Slashdot Mirror
GPS Receivers?
An Anonymous Coward asks: "I need a GPS receiver for some field research, and am having some trouble finding one that meets my requirements. I need to have the coordinates and time recorded once per second for an extended period (12+ hours) with a decent accuracy (5 meters or so). Ideally, I'm looking for a device which will record real-time to my Dell notebook (running Windows XP), but I'd be okay with one that recorded into it's own memory as long as I could download to my Dell easily and: a) it has enough memory to save a days worth of data or b) it will save about an hours worth of points but is cheap enough that I can get two (so I can tag-team them). I'm working on a research budget, so the solution has to be under $800 (but less is better, or course). I thought a PCMCIA card receiver might be a good solution, but it seems like the stand-alone units have more to offer, so I'm open to either option."
You might want to look into a Handspring Visor and a GPS springboard expansion. You can get a refurbished Visor Deluxe for about US$99 and a good GPS package for under $200. That'll give you a good sized display and lots of flexibility.
Why doesn't the 99 dollar WalMart Garmin special with the Serial connector work for ya, should supply all the goods you are looking for. As for the whole accuray bit, that is really going to depend on WHERE you use it, lots of hills, trees and stuff overhead, might get 5 meters, but then again maybe not.
Based on your criteria, I don't see why you'd have any problem finding one. There are dozens of GPS's for < $200 USD that have serial output. You could (Dog forbid) even use Hyperterm to log the data to a file.
Most of the headless GPSs do not have internal batteries, but a cheap 6 amp hour 12V sealed lead acid battery or gel-cell will run it for days. The one I use in an application (which is standalone, but requires a PC board to mount it) pulls 30 milliamps. A 6 amp hour battery would run it for approximately 180 hours.
Check Garmin, Sirf, Motorola, etc. There is no shortage of lost cost GPSs
--jcwren
Here, this is the basic 99 dollar job, pick up the 20 dollar cable and you have what you want, updates every second.
If you want a little more percision then go with this, but the whole DGPS (less then 5 meter accuracy) thing only works if there isn't a large amount of trees or hills, it is about twice as much but comes with the cable.
Without DGPS or WAAS you will not get readings with an accuracy of less than 5 meters more than about 60% of the time. (You could spend 5 years researching and understanding the statistics related to GPS units). Most units are specified to be within 15 meters 95% of the time. They do NOT accuratly tell you when they are NOT within that tolerance. They may give a confidence level of one type or another, but those again, are based on statistics and are not 100% accurate.
With DGPS you can have accuracies within the range you want if your in an area that can receive DGPS signals. A DGPS receiver will cost you a few hundred dollars, and the GPS you use has to be able to use DGPS information. (A DGPS capable GPS receiver will NOT pickup the DGPS signals, rather it needs an external receiver to get the signal, but it will process the data.) (You may find an exception to this, I expect it will be outside your price range by a wide margin).
WAAS might work for you, but then again, it might not. If you on the East, or West coast you will have a higher probability, but even then it has been rather spotty for many people and is really intended for aircraft (significantly above ground) usage. The satelites are too close to the horizon for most people and are blocked by ground clutter. Optimally WAAS should get you within 3 meters, but it might only increase the accuracy of the estimated error.
Note: Getting readings that indicate within a couple meters of each other for a period of days might get you an average location with a low margin of error, but again, it might still be WRONG. Some methods of determining location are known for being inaccurate but consistant. I understand ( but do not know the specifics) that Loran is known for this. In a specific location you can expect a VERY similar reading for the same spot in the future, even if the reading is off by a wide margin from reality it will be consistantly off at that location. GPS doesn't do this significantly. With the satelites moving at a high rate of speed and various objects reflecting the signal depending on the current location of the satelites the error will not necessarily be consistant from day to day, or hour to hour.
(On the other hand, the orbits do repeat and shift slightly over a period of 24hrs so a reading at 4 o'clock today will have similar satelite coverage tomorrow at 3:55, or 4:05 (not sure which off hand).
If you want more stable results try to aquire your data at night. A significant margin of error is introduced by the ionisphere and this is reduced at night.
All of the above is generic GPS information....
Now, the reason I suggested GARMIN, GARMIN GPS units support a mode which allows you to extract data required for RINEX processing. The programs which extract this data are using undocumented functions but they do work. (see: This website) RINEX data can be post-processed and combined with external data (freely available after a delay of a few days generally) which can be used to get very accurate results. I believe the estimated accuracy for a GARMIN unit with RINEX based post processing is about 1/2 meter.
The program to collect the RINEX processable data can be run on a laptop or PocketPC connected to almost any Garmin GPS (from the older GPS12, to the newer etrex)
I have a Garmin Vista (about US$300), RS-232 Cable and Cigarette Lighter plug (about US$30), and Ram-Mount (about US$20).
What a fantastic combination for driving. Tha maps (US and Europe (about US$175 each) rule.
I usually use NiMH batteries when walking. They last a few hours without charging. I recently used a pair of alkaline AAs for a two week trip through Europe and only had to change them once! Granted I didn't use it constantly, but I did use it quite a bit.
I use gpspoint on Linux for saving, uploading, and downloading routes, waypoints, and tracks.
and unlike garmin models, the 'make route (from track)' uses all of the trackpoints (memory limited)- so with your many ($44) 64mb cards, you may find that you can track for, say, 18 hours (it's a hypothetical number)- whereas with garmins, you'd need to take the laptop- b/c when they save routes to tracks on the GPSr- they don't use ALL the trackpoints (breadcrumbs) they use only some, so your get a much 'rougher' faceted travel path.
all the newer magellans can communicate detailed info with your laptop (thanks to the NMEA data stream they support)
try this: http://www.gpsinformation.net/mgoldreview/mag-gold . tm
If she floats, she's a witch.
both have very poor reception under moderate cover.
WAAS may help, may not, but get a better antenna than any eTrex has for better accuracy.
don't beleive me? go to www.geocaching.com forums and see the complaints by eTrex owners who can't find the spots under trees (and they're talking 25m+ in-accuracies, if not total receptoin loss)
If she floats, she's a witch.
Not sure what Garmin recievers these people have, I actually have both the etrex and the venture, and they work great, shit I regularly get 3 meter with my venture. Don't belive everything you read buddy. And lets be a little more constructive here and actually offer help, bot just slam things because we heard from someone that it wasn't good.
You did say +/- 5m so I am guessing you do know this, but if not you may want to be sure you are getting what you think you are getting (apologies to "The Princess Bride").
sPh
This is only a receiver, with single cable ending in a RJ-45 connector for serial and power.
I power the unit from the laptop's keyboard port (lvs == low-voltage supply), using a handcrafted converter to split serial to a Cisco-standard RJ-45 from the Garmin's RJ-45.
One feature not requested -- the GPS-16 provides a PPS output, making it suitable for use as a very accurate time source for NTP or simply for timestamping.
I do not deploy Linux. Ever.
You only need 3 sat for position if altitude isn't taken into account. (My unit switches from 2D to 3D mode if a 4th satelite is aquired). 2D mode isn't as accurate, but is quite usable for a quick position fix.
A typical receiver will quickly locate itself -if- it has a recent almanac and is not significantly relocated since it was on last. (ie: it's best guess is close enough.).
The most extreem cases seem to be a few hours for units that have significantly out of date almanac and have been relocated significant distances. (my older Garmin suggests that a difference of 500 miles is enough to slow down initial acquisition. While I'm sure newer units are even better, this has held true for all uits I have personally seen. I've head descriptions of units being off for -years- and having run out of power (therefor no almanac) taking a few hours to aquire. Not a typical scenario. (As well, Garmin units can retrieve almanac data from another garmin unit and that will help them aquire in such circumstances, I'm sure others can as well).
This extract from the documentation suggests acquisition times a lot quicker than the few hours you mention: