best gps for animal tracking collar?

Hi I’m building some GPS collars for some biologist and i was wondering if you could give me some input on the best gps unit. I’ve be testing some arduino code with the MICRO-MINI, but i’ve been getting some strange readings, and bad accuracy. I think it’s probably a weak signal strength issue. I borrowed a parallax gps with the much larger antenna and the results were much more accurate (no offense) I tested them both with GPSFOX and the parallax had a much stronger signal. I’m looking for another GPS to replace the Micro. I’ve also learned that gps with chip antenna need to be pointed toward the sky, but I’ve heard that helical antenna models can be oriented in any direction and still lock.

This unit needs to be battery powered so the lowest current draw possible is best, and its going to be on a collar, so it needs to be able to get a lock from any orientation. I originally thought of a SUP500F but its a chip antenna and i’ve read about all of the problems on the forums. My current choice is the D2523T http://www.sparkfun.com/commerce/produc … ts_id=9566. It has the lowest current draw of the helicals. and a very quick cold start time so it can be turned on a lock quickly and shut off to save power.

Am I wrong about the D2523T, is it better then the Micro-Mini, do I need the helical antenna?

Thank you for any help

For GPS accuracy is mostly dependent on the number of satellites in view of the device. Either your antenna is an orientation based antenna, or your antenna doesn’t have enough gain to pick up enough sats to be accurate. You need a minimum of 3 connected birds to get a decent fix, but I wouldn’t trust less than 4 or 5. There’s literature on GPS out there that disagrees with me, but it’s been my experience that you can’t get trustworthy locations. Maybe it’s just the lowest bidder MIL-SPEC crap we work with. Anyhoo, I digress.

The best thing you can do for a GPS is give it the biggest highest gain antenna your design can afford. External antennas are best. The higher the incoming gain the better. There are some design considerations with this, however.

What size animals are we talking about here?

How long of a sampling period does the biologist need? (Days, Weeks, Months?)

Is this for data logging the animal’s position over a period of time, or is it to be able to locate the animal?

How big is the tracking area?

Just some thoughts. I’m sure others here have more questions about the design requirements, so feel free to post all the gory details.

Good luck tinkering, and try not to get eaten by a lion (or whatever).

MB

We’d like to avoid external antennas for fear of damage. There maybe another module that does gps data logging or that may be added later, My gps is to determine if the animals enter areas where they aren’t supposed to be (farms, villages,etc) and reports back via rf. The collars are going to be for many species, wild boars, water buffalo, and elephants. the time line is going to be about 3 months for the boars, about 6-8 for the buffalo, and about 18 for the elephants. The GPS is only going to activate once or twice a day and if one of the zones is within the one day range of the animals it will activate more often to determine if it is heading towards a zone. if it gets within a certain range it will send an rf signal to a base station.

The micro-mini i’ve been using usually gets about 3 satellites in its lock, the problem is I can be sitting still with the gps sitting on the ground and drift a few 1,000ths of a degree in just the 30 seconds i was watching it. which should mean i moved about 600 feet. I’m working on modifying the GPS speedtrap for arduino http://nootropicdesign.com/projectlab/2 … peed-trap/. The Point in polygon algorithm that he uses is perfect for what i’m doing, but I’ve been getting some strange readings. If i set gps zones miles away from where i test, it will occasionally report i’m in one of those zones. I switched to a parallax gps that looks like the EM-406 and that mostly fixed the problem, I’ll still get a random wrong zone every now and then (it could be the large buildings i’m testing near) but i’ve fixed that so far with some code.

I need to keep the systems small for the boars, but we can add larger batteries to meet the time lines for the buffalo and elephants.

I tried testing the Parallax gps yesterday by holding it upside down and trying to get a lock, it didn’t work, if the collar shifts then the gps won’t work, will the Helical antennas like the GS407 and D2523T work upside down or horizontal? if so I think i will go with them. otherwise I have to figure out how to keep it upright.

Don’t worry I’m staying nice and safe here. They can go hang with lions if they want!

All of the GPS modules that I’ve used are going to “wander” over time as more and more satellites come into view. You may want to monitor the “HDOP” parameter to determine when it is within acceptable limits, especially if you are going to powerdown the unit in between fixes.

I’ve used the EM406A quite a bit, not sure what it’s power draw is compared to the units you’ve mentioned, but it has very good sensitivity in just about any orientation.

The FALCOM FSA03 is also another very good one with a helical antenna, and quite a bit cheaper than the 2523, but the antenna isn’t very robust unless you provide some means of stabilizing it mechanically.

– Greg

hi, do you have any idea where to buy this animal tracker device? or any other GPS device that is inexpensive and is available in south east asia… I may order a bulk to start a business.

BigRedBee:
All of the GPS modules that I’ve used are going to “wander” over time as more and more satellites come into view. You may want to monitor the “HDOP” parameter to determine when it is within acceptable limits, especially if you are going to powerdown the unit in between fixes.

I’ve used the EM406A quite a bit, not sure what it’s power draw is compared to the units you’ve mentioned, but it has very good sensitivity in just about any orientation.

The FALCOM FSA03 is also another very good one with a helical antenna, and quite a bit cheaper than the 2523, but the antenna isn’t very robust unless you provide some means of stabilizing it mechanically.

– Greg

There is the possibility of encapsulating the circuits with epoxy or something, we’re still working out the details so i don’t think weak antenna would be a problem. Have you used the FSA03 before. It seems to be perfect for what I need.

Thank you

Yes, I have built some custom RF trackers using these modules. let me know if you have any specific questions. The one thing I know is that if you issue commands and then remove power they will be forgotten unless you have the battery backup in place, there is no ‘non-volatile’ storage implemented on the FSA03

greg ‘at’ bigredbee.com

Thanks I was just planning to use the default NMEA at 9600 and to power off between readings. if the Vbat current is low enough i might tie that directly to the system batteries if it would significantly reduce the acquire time. I’ll get some ordered.

Thank you very much

I’ve done a number of GPS based animal tracking systems.

You’ll want to use a helical antenna. As you noted, it is not as sensitive to position + it also doesn’t require a ground plane as a patch antenna does. Also it is not as effected by the human/animal being near it as a patch antenna.

The Sarantel antenna is the one most frequently used, and they are fairly fragile. What I have done in the past is make a plastic cover for the device… it was fairly rugged. I’m actually doing a collar now, and its fairly small…

uBlox GPS’s are good, a bit pricey though. I’ve also had good luck with the Media Tech GPSes’

What are you planning on using for your two way radio or comm device?

I would be interested to know about your dog collar. Have a look at pawtrax.co.uk