HI,
please bear with me. Newbie just jump onboard to Arduino.
I'm looking for a good GPS module which can do 10Hz update for more precise application. I looked up Makershed, Sparkfun, Seeeduino..etc.
Any good recommendation?
Thanks
Roadblock
Hard to answer this question without a ton of caveats.
"Good GPS Module"
http://www.canalgeomatics.com/product_details.php?product_id=142
Just the GPS module:
http://www.canalgeomatics.com/product_details.php?product_id=45
If you're just looking for consumer level GPS, maybe look into:
http://www.gpscity.com/garmin-gps-18x-5hz-high-sensitivity.html
AFAIK, 10hz at "Decent" pricing is going to be difficult. (Beware of advertised lies.)
Please also note that a faster update rate is NOT going to give you more precision. It's still bound to the same error rates, just give you more of them. If its only giving you 5 Meters at 95% confidence level, it's just giving you that 5 Meter radius reading faster at 10Hz then it is at 1Hz.
The Hemisphere GPS modules can give you sub-meter positioning, but you must subscribe to an L-Band carrier, such as omniSTAR.
Quick primer on it:
http://www.hemispheregps.com/About/AboutGPSTechnology/tabid/374/Default.aspx
Somewhat cheaper precision comes from differential positioning, such as using land based beacons to correct GPS signals.
Example of such:
The Beacon receivers work excellent if you are somewhat near the transmitters. This generally means either coast = excellent. Central US/CDN = suckage.
Hmm ... "We lose a little (precision) on each one, but we make it up in volume!" ;D
The question I've never actually investigated is "Is the error systemic? Or is it noise that could be averaged out if you got more data?". If a given system is going to repeatably return an error of, e.g., 3m northeast of your actual position when you're in location "X", and always be off by 2m to the south at location "Y", you're just plain out of luck. But, if the error is "jitter", accumulating and averaging multiple readings could significantly improve accuracy.
Ran
Since we dropped Selective Availability, the error is primarily "jitter". In some cases, for example my case where I am deep in an urban canyon, the position data I get can be off by as much as a city block and sometimes looks pretty random. Averaging might indeed improve the quality of that data. On the other hand, if I'm sitting on the top of a mountain with nothing but clear sky overhead, averaging might not improve the accuracy.
The errors in GPS are due mostly to the ionosphere and multipath signals.
Sitting atop a mountain fixes both of those.
GPS is nothing more than high precision time signals that are triangulated.
The 'errors' happen because although we know precisely where each satellite is located, and we know how much time difference between each signal's time code should be, the signal itself gets delayed as it makes its way to your antenna.
The WAAS system helps add precision by broadcasting current information about the ionosphere so that it can update how long it will in reality take for the signal to get to your location.
Great leaps have been made to analyze this information inside the receivers themselves and try to model and predict the errors, but as soon as mankind figures mother nature out, she changes the game.
In any event, the error rate reported back to you is the receiver's standard deviation at 95%. Basically if it tells you 5M accuracy, it really means "I'm 95% certain I am calculating your horizontal position within 5M"
If you have seen your local survey guys out and about, you'll notice one guy has to stand in one place with a receiver for a while. That's how they are getting their accuracy back. (Not moving = 95% confidence level can correct itself over time.)
The Hemisphere modules I used to test if left for a day or 2 could bring the accuracy down to 2M with no outside corrections. As soon as you moved it, you immediately lost that, and were back to 5M.
Thanks for all the inputs. I'm sure that $1500 GPS wont' work for me ;D
I'm actually looking for a GPS which Arduino community has tested successfully. The only feature I'm looking for is finer details in GPS positions.
No idea on the $1500 one yet, still in the box on a shelf.
The $4000 one works like a charm. 