GPS inaccuracy?

GPS/GNSS Development Boards
1

I’m using the Micro-Mini SMa with the MN5010HS module and reading GGA messages. The reported lat / long for is correct in terms of Deg and Minutes but wrong in seconds when compared to http://itouchmap.com/latlong.html

There is, on average, a 6" Lat and 10" Long discrepancy between reported data and what this site tells me is my exact position.

I have the module next to a window in a two story apartment by a lake. Is it possible there is enough interference to cause the inaccurate reading?

Thanks.

2

  1. I would not trust an online map for an accurate fix.

  2. What datum is your GPS using versus the online? Are they both WGS-84 (WGD/WGE)?

  3. Yes, you could get crappy fixes near a window. You will only get satelites in one quadrant of the sky. You need full sky visibility for a good fix.

I worked with a guy having trouble with his GPS. He sent me a listing of his signal to noise strengths, and I (to his surprise) was able to tell him what way his window faced.

3

From what I’ve gathered, both are WGS84.

I’m hoping the issue is in fact the window. I co-worker has given me an ANT-555 antenna that I can use outside of the window. I will try again today and report back here what I find.

4

If you want a GPS module that works in most locations get a u-blox unit. They are popular with the military and law-enforcement agencies because of their high performance.

5

The accuracy you’re getting is actually very good, considering you’re not using differential GPS…

6

6 seconds of latitude is 6/60 minutes of latitude. Each minute is about 6076 feet so 6 seconds of latitude is 607.6 feet. This is well above what GPS is capable of with a good GDOP. But operating from a window gives you a bad GDOP and hence a 600 foot error. Of course Columbus would have been delighted to have had even an error ten times larger!

To get the best HORIZONTAL solution, you want your four satelites all fairly low in the sky and spaced at 90 degrees around the horizon. To get a good vertical solution you need a satelite above you. If you have this ideal minimum setup plus a few extras, you will get a good solution. But if you cut out half the sky witha building behind you, all bets are off. To understand why, look up “line of position” and “position fix”. It all has to do with intersecting circles.

7

Ok, it’s time for an embarrassing admission of fault. To make a long story short, I’m a failure at reading comprehension.

After trying the ANT-555 and getting identical results, I decided to open up some more documentation to see if there was something I had overlooked. To my surprise, I realized that the NMEA format for a Lat coordinate was NOT “DDMMM.SSSS” I’m not sure why I thought this was the case.

NOW, I’m reading very accurate data. Actually, better than I expected. Thanks to all for the help and your patience with my foolish way. :roll: