Accuracy

[+allan01273]

I purchased a Gps unit second hand I am unsure if the seller got rid of it because it was inaccurate.

How accurate should they be?

 

Standing at the Meridian stone near Newick which should be 000 it read W00 00.087

 

Approaching the GZ the readout changes by 1 every two or three paces. When I reach the GZ and lay it on the ground, knowing that the cache is arround here somewhere, the readout keeps changing by about plus or minus 00.020. Should I take the average or do other gps hold a steady position.

How far is the difference of each 00.001

 

Not such a problem but still annoying, the height above sea level says something like 33 feet plus or minus 67 which is at the bottom of the ocean

 

Allan

[+Lost in Space]

According to a well known aged poster here, the average accuracy is + or - 35'.

 

Where the heck are you that 100' is the bottom of the ocean?

[+jerryo]

Acording to my calculations at that latitude your GPSr seems to be out by 200 feet, which is a lot. As LIS says you should expect an accuracy of +/- 30 feet or so in the worst conditions (from my experience). If your machine is using barometric info to get the elevation, it may need calibrating. If you're using GPS elevation it can be a bit inaccurate. Are you using WGS 84 on your GPS? Which one have you got?

[+Graculus]

All you ever wanted to know about GPS and accuracy (but were afraid to ask...except you did ask )

 

There is a very good resource site for geocachers my alter ego Mr Blorenge put together and part of it is about GPS receivers, how they work, waypoints and more. It isn't too technical and hopefully will explain what you need to know.

 

GPS receivers won't develop an inaccuracy as a result of age. But they can be very innacurate if they are not setup correctly. So go and have a look and if you have any further questions just ask.

 

Chris

Graculus

Volunteer UK Reviewer

 

www.follow-the-arrow.co.uk/resources

(scroll down the page and follow the link to GPS)

[+JeremyR]

It probably is set to the wrong datum but just a thought - I've never heard of it but just how long has the Meridian Stone been there? If it's been there a while, the thought does occur that the Meridian Stone might not be exactly on the Greenwich meridian, which may account for some or all of your error?

[+The Bongtwashes]

It probably is set to the wrong datum but just a thought - I've never heard of it but just how long has the Meridian Stone been there? If it's been there a while, the thought does occur that the Meridian Stone might not be exactly on the Greenwich meridian, which may account for some or all of your error?

It depends what datum you are using, and which Greenwich Meridian you are using, it has changed a few times

[+Team Sieni]

It depends what datum you are using, and which Greenwich Meridian you are using, it has changed a few times

Indeed. It's likely (but not certain) that the meridian stone you were on was placed on the Airy meridian, which is about (depending on your latitude) 100m to the W of the WGS84 Meridian, where your GPSr will show 0.00.

 

Confused?

 

Here are a couple of links that may help.

 

Flamsteed Astronomy Society

History of the Prime Meridian

 

If you are interested in Waymarking, you could record the stone here: Greenwich Meridian Markers

[+allan01273]

It probably is set to the wrong datum but just a thought - I've never heard of it but just how long has the Meridian Stone been there? If it's been there a while, the thought does occur that the Meridian Stone might not be exactly on the Greenwich meridian, which may account for some or all of your error?

It depends what datum you are using, and which Greenwich Meridian you are using, it has changed a few times

Havent got time for a full reply at the moment. Thanks for all your responses and I will look at them further, later tonight.

The stone was part of "Not another Meridian Cache" GC19C3K which I did yesterday with "Upper River Ouse Locks - Iron Gate" GC15QGK in the afternoon.

 

Allan

[+JeremyR]

The stone was part of "Not another Meridian Cache" GC19C3K which I did yesterday with "Upper River Ouse Locks - Iron Gate" GC15QGK in the afternoon.

The cache listing shows that the WGS84 coords for the stone are at W 000° 00.096. Your recorded coordinates are about 34 feet away from that point, well inside standard GPS accuracy especially considering the location appears to be under tree cover - your unit is fine and is set up properly...

Edited October 26, 2008 by JeremyR

[+Team Sieni]

The stone was part of "Not another Meridian Cache" GC19C3K which I did yesterday with "Upper River Ouse Locks - Iron Gate" GC15QGK in the afternoon.

The stone is waymarked here: WM4MHX This waymark shows the co ordinates in both OSGB and WGS84, so you can see the difference.

[+allan01273]

I am using a Garmin GPS45. It was sold as a marine unit and is now long since obsolete. The software is from 1994.

It only cost £11.00 so I realised that it may not be suitable for Geocaching, but risked it anyway.

I thought that "marine" would mean it's more robust and waterproof. After all a position is just a position whether on land or sea.

Why would it give the height above sea level, if it would not work on land? I have set the Units from nautical to statute (I'm too old to work in Metric) and the datum is WGS84 as shown in its instruction book. There were about 100 datums to choose from including Ord Survey GB which does not sound too nautical.

 

If the missus and I find that Geocaching is something that we can get interested in, then if necessary, we will spend a little more money and update our GPSr, but it seems that there is no greater accuracy in spending hundreds of pounds.

It would seem that the Meridian stone was not where I thought it should be, on the Meridian!. What about my other queries how far is 00.001 in inches, and is the readout meant to recalculate and change by plus/minus 00.020 whilst the GPSr is stationary.

 

Allan

[+careygang]

Famous Double Posting glitch!

Edited October 27, 2008 by careygang

[+careygang]

Double posting glitch!

Edited October 27, 2008 by careygang

[+careygang]

I am using a Garmin GPS45. It was sold as a marine unit and is now long since obsolete. The software is from 1994.

It only cost £11.00 so I realised that it may not be suitable for Geocaching, but risked it anyway.

I thought that "marine" would mean it's more robust and waterproof. After all a position is just a position whether on land or sea.

Why would it give the height above sea level, if it would not work on land? I have set the Units from nautical to statute (I'm too old to work in Metric) and the datum is WGS84 as shown in its instruction book. There were about 100 datums to choose from including Ord Survey GB which does not sound too nautical.

 

If the missus and I find that Geocaching is something that we can get interested in, then if necessary, we will spend a little more money and update our GPSr, but it seems that there is no greater accuracy in spending hundreds of pounds.

It would seem that the Meridian stone was not where I thought it should be, on the Meridian!. What about my other queries how far is 00.001 in inches, and is the readout meant to recalculate and change by plus/minus 00.020 whilst the GPSr is stationary.

 

Allan

 

Given the age of your unit, it is going to be less reliable than a more modern item.

 

The fact is that any GPSr relies on receiving the signal from multiple satellites and using the difference in the signals timing to calculate the location. The older the unit the more susceptible to interference. As yours was marketed as a marine unit (I'm only guessing here) it probably was not intended to be able to cope with signal delay caused by bouncing from tall buildings or any obstruction caused by high ground.

 

Even my new eTrex Vista HCx (HCx = better chipset) throws a wobbly when surrounded by high buildings, with the pointer sometimes giving up to a full 180 reverse when it gets a good signal. Not what you want when within 50ft of the cache.

 

You don't have to spend 100's to get a better accuracy. A basic eTrex yellow H model (high sensitivity chipset) can be brought from a well known on-line seller that sounds like a South American River, for under 60 quid, while a Vista HCx will cost about 160. As the price goes up you tend to be paying for features rather than accuracy these days.

 

It's just a thought, but as your software is from 1994 (before the switch off of selective availability) they didn't expect an accuracy of much better than a couple of hundred feet, so you're probably getting the best that they intended at the time.

[+Just Roger]

What about my other queries how far is 00.001 in inches, and is the readout meant to recalculate and change by plus/minus 00.020 whilst the GPSr is stationary.

 

Allan

 

To all intents and purposes a nautical mile is 2000yards (2020 if you want to be fussy). A nautical mile is defined as a Minute of Latitude. Thus 1 minute of Latitude is 2000 yds and 0.001 of a minute will be 2yds or 6 feet. Or, as you wanted it in inches, 72 inches

 

Note all the above applies to Latitude. A minute of Longitude varies with latitude and is (near enough) the same at latitude at the equator but is Zero, nil, very small at the poles. It is actually 1 x Cos(Lat) miles so you can work it out for your own location.

 

The reading will always vary slightly while stationary, the size of the variation will depend on circumstances -tree cover, urban canyons etc. but having said that a variation of 0.02 is on the large side. Try taking your GPS into the middle of a big open space and check the variation there. It is also worth checking the arrangement of the satellites at the time. Occasionally I have seen all the visible satellites an a line across the sky. This gives wonderful "accuracy" along the line but terrible at right angles to it.

 

Hope this helps.

Edited October 27, 2008 by Just Roger

[+Graculus]

It does take a GPS a certain time to lock on to the signals from satellites. If you switch it on and after a few moments it has a 3D lock (3 satellites minimum) the accuracy will be quite poor. As it locks on to more satellites this will improve. I've often found if there are no satellites directly overhead and with trees or an urban canyon blocking those lower in the sky the GPS will struggle to get a decent lock on multiple satellites (accuracy may never improve beyond 40ft or more). The newer chipsets (SiRFstarIII) have more channels which allows them to process signals from multiple satellites and so get a faster lock.

If you look at this display from a Colorado, the outer circle of the satellite display is the horizon, the inner circle is 45 degrees above the horizon (imagine it like a circle drawn round the sky half way between the horizon and directly overhead. The centre of the circle is directly overhead. Any satellites below the 45 degree line may be difficult to receive if your horizon is obscured.

Here satellite 30 is overhead, but 5, 12, 2, 29 show a higher signal strength. This comes down to the way the GPS is held. The Colorado has a helix aerial which means it should be held vertically, the Etrex range have a patch aerial which means they should be held flat in your hand. Also satellite 24 is just being recieved but 20 isn't, this is because my house was to the North of the unit so it obscured 20. If you find there are no satellites directly overhead, don't bother waiting for them to move. The satellites orbit at a height of 12,000Km so take 12 hours to make one orbit (6 hours to travel from one horizon to the other).

So depending on how many satellites are visible, whether you are holding the unit correctly, atmospheric disturbance, ionospheric disturbance, trees and buildings it is a wonder we get a signal at all .

[+Team Sieni]

You probably already know this, but for caching you should always have the datum set to WGS84 (and the co-ords format to DD MM.SSS) Accidentally leaving it set to another value can have "interesting" results.

 

Sounds like you may have a valuable antique.

 

So depending on how many satellites are visible, whether you are holding the unit correctly, atmospheric disturbance, ionospheric disturbance, trees and buildings it is a wonder we get a signal at all .

You don't believe all that mumbo jumbo about great birds in the sky do you? It's all done by pixies really. The pixies are scared of the dwarves who live in the woods, and of busy towns, and that's why reception is bad in woodland and city centres.

[+deejay44]

It does take a GPS a certain time ...

 

Nice one Graculus I enjoyed reading that and found it very informative.

 

Dave

[NickPick]

Just another thing to consider. While GPS signals should be accurate to within 10m or so when you get a good signal, the vertical accuracy is usually nowhere near this.

 

I think the sattelite constellation is optimised for placing you on a globe, where best horizontal accuracy is achieved when you have sattelites nearer the horizons. The vertical accuracy is less accurate in this configuration, so I'd expect vertical accuracy to be about 3 times less than horizontal (So if you're getting an 'accuracy' of 10m, then you can expect the altitude to be within 30m)

 

I don't think your old unit should be any less accurate than a more modern one, but it might not be able to lock onto the satellites as easily, due to the modern ones having improved antennae.

[+Graculus]

GPS elevation is indeed subject to the signal. If you happen to be by a trig point and know its altitude (fom the OS map) you can check how accurate the elevation reading of the GPS is. The Colorado has a barometric altimeter which is very accurate, that is assuming you remembered to set the pressure/or known elevation before leaving home

 

Chris

Graculus

[+allan01273]

Thank you all for all of your replies. I have now found out that the answers are all somewhere else on this site. However it is so much easier and quicker to ask specific questions and get the answers straight away, thank you.

 

My GPSr is old and only tracks eight satellites, it dosen't have the latest chipset, nor the toys, games and maps.

However I have already found a few caches and look forward to finding more. I don't know how people can find a lot of them in one day. The most that I have done is two, perhaps it is because the ones that I do involve searching all round villages for parts of the final location.

 

Perhaps the prices of GPSr on Ebay come down in the winter, when people are not using them and I will look out for them.

 

In the meantime I think that the GPS 45 that I have will sufffice. A major problem may be that I don't have "cachers eye" and don't guess or see the the hiding place. A little film box somewhere a stones throw away is a big task.

 

Thank you all,

 

Allan

[+careygang]

 

Perhaps the prices of GPSr on Ebay come down in the winter, when people are not using them and I will look out for them.

 

 

It's worth putting a few retailers in your favorites and then check them regularly for special offers.