have a GPS at point A it is connected via a 5metre
cable to
the external antenna at point B 5metres away.
GPS receivers report the positions of their antennae.
|
|
|
| Back to top |
|
|
Guest
|
| Posted:
Sun Dec 11, 2005 5:08 pm Post subject:
Re: Antenna cable delay affect on position |
|
|
Apparently on date Sun, 11 Dec 2005 13:43:13 GMT, "john"
<john.davies@rocketmail.com> said:
| Quote: | Hello
A question to the group
Let's suppose I have a GPS at point A it is connected via a 5metre cable to
the external antenna at point B 5metres away.
What position does the gps report A or B?
|
Actually, neither, but it depends on the GPS firmware as to exactly how it
deals with this.
| Quote: | Further A and B still 5metres apart but now connected via 20metres of
cable - what position?
|
The GPS engine calculates position on the basis of signal timing. Therefore,
when the signal from SV15 arrives a fraction of a second earlier than SV23, the
GPS knows it is (fraction of a second) * (Speed of light) = (Meters) further
from SV23 than it is from SV15 at that moment.
When the signals also travel along a cable, it will still perceive this same
difference in the arrival time of each signal.
Trouble appears when you treat this as a 3D lock. You can consider this as a
simplistic model, where the signals arrive down the cable as if the unit were
five, or twenty, meters under the ground and the signals have that extra five
or twenty meters to travel to get there.
That sounds ok, but in practice the lines from each satellite to antenna, and
then down wire to GPS, cannot all be straight unless the geometry of the
satellites happens to be just right so that the extra five meters is the same
in each case. So, some error is necessarily introduced. Basically, in most
cases, the apparent position is impossible. But the system allows a fair bit of
slack so in practice will merely degrade the positional accuracy, usually by
not enough to worry about.
Simplest answer, the GPS will give position where the antenna is but below the
position by roughly the length of the cable. |
|
| Back to top |
|
|
john
Guest
|
| Posted:
Sun Dec 11, 2005 11:13 pm Post subject:
Re: Antenna cable delay affect on position |
|
|
<Nospam@nowhere.com> wrote in message
news:0kjop1h7jk8dvniu9h2tq05facrea90nd1@4ax.com...
| Quote: | Apparently on date Sun, 11 Dec 2005 13:43:13 GMT, "john"
john.davies@rocketmail.com> said:
Hello
A question to the group
Let's suppose I have a GPS at point A it is connected via a 5metre cable
to
the external antenna at point B 5metres away.
What position does the gps report A or B?
Actually, neither, but it depends on the GPS firmware as to exactly how it
deals with this.
Further A and B still 5metres apart but now connected via 20metres of
cable - what position?
The GPS engine calculates position on the basis of signal timing.
Therefore,
when the signal from SV15 arrives a fraction of a second earlier than
SV23, the
GPS knows it is (fraction of a second) * (Speed of light) = (Meters)
further
from SV23 than it is from SV15 at that moment.
When the signals also travel along a cable, it will still perceive this
same
difference in the arrival time of each signal.
Trouble appears when you treat this as a 3D lock. You can consider this as
a
simplistic model, where the signals arrive down the cable as if the unit
were
five, or twenty, meters under the ground and the signals have that extra
five
or twenty meters to travel to get there.
That sounds ok, but in practice the lines from each satellite to antenna,
and
then down wire to GPS, cannot all be straight unless the geometry of the
satellites happens to be just right so that the extra five meters is the
same
in each case. So, some error is necessarily introduced. Basically, in most
cases, the apparent position is impossible.
|
I had come to this conclusion myself - and thought that if one could
compensate for that time delay by subtracting that delay from each signal
then problem solved. It could be done by entering cable length or maybe even
automatically mesauring at power up.
Just my theory.
| Quote: | But the system allows a fair bit of
slack so in practice will merely degrade the positional accuracy, usually
by
not enough to worry about.
|
If I am using dgps or waas(egnos) when it becomes available (Sweden 61N
017E)
and am looking for better than 4metres accuracy -1metre maybe - then is it
significant?
| Quote: | Simplest answer, the GPS will give position where the antenna is but below
the
position by roughly the length of the cable.
|
Can you explain that a little better for me?
Thanks
John |
|
| Back to top |
|
|
Stichting ST
Guest
|
| Posted:
Mon Dec 12, 2005 12:02 am Post subject:
Re: Antenna cable delay affect on position |
|
|
For a normal standalone GPSr that has to make a absolute position, yes.
But when you are working relative as most surveyer GPS equipment does (base
station and rover) then you even need to know where the phase centre of your
antennas are. And you need expensive antennas that keep the phase centre
stable even when the GPS signals come from very sideways.
I was told that some geodetic antennas even have a different (measured) phase
centre for the frequency of L1 and L2. It all makes cm differences (especially
in height) over distances of 20 km.
Piet
On Sun, 11 Dec 2005 15:36:25 GMT, "Woody" <TheDuck@pond.net> wrote:
| Quote: | Since the download speed of the gps is about 1k and the signal on the cable
runs close to the speed of light the delay is negligible...
"john" <john.davies@rocketmail.com> wrote in message
news:9uot63-h3h.ln1@news.infowest.com...
Hej
What about cable delay - is it automatically compensated for? What would
its
affect be on position if it were not compensated for?
Thanks for the help
"Sam Wormley" <swormley1@mchsi.com> wrote in message
news:xrWmf.620177$_o.155087@attbi_s71...
john wrote:
Let's suppose I have a GPS at point A it is connected via a 5metre
cable to
the external antenna at point B 5metres away.
GPS receivers report the positions of their antennae.
|
|
|
| Back to top |
|
|
Frank
Guest
|
| Posted:
Mon Dec 12, 2005 12:12 am Post subject:
Re: Antenna cable delay affect on position |
|
|
:: If I am using dgps or waas(egnos) when it becomes available (Sweden
:: 61N 017E)
:: and am looking for better than 4metres accuracy -1metre maybe -
:: then is it significant?
::
Not even a tiny amount. If you were searching for a millimeter position, and
had a 100-meter cable, then yes, there'd be some offset.
If you had the need for doing such, you'd likely have the equipment to help
with such. 1-4 meters is handheld-averaging range with WAAS/EGNOS.
If you don't understand the explanations, ask for specific re-explaining.
The whole thing has been laid out pretty well. See especially the
"nanosecond per-foot of cable" comment. That should tell you a lot right
there.
Frank |
|
| Back to top |
|
|
Guest
|
| Posted:
Mon Dec 12, 2005 12:51 am Post subject:
Re: Antenna cable delay affect on position |
|
|
Apparently on date Sun, 11 Dec 2005 17:13:15 GMT, "john"
<john.davies@rocketmail.com> said:
| Quote: | Nospam@nowhere.com> wrote in message
Apparently on date Sun, 11 Dec 2005 13:43:13 GMT, "john"
john.davies@rocketmail.com> said:
Let's suppose I have a GPS at point A it is connected via a 5metre cable
to
the external antenna at point B 5metres away.
What position does the gps report A or B?
Actually, neither, but it depends on the GPS firmware as to exactly how it
deals with this.
That sounds ok, but in practice the lines from each satellite to antenna,
and
then down wire to GPS, cannot all be straight unless the geometry of the
satellites happens to be just right so that the extra five meters is the
same
in each case. So, some error is necessarily introduced. Basically, in most
cases, the apparent position is impossible.
I had come to this conclusion myself - and thought that if one could
compensate for that time delay by subtracting that delay from each signal
then problem solved. It could be done by entering cable length or maybe even
automatically mesauring at power up.
Just my theory.
|
Oh sure, if firmware supported an adjustment for cable and length then there'd
be no error and the position reported would be spot on where the antenna was.
That said, I've not seen one on the popular consumer stuff I use.
| Quote: | But the system allows a fair bit of
slack so in practice will merely degrade the positional accuracy, usually
by
not enough to worry about.
If I am using dgps or waas(egnos) when it becomes available (Sweden 61N
017E)
and am looking for better than 4metres accuracy -1metre maybe - then is it
significant?
|
It's on now, has been for years. I reckon it is about 1-2 meters most of the
time (no % statistics from me), from what I can determine. Which isn't too far
off reality, in fact, as I have reference points available, although not
everything that says it is using Egnos is actually accurate enough to persuade
me it is doing it.
It's virtually signed off for SOL stuff, should be weeks or months. So I
wouldn't expect it to be unreliable or intermittent again.
Effect on using cables or a rerad? Well, the error you introduce depends on the
geometry and cable length and I don't know what the range of errors are in
practice, but these will directly add to the uncertainty. WAAS can't help you
on that at all as it is correcting for slack in the space side and atmosphere
and stuff.
Therefore, if it was adding +/- 5M to a +/- 15M regular GPS uncertainty, it
will now be adding +/- 5M to a +/- 3M fix. So the added accuracy of WAAS /
EGNOS won't be wasted, but you would be adding a larger percentage error. I
think. As I say, it may be almost negligible unless you have a really long
cable.
| Quote: | Simplest answer, the GPS will give position where the antenna is but below
the
position by roughly the length of the cable.
Can you explain that a little better for me?
|
Assuming the GPS is measuring only the times at which signals arrive, the only
thing it can conclude is that the signal from a given satellite came in a
direct line from there to the GPS position. Since the antenna lead is
effectively an extension of the signal path the GPS must assume that the signal
continued on down the same line from up in the sky down to the ground,
penetrating underground in effect.
This depends on the geometry, but most geometries are with the satellites up
above the GPS and there are none below the horizon so most solutions will end
up below the plane of the horizon and none will be above.
IYSWIM. Basically, the positional solution will be roughly correct but will be
further away from the SVs than the pickup antenna is. |
|
| Back to top |
|
|
David Lee
Guest
|
| Posted:
Mon Dec 12, 2005 1:08 am Post subject:
Re: Antenna cable delay affect on position |
|
|
john wrote...
| Quote: | What about cable delay - is it automatically compensated for? What would
its
affect be on position if it were not compensated for?
|
Signals from the satellites travels at approximately the speed of light
(slightly slower 'cause it's not always in vacuuo) as does the signal in the
antenna cable. This means that the pseudo-ranges of all the satellites will
be increased by approximately the length of the cable (5m or whatever).
However, since the accurate time at the receiver is not known, the pseudo
ranges are all in error by a fixed amount anyway (which is why it's a
PSEUDO-range of course). The cable length error will be eliminated along
with the timing error when the receiver solves for position and time,
leaving you with a totally insignificant error in the estimate of the time -
of the order of 17ns for a 5m cable.
Hence there will be no effect on the position reported by the GPS.
David |
|
| Back to top |
|
|
john
Guest
|
| Posted:
Mon Dec 12, 2005 1:08 am Post subject:
Re: Antenna cable delay affect on position |
|
|
Thanks for your time and the explanation
John
<Nospam@nowhere.com> wrote in message
news:aorop1d35sv7jb1u1ep2s8lm605f6065il@4ax.com...
| Quote: | Apparently on date Sun, 11 Dec 2005 17:13:15 GMT, "john"
john.davies@rocketmail.com> said:
Nospam@nowhere.com> wrote in message
Apparently on date Sun, 11 Dec 2005 13:43:13 GMT, "john"
john.davies@rocketmail.com> said:
Let's suppose I have a GPS at point A it is connected via a 5metre
cable
to
the external antenna at point B 5metres away.
What position does the gps report A or B?
Actually, neither, but it depends on the GPS firmware as to exactly how
it
deals with this.
That sounds ok, but in practice the lines from each satellite to
antenna,
and
then down wire to GPS, cannot all be straight unless the geometry of
the
satellites happens to be just right so that the extra five meters is
the
same
in each case. So, some error is necessarily introduced. Basically, in
most
cases, the apparent position is impossible.
I had come to this conclusion myself - and thought that if one could
compensate for that time delay by subtracting that delay from each signal
then problem solved. It could be done by entering cable length or maybe
even
automatically mesauring at power up.
Just my theory.
Oh sure, if firmware supported an adjustment for cable and length then
there'd
be no error and the position reported would be spot on where the antenna
was.
That said, I've not seen one on the popular consumer stuff I use.
But the system allows a fair bit of
slack so in practice will merely degrade the positional accuracy,
usually
by
not enough to worry about.
If I am using dgps or waas(egnos) when it becomes available (Sweden 61N
017E)
and am looking for better than 4metres accuracy -1metre maybe - then is
it
significant?
It's on now, has been for years. I reckon it is about 1-2 meters most of
the
time (no % statistics from me), from what I can determine. Which isn't too
far
off reality, in fact, as I have reference points available, although not
everything that says it is using Egnos is actually accurate enough to
persuade
me it is doing it.
It's virtually signed off for SOL stuff, should be weeks or months. So I
wouldn't expect it to be unreliable or intermittent again.
Effect on using cables or a rerad? Well, the error you introduce depends
on the
geometry and cable length and I don't know what the range of errors are in
practice, but these will directly add to the uncertainty. WAAS can't help
you
on that at all as it is correcting for slack in the space side and
atmosphere
and stuff.
Therefore, if it was adding +/- 5M to a +/- 15M regular GPS uncertainty,
it
will now be adding +/- 5M to a +/- 3M fix. So the added accuracy of WAAS /
EGNOS won't be wasted, but you would be adding a larger percentage error.
I
think. As I say, it may be almost negligible unless you have a really long
cable.
Simplest answer, the GPS will give position where the antenna is but
below
the
position by roughly the length of the cable.
Can you explain that a little better for me?
Assuming the GPS is measuring only the times at which signals arrive, the
only
thing it can conclude is that the signal from a given satellite came in a
direct line from there to the GPS position. Since the antenna lead is
effectively an extension of the signal path the GPS must assume that the
signal
continued on down the same line from up in the sky down to the ground,
penetrating underground in effect.
This depends on the geometry, but most geometries are with the satellites
up
above the GPS and there are none below the horizon so most solutions will
end
up below the plane of the horizon and none will be above.
IYSWIM. Basically, the positional solution will be roughly correct but
will be
further away from the SVs than the pickup antenna is.
|
|
|
| Back to top |
|
|
David Lee
Guest
|
| Posted:
Mon Dec 12, 2005 1:08 am Post subject:
Re: Antenna cable delay affect on position |
|
|
Nospam@nowhere.com
| Quote: | Can you explain that a little better for me?
Assuming the GPS is measuring only the times at which signals arrive, the
only
thing it can conclude is that the signal from a given satellite came in a
direct line from there to the GPS position. Since the antenna lead is
effectively an extension of the signal path the GPS must assume that the
signal
continued on down the same line from up in the sky down to the ground,
penetrating underground in effect.
This depends on the geometry, but most geometries are with the satellites
up
above the GPS and there are none below the horizon so most solutions will
end
up below the plane of the horizon and none will be above.
IYSWIM. Basically, the positional solution will be roughly correct but
will be
further away from the SVs than the pickup antenna is.
|
No. The GPS measures the time at which signals arrive according to its own,
inaccurate, internal clock. Thus it doesn't know the exact time when
signals arrive. Each arrival time is in error by the same time. Adding a
length of antenna cable adds another bit of delay but as far as the GPS is
concerned this is no different from the uncertainty due to the clock error,
since the same delay is added to each satellite signal. Effectively the GPS
solves a system of simultaneous equations to determine the position in three
dimensions and the accurate time (although actually it uses Kalman filter
techniques). The effect of the cable-delay will be to add an extremely
small error to the time but the positional estimate will still be correctly
derived for the point at which all the signals were received together - at
the antenna.
David |
|
| Back to top |
|
|
peter
Guest
|
| Posted:
Mon Dec 12, 2005 1:08 am Post subject:
Re: Antenna cable delay affect on position |
|
|
David Lee wrote:
| Quote: | Effectively the GPS
solves a system of simultaneous equations to determine the position in three
dimensions and the accurate time (although actually it uses Kalman filter
techniques). The effect of the cable-delay will be to add an extremely
small error to the time but the positional estimate will still be correctly
derived for the point at which all the signals were received together - at
the antenna.
|
Agreed, but there is a very slight effect on the positional accuracy
since the ephemeris data predicts the satellite position based on
knowing the accurate actual time rather than the slightly offset time.
For a 200m cable length the satellite will move a little under half a
centimeter in the time it takes the signals to traverse the cable. I'd
therefore expect the calculated position to have an additional error of
about that magnitude - i.e. negligible for typical applications. |
|
| Back to top |
|
|
john
Guest
|
| Posted:
Wed Dec 14, 2005 4:13 pm Post subject:
Re: Antenna cable delay affect on position |
|
|
Thanks again - It seems that David Lee and Peter have explained it very
well for me - along with the knowledge I have gained from reading the other
posts.
Once again thanks to all that contributed
John
"peter" <prathman@comcast.net> wrote in message
news:1134341974.405756.229070@g44g2000cwa.googlegroups.com...
| Quote: | David Lee wrote:
Effectively the GPS
solves a system of simultaneous equations to determine the position in
three
dimensions and the accurate time (although actually it uses Kalman
filter
techniques). The effect of the cable-delay will be to add an extremely
small error to the time but the positional estimate will still be
correctly
derived for the point at which all the signals were received together -
at
the antenna.
Agreed, but there is a very slight effect on the positional accuracy
since the ephemeris data predicts the satellite position based on
knowing the accurate actual time rather than the slightly offset time.
For a 200m cable length the satellite will move a little under half a
centimeter in the time it takes the signals to traverse the cable. I'd
therefore expect the calculated position to have an additional error of
about that magnitude - i.e. negligible for typical applications.
|
|
|
| Back to top |
|
|
|
|
|
|
FAQ
Memberlist
Register
Profile
Log in to check your private messages
Log in