|
Hi Group, i received my new OLG1 kit this week and constructed it yesterday.
It is working ok but i have noticed that there is quiet a drift in Altitude within a 2 hour time period.
As my QTH is only a couple of hundred meters from the water, South Coast of South Australia.
I am noticing that it has quiet a large drift from -4.5m - 40m above sea level.
Just a bit concerned that this is normal for this amount of drift or is there a issue with the components.
Regards
73
Stuart Vk5ade
|
Stuart Vk5ade
When you first turn on a GPS receiver it has to download a lot of information (called
ephemeris data) from the satellites.? Since this is sent in slow bursts between other
satellite information, it can take over an hour in some of the older GPS chipsets to
complete.? Newer GPS chips are faster, but there is some speed limitation due to
limited transmission rate of this secondary information from the satellites.?
Once the ephemeris data is complete and processed most of the GPS measurements
should stabilize.? I just pulled up the altitude, speed, and heading page on my local
GPS unit.? It took 12 seconds to give any display, and this is with ephemeris data already
loaded.? Speed is moving back and forth between "0" and 1 MPH, although the antenna is
lying here motionless on my desk.? Heading shows 13-degrees, but this is possibly an
artifact of atmospherics because the unit is not moving.? Altitude was changing between
2768 and 2714 feet but when I sat the antenna in the window this stabilized at between
2768 and 2765 feet.?
Hmmm...when I turned off the fan next to the window the speed immediately dropped to
zero and stayed there.? Heading is still 13 degrees, and altitude has stabilized to 2645 feet.
It is possible to use software to average GPS readings to something that may be more
accurate in the long term.? DGPS, or Differential GPS, does something similar by using a
known exact location reference and comparing that with that received from the GPS receiver.
With both received and known locations it is possible to calculate correction factor and
arrive at location measurements that are very accurate.? Surveyors and construction
architects usually use GPS and DGPS correction for height and position measurements.?
Arv
_._
On Thu, Jul 5, 2018 at 6:57 PM Stuart Cameron Vk5ade < stuart@...> wrote: Hi Group, i received my new OLG1 kit this week and constructed it yesterday.
It is working ok but i have noticed that there is quiet a drift in Altitude within a 2 hour time period.
As my QTH is only a couple of hundred meters from the water, South Coast of South Australia.
I am noticing that it has quiet a large drift from -4.5m - 40m above sea level.
Just a bit concerned that this is normal for this amount of drift or is there a issue with the components.
Regards
73
Stuart Vk5ade
|
Your fan was acting as a reflector and causing doppler shift of the received GPS signal.? ;) Pretty cool, eh?
On Thu, Jul 5, 2018 at 9:48 PM, Arv Evans <arvid.evans@...> wrote: Stuart Vk5ade
When you first turn on a GPS receiver it has to download a lot of information (called
ephemeris data) from the satellites.? Since this is sent in slow bursts between other
satellite information, it can take over an hour in some of the older GPS chipsets to
complete.? Newer GPS chips are faster, but there is some speed limitation due to
limited transmission rate of this secondary information from the satellites.?
Once the ephemeris data is complete and processed most of the GPS measurements
should stabilize.? I just pulled up the altitude, speed, and heading page on my local
GPS unit.? It took 12 seconds to give any display, and this is with ephemeris data already
loaded.? Speed is moving back and forth between "0" and 1 MPH, although the antenna is
lying here motionless on my desk.? Heading shows 13-degrees, but this is possibly an
artifact of atmospherics because the unit is not moving.? Altitude was changing between
2768 and 2714 feet but when I sat the antenna in the window this stabilized at between
2768 and 2765 feet.?
Hmmm...when I turned off the fan next to the window the speed immediately dropped to
zero and stayed there.? Heading is still 13 degrees, and altitude has stabilized to 2645 feet.
It is possible to use software to average GPS readings to something that may be more
accurate in the long term.? DGPS, or Differential GPS, does something similar by using a
known exact location reference and comparing that with that received from the GPS receiver.
With both received and known locations it is possible to calculate correction factor and
arrive at location measurements that are very accurate.? Surveyors and construction
architects usually use GPS and DGPS correction for height and position measurements.?
Arv
_._
On Thu, Jul 5, 2018 at 6:57 PM Stuart Cameron Vk5ade < stuart@...> wrote: Hi Group, i received my new OLG1 kit this week and constructed it yesterday.
It is working ok but i have noticed that there is quiet a drift in Altitude within a 2 hour time period.
As my QTH is only a couple of hundred meters from the water, South Coast of South Australia.
I am noticing that it has quiet a large drift from -4.5m - 40m above sea level.
Just a bit concerned that this is normal for this amount of drift or is there a issue with the components.
Regards
73
Stuart Vk5ade
|
Hi every one and thank you for this information, as this is the first OLG1 GPS i have made and used i was not sure of the accuracy.
This is interesting? as i was not sure what to expect.
My GPS is hanging in mid about a foot of the work bench above the U3S i have a 3 meter shielded lead attached so i can move it if needed for better accuracy.
I would be most interested in the software you are using.
If you can please advise which would be best software this would be most helpful as i would like to understand more of what it is actually telling me.
Kind Regards
Stuart Vk5ade
?
|
To add to the comments from Arv-
Any GPS/GNSS receiver will find its position in a series of single fixes, each resulting in Time, Lat, Long, and Height.
The position is resolved for Time, Latitude, and Longitude. the computing of Height is used as a trade off allowing the Lat and Long to be most accurate. The Height will not normally be as accurate as Lat and Long and can wander around, esp. when the signal path is affected by multipath or whatever.
Another factor is that the position fix is resolved from a collection of Lines Of Position, (LOP) each coming from one of those satellites in view. These LOPs will seldom intersect exactly over each other, in fact estimating the small errors of each LOP in each fix is a major part of the computation process. In successive fixes these LOPs will be seen to move to-and-fro, as the small irregularities of transmission have effect as well as noise. See 'Least Mean Squares Solution'.
Also the geometry of these LOPS has a major effect. If the shape generated by the LOPs is regular and approaches a circle/spheroid then errors are small and accuracy much better.
If the shape is elongated, a triangle, then some of the errors are large and accuracy be poorer, and remember these effects are ALL occurring in 3 dimensions. The result will easily appear to jump around in these circumstances, Maybe as seen by Stuart?
These small errors or residuals are labelled 'Dilution of Precision' or DOPs, the HDOP values and GDOP values are the most commonly used quality figures .
Regarding changes from fix to fix. 'Normal' GNSS receivers are made to plot from A-B so they will always look for the direction and speed towards the next points, or Target. They will hunt for this and the heading and speed values will constantly change a little to-and-fro.
Prof GNSS units for surveying can treat the data in a different manner and then they tend towards a static solution.?
GNSS units can easily be affected by digital noise from other units so one should aim to get a good separation from any of these other units, one can't say at least XXcm but instead 'the further, the better'.? I use a 1mtr minimum.
Hans' use of the small ground plane with this QLG1 means one can place these directly over a digital unit at smaller separations.
Spheroids. These are mathematical models that allow us to make computations and get reasonable answers. They are NOT 100% accurate shapes for good old Earth. The Earth has an irregular shape with some bits sticking out and some bits sunken in, depending on gravity and density of the rocks. So one can well see negative height values in some areas even when one is sitting on a beach. E.g. the Dead Sea area is about 200m below mean sea level (on the spheroid). After all they are only numbers, the reality is around you!
Accuracy?? Read the specs of the systems and the receivers used and you'll see - as good as +/-10m in Lat/Long for a single fix, Height will be more.
These are general worst case values, and can be easily improved on, with monitoring over time, with sophisticated statistical analysis, or with using something better than a QLG1. ?
The QLG1 are nice little units, I have three, but not quite up to Topcon or Trimble units that use antennas 10 times as big.
For deeper understanding one can TRY reading-
Heavy stuff I'm afraid, and there is lots and lots more listed in the Reference sections.
|
It may be possible to imprvoe the accuracy if the module supports a 'survey' mode but that can take days to run...
On 13 July 2018 at 15:12, Peter LB0K <lb0k@...> wrote: To add to the comments from Arv-
Any GPS/GNSS receiver will find its position in a series of single fixes, each resulting in Time, Lat, Long, and Height.
The position is resolved for Time, Latitude, and Longitude. the computing of Height is used as a trade off allowing the Lat and Long to be most accurate. The Height will not normally be as accurate as Lat and Long and can wander around, esp. when the signal path is affected by multipath or whatever.
Another factor is that the position fix is resolved from a collection of Lines Of Position, (LOP) each coming from one of those satellites in view. These LOPs will seldom intersect exactly over each other, in fact estimating the small errors of each LOP in each fix is a major part of the computation process. In successive fixes these LOPs will be seen to move to-and-fro, as the small irregularities of transmission have effect as well as noise. See 'Least Mean Squares Solution'.
Also the geometry of these LOPS has a major effect. If the shape generated by the LOPs is regular and approaches a circle/spheroid then errors are small and accuracy much better.
If the shape is elongated, a triangle, then some of the errors are large and accuracy be poorer, and remember these effects are ALL occurring in 3 dimensions. The result will easily appear to jump around in these circumstances, Maybe as seen by Stuart?
These small errors or residuals are labelled 'Dilution of Precision' or DOPs, the HDOP values and GDOP values are the most commonly used quality figures .
Regarding changes from fix to fix. 'Normal' GNSS receivers are made to plot from A-B so they will always look for the direction and speed towards the next points, or Target. They will hunt for this and the heading and speed values will constantly change a little to-and-fro.
Prof GNSS units for surveying can treat the data in a different manner and then they tend towards a static solution.?
GNSS units can easily be affected by digital noise from other units so one should aim to get a good separation from any of these other units, one can't say at least XXcm but instead 'the further, the better'.? I use a 1mtr minimum.
Hans' use of the small ground plane with this QLG1 means one can place these directly over a digital unit at smaller separations.
Spheroids. These are mathematical models that allow us to make computations and get reasonable answers. They are NOT 100% accurate shapes for good old Earth. The Earth has an irregular shape with some bits sticking out and some bits sunken in, depending on gravity and density of the rocks. So one can well see negative height values in some areas even when one is sitting on a beach. E.g. the Dead Sea area is about 200m below mean sea level (on the spheroid). After all they are only numbers, the reality is around you!
Accuracy?? Read the specs of the systems and the receivers used and you'll see - as good as +/-10m in Lat/Long for a single fix, Height will be more.
These are general worst case values, and can be easily improved on, with monitoring over time, with sophisticated statistical analysis, or with using something better than a QLG1. ?
The QLG1 are nice little units, I have three, but not quite up to Topcon or Trimble units that use antennas 10 times as big.
For deeper understanding one can TRY reading-
Heavy stuff I'm afraid, and there is lots and lots more listed in the Reference sections.
-- Clint. M0UAW IO83
No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.
|
It may be possible to imprvoe the accuracy if the module supports a 'survey' mode but that can take days to run... I think this is in PC software, not in a GPS module? 73 Alan G4ZFQ
|
Improving accuracy could be accomplished by establishing your own local DGPS (Differential
GPS station) using a second GPS receiver.? The DGPS concept involves using a GPS receiver
that is located at a proven fixed location and comparing information received from satellites
versus the known location.? The differences are then used as dynamic correction factors to be
applied to the locations given by the portable GPS receiver unit.? This provides a way to correct
for atmospherics, signal reflections, and other strange phenomena.
You may have a nearby DGPS station that can be used.? They are usually on an LF frequency
(the closest one here is on 325 KHz).?
Arv
_._
On Fri, Jul 13, 2018 at 10:24 AM Alan G4ZFQ < alan4alan@...> wrote: > It may be possible to imprvoe the accuracy if the module supports a
> 'survey' mode but that can take days to run...
I think this is in PC software, not in a GPS module?
73 Alan G4ZFQ
|
Pretty sure it¡¯s a feature of the GPS module but initiated by the controlling device
> It may be possible to imprvoe the accuracy if the module supports a
> 'survey' mode but that can take days to run...
I think this is in PC software, not in a GPS module?
73 Alan G4ZFQ
-- Clint. M0UAW IO83
No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.
|
Arv
This subject seems to drift off thread so I'll limit myself to just this reply.
Setting up a new ref station for a DGPS service would be an interesting project. To whit, getting a new station with an ten-fold improvement in accuracy AND regularity, licensed, working and on the air/active.
The DGPS services available on LF in the USA require a dedicated decoder, as far as I know and have experienced.
Are there any free solutions available?
And are there any other generally available services, and associated decoders, in other countries? I throw this question out to a wider circle.
We're using the QLG1 to regulate a transceiver frequency and synchronise it for WSPR transmissions, small variations of Lat, Long, and Height are of little importance but of great academic/radio amateur interest.? (I wanna fix it! Is it broke?)
?
|
Peter LB0K
I do not know of any free DGPS code that might be available, but Black Cat Systems
does have information available that could help to build something workable:
Many years ago I used Garmin GPS receivers to do something similar.? We were
not interested in all the facets of DGPS so not all the functions were available
in our systems.? Today we have the larger screens being used for BITX projects so
this might drive use for a more complete GPS with DGPS capability.
I would start with a fixed location GPS receiver and attached processor to store and
average the data received over time from several satellites.? With sufficient averaged
data this could give the actual physical location of that GPS unit.? Each subsequent
stanza from this fixed location receiver could then be compared with the averaged data
to generate a correction factor for singular data stanzas.? By using a communications
link it should then be possible to apply this correction factor to mobile GPS units in the
vicinity of the fixed unit.? Processing all the information may be beyond the capability
of an Arduino but if inputs were limited it just might be possible.
You are correct that there are a lot of aspects related to building your own Amateur
radio DGPS? (ADGPS...?) for making GPS more accurate, and this may not be of
interest to most on this group.? For this reason I too have limited my response.?
Arv? K7HKL
_._
On Sun, Jul 15, 2018 at 1:03 AM Peter LB0K < lb0k@...> wrote: Arv
This subject seems to drift off thread so I'll limit myself to just this reply.
Setting up a new ref station for a DGPS service would be an interesting project. To whit, getting a new station with an ten-fold improvement in accuracy AND regularity, licensed, working and on the air/active.
The DGPS services available on LF in the USA require a dedicated decoder, as far as I know and have experienced.
Are there any free solutions available?
And are there any other generally available services, and associated decoders, in other countries? I throw this question out to a wider circle.
We're using the QLG1 to regulate a transceiver frequency and synchronise it for WSPR transmissions, small variations of Lat, Long, and Height are of little importance but of great academic/radio amateur interest.? (I wanna fix it! Is it broke?)
?
|
