PDA

View Full Version : Digital Compass Software



bshwckr
4th February 2008, 09:24 AM
I cannot find a program that does this but I think it would be possible.
I have a PDA with built-in GPS. I also have a bluetooth GPS. I am thinking that if I placed them say 5 meters apart (my PDA bluetooth reaches that far), a software program (theoretical) could use the 2 co-ordinates and display a digital compass.
To build onto this, perhaps this same program could use both GPS units and average the values for greater accuracy? For this part, they could be sitting next to each other.
Where I would find this useful;
Geocaching - Once I get close to GZ (ground zero), generally the navigation pointer starts to swing around as it is usually necessary to stop moving. I could place the bluetooth GPS down near GZ and move around with my PDA. The direction pointer should be more stable.
Placing a new geocache - The co-ordinates for the new cache could be more accurate as it would be derived from the averaging of both GPS's.
Is there anybody in our forum with the skills to write this code?
Cheers
bshwckr

ozmidnight
4th February 2008, 09:54 AM
Hi Bshwekr

I don't know if this site mite help you or not but it could be interesting

http://www.arrickrobotics.com/arobot/compass.html

OZ

bshwckr
4th February 2008, 12:12 PM
Is this the sort of gear that a high end GPS with digital compass employs?

the evil twin
4th February 2008, 12:18 PM
Not to my knowledge... I would certainly expect better resolution.

A lot of "high end" gear simply has a seperate fluxgate compass cct built in so you kinda have a 2 in 1 device instaed of buying a GPS and a Digital Compass.

Has it's advantages and disadvantages.... biggest disadvantage is that if your GPS fails so does your Compass.

Nicko
4th February 2008, 01:26 PM
If you are thinking a software compass, it is only a pointer and not a true compass. OziExplorer CE I think now has that in it's new version.

4X4
4th February 2008, 02:46 PM
If you are thinking a software compass, it is only a pointer and not a true compass. OziExplorer CE I think now has that in it's new version.

It is only available in OziExplorer CE development version 2.02

bshwckr
4th February 2008, 03:19 PM
It is only available in OziExplorer CE development version 2.02
I am currently using OziCE 2.02 and yes it does have a compass. It is only useful when you are moving because if you stand still, it has no idea which way you are looking. By being on the move it can tell which way I am heading because it compares where I was a second ago with where I am now. My idea is;
If my PDA, with its own gps knows where it is by its co-ordinates and it talks to another gps nearby via bluetooth, it will know another set of co-ordinates. By using both sets of co-ordinates it should be able to point north while I am stationary (which it cannot do by itself). I keep almost saying that it can triangulate but that would just refer to 3 points. If I have 2 GPS co-ordinates, I have 4 reference points, 2 x latitude & 2 x Longitude.
Also, I do really like the idea of being able to use both GPS units to average a more precise location.
Unfortunately, I'm not smart enough to write the software. If Des can write it into OziCE, I would be very happy.

The Explorer
4th February 2008, 04:45 PM
Hello - all very interesting and good in theory but keep in mind that you could get into trouble looking for a geocache that is +/- 5 metres from its published location (in any direction) using a compass of any type when your gps unit tells you you are +/- 5 metres away from it (in any direction). I figure it will only create confusion, but I could be wrong. Main problem I see is that you have no idea how accurate the original cache location is or on other hand if you publish a cache with very accurate coords how accurate the searchers GPS reading will be on the day. Based on my experience if you are within 5 or 10 metres of a cache it wont be too hard to find (assuming coords are as correct as they can be).

Also based on my understanding you can only use two GPS units to improve accuracy if one is placed at a position that has a known coordinate - otherwise all you will be getting will be two lots of "inaccurate" data with no idea of the error involved with either. I get what you mean by averaging the result but you can do that with one GPS - just wait a couple of extra minutes. Just my quick thoughts.

Cheers
Greg

bshwckr
4th February 2008, 05:14 PM
Hello - all very interesting and good in theory but keep in mind that you could get into trouble looking for a geocache that is +/- 5 metres from its published location (in any direction) using a compass of any type when your gps unit tells you you are +/- 5 metres away from it (in any direction). I figure it will only create confusion, but I could be wrong. Main problem I see is that you have no idea how accurate the original cache location is or on other hand if you publish a cache with very accurate coords how accurate the searchers GPS reading will be on the day. Based on my experience if you are within 5 or 10 metres of a cache it wont be too hard to find (assuming coords are as correct as they can be).

Also based on my understanding you can only use two GPS units to improve accuracy if one is placed at a position that has a known coordinate - otherwise all you will be getting will be two lots of "inaccurate" data with no idea of the error involved with either. I get what you mean by averaging the result but you can do that with one GPS - just wait a couple of extra minutes. Just my quick thoughts.

Cheers
Greg

G'day Greg
Yes you are right on all points. Still, I would like that needle to stop spinning around when I get close to the cache. I suppose I could buy a Garmin 60CSx but the wife has pulled the pin on my gadget budget (she thinks I can suffer with just 2 GPS's having left the third in a cache for a gift).
With the averaging, I have had both gps's averaging at the same location for several thousand samples but still had a slightly different reading out of each. My thinking is that if both were talking to the same piece of software, my two "wrongs" would be nearer the "right".
As for the accuracy of the gps of someone trying to find my super accurately placed cache, at least I would know my co-ordinates were as close as could be reasonably expected.
Regards
Mark

Nicko
4th February 2008, 08:50 PM
Buy a differential beacon receiver ($3k) and if within 200km of a free maritime beacon it will give you anywhere between 1 and 3 metres, as explained
http://www.gpsaustralia.net/forums/showthread.php?t=2473

Otherwise buy a Real Time Kinematic (RTK) system with base, radio and rover to give you 1cm accuracy in the horizontal and 2cm in the vertical but only if you have $45k spare.

Another method is GPSGate, where a differential system outside of the free maritime beacon. You anchor a base station to a known survey marker (SSM) and walk away with a receiver; it will give you sub metre for a mere $15,000.

So, what can I sell you ;)

bshwckr
5th February 2008, 02:04 AM
Buy a differential beacon receiver ($3k) and if within 200km of a free maritime beacon it will give you anywhere between 1 and 3 metres, as explained
http://www.gpsaustralia.net/forums/showthread.php?t=2473

Otherwise buy a Real Time Kinematic (RTK) system with base, radio and rover to give you 1cm accuracy in the horizontal and 2cm in the vertical but only if you have $45k spare.

Another method is GPSGate, where a differential system outside of the free maritime beacon. You anchor a base station to a known survey marker (SSM) and walk away with a receiver; it will give you sub metre for a mere $15,000.

So, what can I sell you ;)
Can you give me a bundled price for all 3? I would also need an easy payment plan. :grinner

Art
16th February 2008, 10:55 AM
These are good ideas, and sounds like a good exercise for the
programmer, and functional in the way you describe only if both units have a fix of great accuracy.
I might ask the obvious question.. why not use one of these:

http://images.google.com.au/url?q=http://www.stanleylondon.com/compass.jpg&usg=AFQjCNHcFmlRgUj-D3mR6qkLVCfX9lA98A

StormyKnight
16th February 2008, 03:12 PM
Using simple mathematics it wouldn't be too hard to implement. Really all you are doing is finding the midpoint between two time accurate positional measurements. Actually you could triangulate if you had three gps's to.

The big question is would it be more accurate using this combine information to make a real-world difference considering that they both are receiving identical information & the only difference between would be the quality of the reception equipment & the accuracy of the algorithms used to calculate your position.

I suspect the accuracy would be similar to having 100 gps's & averaging their indicated positions.

I found this link, it may be of interest.
http://www.syz.com/gps/gpsaveraging.html

here is a quote from it...

The graph can be used to determine "how long is long enough" when averaging GPS positions in the following manner. For example, assume that you want to be accurate to within 10 metres in an east-west direction 68.27% of the time (1 standard deviation), using all available data (not necessarily waiting for a more optimal D.O.P. in satellite configuration). From the pink line, we can readily read that approximately 15 minutes of data are required (gathered once per second, or generally as quickly as possible).

& for a more detailed paper on the effects of averaging a track on a moving object....a military document of 40 pages.

http://stinet.dtic.mil/cgi-bin/GetTRDoc?AD=ADA369222&Location=U2&doc=GetTRDoc.pdf

Food for thought?

mlenser
16th February 2008, 10:49 PM
If my PDA, with its own gps knows where it is by its co-ordinates and it talks to another gps nearby via bluetooth, it will know another set of co-ordinates. By using both sets of co-ordinates it should be able to point north while I am stationary (which it cannot do by itself). I keep almost saying that it can triangulate but that would just refer to 3 points. If I have 2 GPS co-ordinates, I have 4 reference points, 2 x latitude & 2 x Longitude.
I don't understand the principle that more than one gps input would allow you to determine which direction your (gps) is facing? Can some explain it in simple terms please.

Regards,
Mike.

P.S
I tried an averaging way point the other day for approx 2 min's at 1 sec recordings and the estimated error actually increased from 4m at start to 4.5m. I was only playing and was marking the cars location before heading of on a hike so high accuracy wasn't an issue and I didn't want to wait around longer to see what happened. I'll have to try it again when I'm bored.

Nicko
17th February 2008, 12:59 AM
Mike it is like shooting at a bullseye, even though you may have 4m accuracy shifting to 4.5 the fact is the majority will be very close to the true position and may over time average out all results to say 3 metres from actual position. Not a great improvement but an improvement all the same.

The two GPS theory is in actual fact being used on ships. I have seen 4 GPS on a car.

In theory if you position a GPS over a known point and a second is attached to the first but a fixed distance then as soon as you move the two can update the other in a differential manner.

mlenser
17th February 2008, 07:38 AM
The two GPS theory is in actual fact being used on ships. I have seen 4 GPS on a car.

In theory if you position a GPS over a known point and a second is attached to the first but a fixed distance then as soon as you move the two can update the other in a differential manner.I understand the idea of differential gps corrections when you use a fixed land gps (with a highly accurately known position) sending out the correction figures.
I assume that the shipping system doesn't get anywhere near 1cm accuracy and that they must use radio signals to communicate between the different GPS's/ships but how would ships know the distances between each other accurately? Or do they relay correction information from landbased stations from ship to ship as well? I might have to Google it later.

Regards,
Mike.

The Explorer
17th February 2008, 08:27 AM
I don't understand the principle that more than one gps input would allow you to determine which direction your (gps) is facing? Can some explain it in simple terms please.

Regards,
Mike.



Yes, you are right, using the scenario from the first post in this thread (ie a gps with a compass display that requires you to be moving to determine the direction it is facing) it would be impossible for the compass to work as a "normal" compass if you were stationary, no matter what addition gps information was provided to it. The gps unit in question would have no idea at what orientation it was being held.

You can of course get gps units that have a digital compass (ie you dont have to move for it to work) so if you could improve the accuracy of your coordinates then theoretically it would improve the accuracy of the compass. All the technolgy to do this already exists (differential or Real time systems) but as I mentioned previously it requires one unit to be at a known point so that the error can be calculated.

My understanding is that it is of no use getting data from gps units in "unknown" locations and then trying to use this inaccurated data to calculate an accurate cordinate for a GPS somewhere else. Even if you could do it I see little point in using it to find a geocache that has (very likely) inaccurate (+/- 5m) coordinates anyway. You can use the most accurate gps system in the world for geocaching but if everyone else is using normal handhelds (to hide and seek) it wont help your cause.

Cheers
Greg

Dooghan
17th February 2008, 10:03 AM
but how would ships know the distances between each other accurately? Or do they relay correction information from landbased stations from ship to ship as well? I might have to Google it later.

Regards,
Mike.I would think radar would take care of that.

Not_lost_just_resting
17th February 2008, 01:28 PM
I would think radar would take care of that.

I would think Automatic Identification System (AIS) takes care of it.

N

Nicko
17th February 2008, 04:28 PM
I don't think anyone said ships do get 1cm, in fact they use the maritime differential beacon service which is around 1 -2 metres close to port of say a few kilometres. That is sufficient accuracy for them to go through narrow channels.

If are happy with a local grid then you can place a nail in the ground anywhere on the planet and use that as the reference. Even an eTex has that facility.

mlenser
17th February 2008, 06:39 PM
I don't think anyone said ships do get 1cm, Sorry I didn't mean to imply that anyone had said that it could achieve 1cm accuracy. It was just that I had read/heard that differential gps systems (surveyor quality) can get down to a few mm accuracy and I was just thinking (typing) aloud that I was guessing/assuming that the shipping system couldn't attain the same accuracy. As you stated 1m - 2m accuracy seems much more likely the best achievable without ground based correction data.

Regards,
Mike.

Dooghan
17th February 2008, 06:44 PM
I would think Automatic Identification System (AIS) takes care of it.

NThat's how much I know about ships :lol2

Nicko
17th February 2008, 08:01 PM
As you stated 1m - 2m accuracy seems much more likely the best achievable without ground based correction data.

Maritime Differential Beacon is a Ground based service, but permanent, (except the receiver on the ship of course0 unlike the typical survey grade RTK (Real Time Kinematic) system with a base on a known point. You may see a tripod with a white dome shape antenna on the side of the road, that's the base. It sits directly above a peg (wooden stake) or a square thin piece of aluminum with a nail through it and a number placed there by a surveyor.

However, I have read/heard somewhere that a system is being used with high accuracy on a ship without a base station. How it does it I can't remember.

Not_lost_just_resting
17th February 2008, 10:25 PM
That's how much I know about ships :lol2

Have a read of this for a bit more information.

http://www.navcen.uscg.gov/enav/ais/how_ais_works.htm

If you want to watch the ships in your nearby port have a look at Ship Plotter here http://www.coaa.co.uk/shipplotter.htm

and the AMSA fact sheet here http://www.amsa.gov.au/Publications/Shipping/AIS_fact.pdf

Cheers,

N

BillForShort
24th June 2008, 07:58 AM
Hi, AgMapper software has the electronic compass as well as ALOT of other features: www.agmapper.com (http://www.agmapper.com) .

Let me know what you think.

bshwckr
24th June 2008, 08:54 AM
Hi, AgMapper software has the electronic compass as well as ALOT of other features: www.agmapper.com (http://www.agmapper.com) .

Let me know what you think.
Its not really an electronic compass. This uses the gps co-ordinates while you are on the move to simulate a compass. If you are standing still, it will have no idea in which direction you are looking.
IMHO an electronic compass is only an electronic compass when it has the appropriate hardware as in a high-end stand alone GPS unit. I am yet to find a PPC that has this hardware.
bshwckr

BillForShort
24th June 2008, 09:05 AM
bshwckr, yeah right... not some thing I had thought about, sorry.

It would have to be a pretty speciallised bit of hardware for the manufacturer to warrant putting a magnet insde one huh.

All the best with your search...

adjohnno
27th April 2009, 09:17 PM
This was very interesting to read, your thread, I will have to try this.

Buco_73
31st October 2010, 09:24 PM
use one of these devices..

http://www.active-robots.com/products/sensors/sparkfun/magneto.shtml

some custom programing and you should be able to achieve what you are after....