GPS setting with Memory Map

Leonard Trim said:

Normally my GPS setting is British Grid/Ord Survey GB. This seems to
place me about 100m. out on my Memory Maps. Their FAQ suggests
setting my GPS to WGS84. Can't get my head round this. What's the
reason?

There are various methods of doing WGS84 to OSGB conversions, some are
more accurate than others. Typically the one done by a GPS receiver
is not that accurate due to the limited processing power available.

Around here if you let the GPS do the conversion you get about a 45m
error WSW of the true position.

Chris
--
Chris Eilbeck
MARS Flight Crew http://www.mars.org.uk/
UKRA #1108 Level 2 UYB
Tripoli UK Member #9527 LSMR

Leonard Trim said:

Normally my GPS setting is British Grid/Ord Survey GB. This seems to
place me about 100m. out on my Memory Maps. Their FAQ suggests setting
my GPS to WGS84. Can't get my head round this. What's the reason?

Use WGS84 to *communicate* with the GPS (send data between the GPS and the
software).

Use Ord Survey to *display* the data (on the software and on the GPS).

Paul

The message <[email hidden]>
from "Paul Saunders" <[email hidden]> contains these words:

Quoted message said:

Leonard Trim said:

Normally my GPS setting is British Grid/Ord Survey GB. This seems to
place me about 100m. out on my Memory Maps. Their FAQ suggests setting
my GPS to WGS84. Can't get my head round this. What's the reason?

Quoted message said:

Use WGS84 to *communicate* with the GPS (send data between the GPS and the
software).

This certainly puts me in my back garden as I should be!

Quoted message said:

Use Ord Survey to *display* the data (on the software and on the GPS).

Does this mean I should change back to Ord Survey when I am following a
series of downloaded waypoints?

Where do you learn all this apart from doing it the hard way?

One annoying problem I am getting is an occasional clash with the
software causing the cursor to leap all over the place when the GPS is
connected. A few reboots solves the problem, but annoying.

Thanks for the help Guys.

--
Len Trim
[email hidden]

Leonard Trim said:

Quoted message said:

Use WGS84 to *communicate* with the GPS (send data between the GPS
and the software).

This certainly puts me in my back garden as I should be!

Jolly Good.

Quoted message said:

Quoted message said:

Use Ord Survey to *display* the data (on the software and on the
GPS).

Does this mean I should change back to Ord Survey when I am following
a series of downloaded waypoints?

Yes. I don't change anything on my GPS, I set the communication to WGS84 in
the software. The GPS sends in WGS84 no matter what, so you can leave the
GPS set to Ord Survey - that setting is only for display.

Quoted message said:

Where do you learn all this apart from doing it the hard way?

I read it somewhere, can't remember where, it was a long time ago.

Quoted message said:

One annoying problem I am getting is an occasional clash with the
software causing the cursor to leap all over the place when the GPS is
connected. A few reboots solves the problem, but annoying.

Don't know about that.

Paul

Chris Eilbeck said:

There are various methods of doing WGS84 to OSGB conversions, some are
more accurate than others. Typically the one done by a GPS receiver
is not that accurate due to the limited processing power available.

True, but it's not far off. Compared to accurate published OS trig point
GRs, the error is rarely more than 3-5m. Worst I've seen in Wales is 8m.

Not really an issue though since the maps themselves aren't that accurate.
At 50k, 1 pixel is equal to 5m, and even if you do set your waypoints with
pixel precision, a spot height mark is often 2 pixels wide, and they're
often not in exactly the right position. Things are much better at 25k,
with 1 pixel equalling 2.5m, and the spot heights and other features placed
much more accurately, but again, the slightest error in placing the waypoint
can put you 5-10m out. I'm quite happy with 10m accuracy though.

Quoted message said:

Around here if you let the GPS do the conversion you get about a 45m
error WSW of the true position.

45m? Are you sure? Where are you?

Paul

In message <[email hidden]>, Leonard Trim
<[email hidden]> writes

Quoted message said:

Normally my GPS setting is British Grid/Ord Survey GB. This seems to
place me about 100m. out on my Memory Maps. Their FAQ suggests setting
my GPS to WGS84. Can't get my head round this. What's the reason?

For more information see "A guide to coordinate systems in Great
Britain" at:

http://www.gps.gov.uk/guidecontents.asp

Regards

--
Neil Pugh

Paul Saunders said:

Chris Eilbeck said:

There are various methods of doing WGS84 to OSGB conversions, some
are more accurate than others. Typically the one done by a GPS
receiver is not that accurate due to the limited processing power
available.

True, but it's not far off. Compared to accurate published OS trig
point GRs, the error is rarely more than 3-5m. Worst I've seen in
Wales is 8m.

Not really an issue though since the maps themselves aren't that
accurate. At 50k, 1 pixel is equal to 5m, and even if you do set
your waypoints with pixel precision, a spot height mark is often 2
pixels wide, and they're often not in exactly the right position.
Things are much better at 25k, with 1 pixel equalling 2.5m, and the
spot heights and other features placed much more accurately, but
again, the slightest error in placing the waypoint can put you 5-10m
out. I'm quite happy with 10m accuracy though.

Road centrelines are the most accurate mapping features but you only
really have access to this on 10K or better mapping.

Quoted message said:

Quoted message said:

Around here if you let the GPS do the conversion you get about a 45m
error WSW of the true position.

45m? Are you sure?

Yes, certain. I have access to survey grade GPS and a cheap one
(Magellan 3000 or Garmin GPSIII+, both with excellent constellation
visibility) claims my survey point is on top of another building, not
on mine.

Quoted message said:

Where are you?

The Malverns

Chris
--
Chris Eilbeck
MARS Flight Crew http://www.mars.org.uk/
UKRA #1108 Level 2 UYB
Tripoli UK Member #9527 LSMR

"Leonard Trim" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Normally my GPS setting is British Grid/Ord Survey GB. This seems to
place me about 100m. out on my Memory Maps. Their FAQ suggests setting
my GPS to WGS84. Can't get my head round this. What's the reason?

The base map is WGS84 and the grid (overlay) is OS, as with my
Geko101.

Nick

Paul Saunders <[email hidden]> wrote

Quoted message said:

Something about GPSRs..

All's right with the world again.
--
Gordon Harris

Chris Eilbeck said:

Road centrelines are the most accurate mapping features but you only
really have access to this on 10K or better mapping.

Really? I thought trig points were the most accurate.

Quoted message said:

Quoted message said:

45m? Are you sure?

Yes, certain. I have access to survey grade GPS and a cheap one
(Magellan 3000 or Garmin GPSIII+, both with excellent constellation
visibility) claims my survey point is on top of another building, not
on mine.

Quoted message said:

Where are you?

The Malverns

I'm very surprised. Never encountered an error anywhere near that big
myself (assuming good reception of course).

Paul

Paul Saunders said:

Chris Eilbeck said:

Road centrelines are the most accurate mapping features but you only
really have access to this on 10K or better mapping.

Really? I thought trig points were the most accurate.

As far as I know GPS is the reference now and hardly anyone uses and
maintains the calibration on trig points. Keep an eye out for nails
hammered through washers with a ring of paint around them on the edge
of a pavement. They are accurately surveyed locations, usually to
better than 10cm accuracy.

Quoted message said:

Quoted message said:

Quoted message said:

Where are you?

The Malverns

I'm very surprised. Never encountered an error anywhere near that
big myself (assuming good reception of course).

I've done this with several cheap receviers and found similar errors,
all with excellent sky visibility and no obscuration.

Chris
--
Chris Eilbeck
MARS Flight Crew http://www.mars.org.uk/
UKRA #1108 Level 2 UYB
Tripoli UK Member #9527 LSMR

On Mon, 07 Nov 2005 08:32:01 +0000, Chris Eilbeck
<[email hidden]> wrote:

| Keep an eye out for nails
| hammered through washers with a ring of paint around them on the edge
| of a pavement. They are accurately surveyed locations, usually to
| better than 10cm accuracy.

Not by the Ordnance Survey. We have some on the pavement near us, they
were surveyed by some private surveyors laying out a group of new houses.

The only ones which the OS now maintain ridiculously accurate positions for
are the Passive Station Database on http://www.gps.gov.uk/
--
Dave Fawthrop <dave hyphenologist co uk>
Help make Usenet a better place: English is read downwards,
please don't top post. Trim replies to quote only relevant text.
Check groups.google.com before asking an obvious question.

In article <[email hidden]>, Chris Eilbeck
<[email hidden]> writes

Quoted message said:

Leonard Trim said:

Normally my GPS setting is British Grid/Ord Survey GB. This seems to
place me about 100m. out on my Memory Maps. Their FAQ suggests
setting my GPS to WGS84. Can't get my head round this. What's the
reason?

There are various methods of doing WGS84 to OSGB conversions, some are
more accurate than others. Typically the one done by a GPS receiver
is not that accurate due to the limited processing power available.

It is actually more due to the availability of accurate transformation
parameters. There are widely available parameters for the Molodensky
transformation for most of the datums a user might come across. That is
why that it the usual method implemented in receivers.

Quoted message said:

Around here if you let the GPS do the conversion you get about a 45m
error WSW of the true position.

I have never seen anywhere near that size of error in the conversion.
Can you give some example co-ordinates for us to try and an example
receiver that displays that size of error?

--

Dominic Sexton

The message <[email hidden]>
from Dominic Sexton <[email hidden]> contains these words:

Quoted message said:

Quoted message said:

Around here if you let the GPS do the conversion you get about a 45m
error WSW of the true position.

Quoted message said:

I have never seen anywhere near that size of error in the conversion.
Can you give some example co-ordinates for us to try and an example
receiver that displays that size of error?

Might be a complete red herring but it has crossed my mind that errors
of that magnitude could be a result either of the OSs well known
tendancy to build in inaccuracies to help protect copyright or their
more laudable practice of moving adjacent features so that they can be
shown larger than to strict scale and hence actually be seen without the
aid of a magnifying glass.

--
Roger Chapman so far this year 61 summits
New - 28 (Marilyns 14, Nuttalls 5, Outlying Fells 10)
Repeats - 34 (Marilyns 16, Nuttalls 24, Wainwrights 12, Outlying Fells 0)

Dominic Sexton said:

In article <[email hidden]>, Chris Eilbeck
<[email hidden]> writes

Quoted message said:

Leonard Trim said:

Normally my GPS setting is British Grid/Ord Survey GB. This seems to
place me about 100m. out on my Memory Maps. Their FAQ suggests
setting my GPS to WGS84. Can't get my head round this. What's the
reason?

There are various methods of doing WGS84 to OSGB conversions, some
are more accurate than others. Typically the one done by a GPS
receiver is not that accurate due to the limited processing power
available.

It is actually more due to the availability of accurate
transformation parameters. There are widely available parameters for
the Molodensky transformation for most of the datums a user might
come across. That is why that it the usual method implemented in
receivers.

Thanks for the info. It amounts to the same thing, limitations in the
receiver. When I was doing this stuff the OS had a piece of software
on their website which would do the simple transform and a link to a
more accurate transform. The poor one matched what the GPS receiver
itself was doing, the more accurate one matched the mapping I had
available.

Quoted message said:

Quoted message said:

Around here if you let the GPS do the conversion you get about a 45m
error WSW of the true position.

I have never seen anywhere near that size of error in the
conversion. Can you give some example co-ordinates for us to try and
an example receiver that displays that size of error?

I don't have access to the machine I had the data on, hard disk
crashed a while back. Next time I'm up on Worcestershire beacon I'll
take a few different receivers with me and record some data with them
sat on the trig point. I still have the Garmin and the Magellan but
I'd need to bodge the Rockwell Jupiter to run off a DC supply. I have
a few other receivers that could also be bodged up to get some data.

Chris
--
Chris Eilbeck
MARS Flight Crew http://www.mars.org.uk/
UKRA #1108 Level 2 UYB
Tripoli UK Member #9527 LSMR

Roger said:

The message <[email hidden]>
from Dominic Sexton <[email hidden]> contains these words:

Quoted message said:

Quoted message said:

Around here if you let the GPS do the conversion you get about a
45m error WSW of the true position.

Quoted message said:

I have never seen anywhere near that size of error in the
conversion. Can you give some example co-ordinates for us to try
and an example receiver that displays that size of error?

Might be a complete red herring but it has crossed my mind that
errors of that magnitude could be a result either of the OSs well
known tendancy to build in inaccuracies to help protect copyright or
their more laudable practice of moving adjacent features so that
they can be shown larger than to strict scale and hence actually be
seen without the aid of a magnifying glass.

I've seen errors in mapping where you put two sheets together and
there's a forest on one side of an edge but it doesn't carry on over
the corresponding edge of the other map. I've also seen rivers and
streams have upto 100m discontinuities across the join of two map
sheets. Supposedly a lot of mapping is still based on what they call
base maps with roads, rivers and fence boundaries and these features
had not been updated for 90 years or so. This was several years ago
at the start of the digital mapping era so things may have changed
now.

Chris
--
Chris Eilbeck
MARS Flight Crew http://www.mars.org.uk/
UKRA #1108 Level 2 UYB
Tripoli UK Member #9527 LSMR

In article <[email hidden]>, Roger
<[email hidden]> writes

Quoted message said:

The message <[email hidden]>
from Dominic Sexton <[email hidden]> contains these words:

Quoted message said:

Quoted message said:

Around here if you let the GPS do the conversion you get about a 45m
error WSW of the true position.

Quoted message said:

I have never seen anywhere near that size of error in the conversion.
Can you give some example co-ordinates for us to try and an example
receiver that displays that size of error?

Might be a complete red herring

As I read it the error was in the conversion of co-ordinates from one
datum to another rather than in relation to a map.

Quoted message said:

but it has crossed my mind that errors
of that magnitude could be a result either of the OSs well known
tendancy to build in inaccuracies to help protect copyright

I'm not sure that many of them are deliberate errors. There are plenty
of accidental errors which when noticed and not corrected serve the same
purpose.

--

Dominic Sexton

In message <[email hidden]>, Chris Eilbeck
<[email hidden]> writes

Quoted message said:

I don't have access to the machine I had the data on, hard disk
crashed a while back. Next time I'm up on Worcestershire beacon I'll
take a few different receivers with me and record some data with them
sat on the trig point. I still have the Garmin and the Magellan but
I'd need to bodge the Rockwell Jupiter to run off a DC supply. I have
a few other receivers that could also be bodged up to get some data.

Chris

For information I sat my GPS on the trig point on a couple of occasions
a few years age:

The Ordnance Survey, from their website, reckon that the Worcester
Beacon trig point is at:

Grid ref: SO 76882.307 45218.178
Altitude: 425.252 metres

My GPS (Garmin 12) read:

First reading - Grid ref: SO 76884 45227 - Altitude 431 metres

Second reading - Grid ref: SO 76882 45227 - Altitude 429 metres

Regards
--
Neil Pugh

The message <[email hidden]>
from Dominic Sexton <[email hidden]> contains these words:

Quoted message said:

As I read it the error was in the conversion of co-ordinates from one
datum to another rather than in relation to a map.

In which case I am totally lost. I thought a GPS had to convert its
position under WGS 84 to the datum of choice and thus a 45m error in
conversion would automatically be a 45m displacement on the map.

--
Roger Chapman so far this year 61 summits
New - 28 (Marilyns 14, Nuttalls 5, Outlying Fells 10)
Repeats - 34 (Marilyns 16, Nuttalls 24, Wainwrights 12, Outlying Fells 0)