etrex venture HC - waas satellites , and claimed accuracy

[borienteer]

Father xmas duly delivered my etrex venture hc and I have been experimenting with it. I have two observations, I wonder if others would care to comment.

1) The accuracy figure quoted is pretty meaningless. For example I measured a waypoint averaged over 100 measurements, with a claimed accuracy of ~6m. Checking on the map it was 100m out!

Repeated the same procedure at the same location today having ensured I was receiving waas signals from satellite 33 and the position was dead on. Quoted accuracy on both occasions was similar.

My conclusion is, trust measurements when you have the little "D"s on the display, otherwise forget it. For my application (surveying orienteering maps) accuracy <~5m would usually be fine.

2) The unit continuously cycles through waas/egnos satellites 33/34/39/46/49, but only every gets information from 33. Shame I can't force it to just stick with the one that is actually useful round here. (N. Lancashire, UK)

[philpugh]

That's the problem with EGNOS (and WASS) so far away from the equator, the sats are low down to the south (ish). That and a relatively low signal strength leads to problems with reception.

Recheck without EGNOS. Make sure your GPS is placed in the open and getting 6+ sats locked on.

My Colorado 300 (and its predecesors GPS60 and GPSII+, both low sensitivity pre-SiRF days) all manage to get a 15 min average position to 4 digits of OSGB (that is a 10 digit position AA EEEE NNNN).

I did this by logging every few seconds, downloading the tracklog and processing with Excel on the computer. I use OSGB as it's very easy to get very accurate actual OSGB position (we have 1:2500 mapping of the farm for other purposes). This is done outside. I'm lucky in that I live on a farm in the Cheshire plain so clear skies/horizons are very easy to find! In a suburban area you may have to go hunting for a decent open space. Your figure of 100 meters seems to be very excessive and would indicate a blocking of sats from one half of the sky - at least. Even in poor conditions I would expect better average accuracy with a few log positions off by this sort of figure. Also the 100m figure is close to the sort of error you may get by using OSGB position information but having the WGS84 datum still set. (Or vice-versa). I could set this sort of wrong condition on my older units but the Colorado is more or less failsafe as setting BNG display automatically sets the datum to OSGB36.

Satellites (geometry and numbers seen) are the usual factor affecting absolute accuracy. the HDOP (Horizontal Dilution Of Precision) figure shown by many GPSrs, give a reasonable idea of the accuracy they are getting. It's a statistically based figure so it's not an absolute value of error (and it doesn't say anything about the direction of the possible positional error).
_________________
Phil

[borienteer]

Thanks for the answer. My unit will do an average position for you, it does not seem to make a lot of difference, even with 100+ readings. I am a little bit north of you on the Lancashire/Cumbria border. I will continue experimenting. Waiting 15 minutes for an accurate reading will make it of limited use for me. In that time in most circumstances I can do an accurate enough survey by old fashioned compas and pacing. I have started reading up about post processing, but that will require taking a laptop out to the woods with me to record the "raw" data.

[philpugh]

If you need very accurate positions then old fashioned survey techniques will give the most accurate answers - providing you have an accurate known fixed location to work from - and on many OS maps these are actually few and far between.

The point is that given good reception conditions you should be accurate to 15mtrs or better on most modern GPS systems. People use them for geo-caching and you have to get very close to find the correct spot. Surveying GPS systems have better aerials and use radio based differential GPS systems in many cases.

Taking laptops/PDAs out with you may not help as the result you will get is only as good as the data you put in - and if it is out by 100mtrs then this is going to be the error of the 'corrected' figure also. A lot of the post processing for position usually relies on an internet connection so that DGPS corrections can be obtained.

What map did you check the 100 point average on and what datum did you have set on the GPS? Don't forget the GPS will work on WGS84 out of the box (unless the shop had been playing) and the Lat Long you get on the edge of OSGB maps is OSGB36 datum - which gives rise the the sort of errors you are seeing. Go on to Google Earth, find the Prime Meridian at Greenwich and see what Google Earth shows as the Longtitude!
_________________
Phil

We see you’re using an ad-blocker. We’re fine with that and won’t stop you visiting the site.
Have you considered making a donation towards website running costs?.
Or you could disable your ad-blocker for this site. We think you’ll find our adverts are not overbearing!