PDA

View Full Version : Anyone using a Parallax GPS Module?



rixter
04-11-2007, 11:11 AM
Hi,

I am trying to get some feedback (perhaps I should have done this before my purchase), from others that may be using a Parallax GPS module. I've been in discussion with a couple of individuals off of this forum who have various brands of GPS units other than the Parallax unit. I've been learning quite a bit about GPS technology in my attempts to explain why this unit isn't returning latitude and longitude data that is "near enough" to my actual location. I was originally comparing results with what a factory automobile GPS unit is doing. Basically, they in effect take a bunch of data from various sensors to conclude where on the road you must be. This may not be exactly where the GPS got the coordinates from the satellites. The Parallax GPS unit claims a +/- 5 meter position accuracy on average, but I rarely if ever get coordinates that are within this range. 15 feet off wouldn't be anything to cry about, but I regularly get readings 100 feet or more off. This is the case even in the open with a clear sky. I don't know how I can improve the situation to get better results. The demo code that Parallax supplies will not display information if there is not a sufficient number of satellites acquired. I removed the section of code that checks for invalidity just so I could see if the device would tell me what data it WAS holding. Essentially, this is just the latest valid data it acquired plus time and date info, which is always spot on. This device will blink an LED if there is an issue getting a lock on satellites. I've found this device to be rather finicky as it tends to go into blinking mode for long bouts for some unexplained reason. The fact that I do get data and the coordinates are "on my street", if not where I am, tells me that at least I have everything attached properly.

I would appreciate comments or feedback from others as to the success or failure of getting decent data when using this module.

Thanks

Rick

p.s. I use "Smart Mode" and display the latitude and longitude figures on an LCD and map them in Google's GPS Visualizer

Chris Savage
04-11-2007, 11:22 AM
rixter,

Two things to consider…First of all if the LED is blinking the data in the GPS is not valid. That is not something we implemented, but how the GPS Module itself works. Our board simply buffers the NMEA messages and converts them to simpler data which a BASIC Stamp 2 can deal with more easily. You can always get the RAW data from the module but you will find when there are less than 3 satellites locked the LED will blink (this is a function of the Polstar GPS Module) and the flag within the string will indicate not valid. This would occur on any GPS unit that did not have a valid signal.

The second thing to consider is relating the coordinates on Google Maps with another mapping program. All systems have a margin of error that you are inherently adding to the GPS error. For example, plot coordinates on Google Maps, then go to another non-related site with similar functionality. Plot the same coordinates. You will see variation which is inherent. I have seen Google Maps show road lines off the road from the Satellite view. All I am saying is you need to be able to compare Apples to Apples. The only way to do that would be able to read the NMEA strings from both devices in the same location and compare them. We have a program that does this among the GPS Modules we sell. It also compares the Smart Mode values with those given by RAW mode. Take care.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support

rixter
04-11-2007, 07:20 PM
Chris,

Thanks for the information. I realize that the invalid flag is telling me that there aren't enough satellite acquisitions to get valid GPS data. Even my car unit has a GPS Data screen that will show "out of range" for several minutes when I pull out of the garage in the morning, and yet the mapping function in the car will correctly mark me as I drive for miles. Perhaps a credit to the technique they use to perform this.

I suspected that the mapping software I was using may just be off. I in fact live on one of those streets that Google doesn't line up with the true location as shown when you view it in satellite or hybrid view. I was in communication from the support guy there who informed me this does happen. In fact, I plotted some single GPS coordinates in Mapquest and got slightly different results. This morning I did just this for my home location, but it still plots me over 75 feet away in the neighbor's back yard.

I will certainly be studying up more on NMEA strings. Aside from comparing to my auto GPS unit, are you saying that those strings can return more accurate data as to my location if I use them properly or in some different way than what is being returned in Smart Mode? I would really be happy getting data that I feel confident is close to the error margin for the device.

When you say "among the GPS Modules we sell", is this a reference to testing all of the GPS modules of the same type, or do you sell more than one model that I am not aware of? Is there a chance that this one is just off by more than the position accuracy spec?

Thanks

Rick

Chris Savage
04-11-2007, 10:06 PM
Rick,

It seems you understand more now about what’s happening. Also bear in mind most Car GPS Systems will auto correct positioning on the road. If typically the current coordinate is off the road, the software keeps it on the road so you don’t look like you’re driving off the map. That’s part of the inherent error I was referring to when comparing to another program or website that doesn’t do this.

I’m not saying the RAW Mode will give you more accuracy…What I am saying is that the more satellites you have the better your accuracy. One our other Tech Guys (Dave) has managed to acquire 8 satellites from his desk (typically not possible indoors). The more you have the better the internal firmware of the module can triangulate your position.

We only sell the one module, but out test procedure doesn’t just verify coordinates from the module in both modes…They are all compared. That is, if when testing the 5th GPS Module it returns coordinates off by too much from previous units it fails. Again, the error is adjusted for the minimum number of satellites. It’s just a matter of how we test the units. Take care.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support

Fe2o3Fish
04-12-2007, 02:53 AM
Rick,
Please keep in mind something with regards to GPS:

Along the lines of what Chris said in his first reply, Lat/Lon doesn't not always equal Lat/Lon.
A given Lat/Lon determination is made for a given map datum. Map datums are mathmatical
models describing the earth's shape. NAD27, NAD83, and WGS85 are two of the more common
map datum, at least in the US. Don't "decry" GPS accuracy if you don't know which map datum
your GPS is using (usually WGS85) AND the map datum your map is drawn against.

If you'd like to learn about cartography and such, browse through the US Geological Survey website
sometime! http://www.usgs.gov

If you really want to check the GPS's accuracy, go find yourself a survey marker, sit your
GPS on top of it (or as close as possible), and compare readings keeping in mind what I've
written above AND that you have a sufficient # of satellites locked in. Three is the bare minimum
for a useful lat/lon but four or more is better. Check here to find a NGS survey marker near
you: http://www.ngs.noaa.gov/cgi-bin/ds_radius.prl

For what it's worth, the output from my Parallax GPS module from Parallax pretty well
matches the output of my Garmin 12XL.

-Rusty-

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
-Rusty-
--
Rusty Haddock = AE5AE = rusty@fe2o3.lonestar.org
**Out yonder in the Van Alstyne (TX) Metropolitan Area**
Microsoft is to software what McDonalds is to gourmet cooking

rixter
04-12-2007, 05:44 AM
Thanks for the input guys. I will check out that website, do some more testing and seek some other GPS mapping alternatives. I am not knocking GPS technology. I just want to know if the device I having is working properly. If I can get assurance that this device is at least returning data that closely matches any standard, I can go from there. My cause for concern though is that I get readings that are not consistent. In other words, if I am saying that it can be 100 feet off, that is just one of the readings.... it can be 75 feet off in the opposite direction on the next reading, then back to 60 feet off in yet a different direction the next reading. Odd.

My goal was to do data logging for cycling routes... the log containing other data in addition to coordinates. With the readings I was getting in my tests, I didn't want to let anyone see a map of coordinates I record and plot in GPS mapping software for fear of getting pulled over for eradic cycling.... cutting through people's yards and riding through buildings.

I'm sure any GPS device that does mapping is doing the course correction process, even though the car ones will let you drive off into a field somewhere if you want. As long as there is no known road there in their map data, they let you go where ever you want and not "attach" you to a road. But I can't build this complex correction process into my project.... at least not automatically.

Off to investigating those websites,

Thanks again.

Rick

Jonathan
04-12-2007, 05:57 AM
As well as the error in Google maps etc, is the fact that the USGS map they are taken frome aren't perfect either. Some areas have not been surveyed since the '40's or even earlier. I know this from much backpacking and taking GPS readings from places like Mount Shasta in CA. It just doesn't always line up with major landmarks, and the transition between quads can be off as well.

Also, while it was reduced a while back, doesn't the Govt. stiil introduce some error in the GPS readings? Both of my mobile GPS units can be set to an averaging mode, where the longer you let it sit in one place the more accurate the reading becomes.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.madlabs.info (http://www.madlabs.info) - Home of the Hydrogen Fuel Cell Robot

jhoyoza
04-12-2007, 10:54 AM
Hello,

I don’t know a lot about Parallax GPS units, however I know when building my GPS-Toy-Boat Project the reasons I selected a boat was due to the inaccuracies of GPS units in general. Having an unobstructed view of the sky is extremely helpful, and any signal deficiencies would still allow navigation without hitting obstructions.

http://forums.parallax.com/showthread.php?p=621606

I first used an older Garmin eTrex-Summit and it worked relatively well. I later upgraded to a Garmin eTrex-Legend with WAAS (Wide Area Augmentation System) and it got even better. I have seen accuracy down to about a meter. That’s pretty good. I’m very skeptical when seeing others land-based-vehicles (on the web) that appear to steer thru traffic cones and hit waypoints within inches. (Fat chance in H - E- double hockey sticks IMHO.) I built a land based vehicle 1/6 scale Monster-Truck, and I was hitting everything in the book, trees, telephone poles, parking lot barriers, lighting fixtures and parked cars. And as I approached the tree line, and/or buildings I would often lose signal totally.

(WAAS is basically a system of satellites and ground stations that provide GPS signal corrections, giving you even better position accuracy on average up to five times better. A WAAS-capable receiver can give you a position accuracy of better than three meters 95 percent of the time. And you don't have to purchase additional receiving equipment or pay service fees to utilize WAAS.)

The Garmin unit has many advantages as compared to simply providing longitude and latitude data with the addition of additional NEMA strings and up to 500 waypoints. As each waypoint is reached, it automatically switches to the next waypoint making it outstanding for navigation. The GPRMB and the GPRMC string provides advantages when using a BS2. GPRMB essentially provides general current heading data (In degrees) and GPRMC provides the desired heading (In degrees) to the current waypoint. The difference between the two provides steering data. I simply truncate the decimal and it still works great as it self-adjusts as you go.

Example of my method of removing decimal data:

SERIN4,16572,[WAIT("GPRMC,"),WAIT(","),WAIT(","),WAIT(","),WAIT(","),WAIT(","),WAIT(","),WAIT(","),string1,string2,string3]
IF(string1 < 48 OR string1 > 57)THEN GOTO GPS_Mode 'Bad data
IF(string2 = 46)THEN x=string1:x=x-48:ch=x:GOTO Continue_A '1 digit#
IF(string2 < 48 OR string2 > 57)THEN GOTO GPS_Mode 'Bad data
IF(string3 = 46)THEN x=string1:x=x-48:x=x*10:y=string2:y=y-48:ch=x+y:GOTO Continue_A '2 digit#
IF(string3 < 48 OR string3 > 57)THEN GOTO GPS_Mode 'Bad data
x=string1:x=x-48:x=x*100:y=string2:y=y-48:y=y*10:z=string3:z=z-48:ch=x+y+z '3 digit#
Continue_A: IF(ch > 360) THEN GOTO GPS_Mode 'Bad data

I would highly recommend Parallax upgrade to waypoint and WAAS capabilities. The amount of trigonometry involved to calculate any type of advanced navigation without it would appear to be mind-boggling to say the least for a basic-stamp. But, there are a lot of smart guys here who could probably do it. Unfortunately I’m not one of them.

Don’t get me wrong, I think Parallax stuff is a wonderful and I spend nearly everyday fooling around with something or another. Well that’s my two cents for what it’s worth.

Enjoy!

-J

Fe2o3Fish
04-12-2007, 02:00 PM
Jonathan said...
As well as the error in Google maps etc, is the fact that the USGS map they are taken frome aren't perfect either. Some areas have not been surveyed since the '40's or even earlier. I know this from much backpacking and taking GPS readings from places like Mount Shasta in CA. It just doesn't always line up with major landmarks, and the transition between quads can be off as well.

Also, while it was reduced a while back, doesn't the Govt. stiil introduce some error in the GPS readings? Both of my mobile GPS units can be set to an averaging mode, where the longer you let it sit in one place the more accurate the reading becomes.

If the area hasn't been surveyed since the '40's... any associated maps would be using datum NAD27...
which, unless you re-configure your GPS, is gonna be a bit different from WGS83 that your GPS usually
defaults to. That alone can be plenty of discrepancy 'tween the maps and your GPS.

On May 1, 2000, then President Clinton, issued an announcement that Selective Availability would be
turned off... and it was. This was the "jitter" induced into each of the GPS satellites' signals that would
cause some error in your calculated position. So no, I doubt much, if any error, is getting introduced
as it would be noticed by the folks (usually academia) that monitor their surveyed positions versus
those calculated from GPS.

Keep in mind, y'all have consumer grade GPS equipment -- currently it only receives one of the channels
coming down from the GPS satellites. Because of this, additional information from the GPS sat's can not
be obtained nor can your receiver monitor how the ionosphere is bending the signal from the GPS to you.
GPS position calculations are based upon the signal from the satellite travelling along a straight line to your
position. The ionosphere will refract that path somewhat and introduce a bit of error in your position.
The gov't ain't doing that -- Mother Nature is! http://forums.parallax.com/images/smilies/smile.gif And we won't talk about signals bouncing off of buildings,
geological formations, etc and a few other things that can mess with a GPS signal. Oh, and the calculations
done in your GPS receiver will introduce some error in your position simply because of how computers
crunch floating point numbers, or how good their internal crystals/timekeeping is, or how good their
algorithms are, or... [Sorry, I'll stop now] And this is just the tip of the iceberg.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
-Rusty-
--
Rusty Haddock = AE5AE = rusty@fe2o3.lonestar.org
**Out yonder in the Van Alstyne (TX) Metropolitan Area**
Microsoft is to software what McDonalds is to gourmet cooking

steve_b
04-12-2007, 07:33 PM
Fe2o3Fish said...


Keep in mind, y'all have consumer grade GPS equipment -- currently it only receives one of the channels
coming down from the GPS satellites. Because of this, additional information from the GPS sat's can not
be obtained nor can your receiver monitor how the ionosphere is bending the signal from the GPS to you.
GPS position calculations are based upon the signal from the satellite travelling along a straight line to your
position. The ionosphere will refract that path somewhat and introduce a bit of error in your position.
The gov't ain't doing that -- Mother Nature is! http://forums.parallax.com/images/smilies/smile.gif And we won't talk about signals bouncing off of buildings,
geological formations, etc and a few other things that can mess with a GPS signal. Oh, and the calculations
done in your GPS receiver will introduce some error in your position simply because of how computers
crunch floating point numbers, or how good their internal crystals/timekeeping is, or how good their
algorithms are, or... [Sorry, I'll stop now] And this is just the tip of the iceberg.


Just to add a little fuel....we have some scientists that are looking at the GPS signal lag when there's cloud/precip in the area. Apparently it's a measureable lag that correlates to a certain amount of liquid in the air.

So....rain, rain go away, My GPS can't find my way! http://forums.parallax.com/images/smilies/tongue.gif cheez

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Steve

"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."

Tracy Allen
04-13-2007, 12:04 AM
FYI: Here is a link to the Plate Boundary Observatory (http://pbo.unavco.org/), which is set up as an earth scientific study of the continental plates. However, a lot of the data is available to the publc. These stations are based on the Trible netRS, which claims an accuracy in fixed stations down to less than 10mm. I found out about this from an CalTrans survey engineeer--CalTrans has provided several of the station locations, and they use the data for bridge and highway planning/construction.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com (http://www.emesystems.com)