SparkFun Forums 

Where electronics enthusiasts find answers.

Tips and questions relating to the GPS modules from SFE
By soulchyld
#119426
Hi, I would like to get my hands on as much information as possible about GPS systems, I plan to set up a tracking system in a developing country and was wondering what exactly would be required, I know the GPS module needs to be able to send data, therefore a GSM/GPRS/3G package of some sort would be required on the tracker (GPS) end, but what is the easiest and most cost effective solution on the receiving end, and would I be able to get these components from Spark Fun?
User avatar
By itikhonov
#119504
If you are going to send data via GPRS/3G on the receiving end the only thing you need is an internet connection.

So you need GPS, GSM module, basic microcontroller to couple them and power supply. Sparkfun sell them all. Be sure to check datasheets for additional components needed (like capacitors, antenna's and etc). Or just buy a breakout boards. Or just grab arduino versions of boards and arduino itself (pricy but everything already assembled and probably you'll find complete code for your solution).
By soulchyld
#119564
itikhonov, thanks, how about if I use GSM, obviously now making a call, could I get a GSM module on the computer end and make the call from there? this would be cheaper for me because I would only make calls when I wanna know where the vehicle is,

Ohara, yes this looks interesting, you think its possible for me to have my own solution instead of using twitter? would I be able to make a call from the receiving end and receive the co-ordinates that way? roughly how much would each tracker unit cost?
User avatar
By elevator4
#119580
Hi soulchyld,
it is certainly possible to have it your way, ie have the base station interrogate the tracker.
You can either 1) set up a 9600 baud GSM connection, or 2) do it with SMS messages (an interrogation message from the base station to the tracker, then have the tracker send a reply SMS.
There are pros and cons both ways, these come to mind:
with 1) the tracker needs enough time, during the call, to get a valid position fix. this may need up to a minute or more, depending.
with 2) the tracker gets billed for the SMS transmission, which may be impractical for your application.
Also, you want to be sure there is GSM coverage in the area :wink:
Cost - over the thumb: 55$ for the GSM including antenna; 30$ for the GPS including antenna; 12$ for the rest (micro, battery, PCB, case etc.) Obviously depending on quantities.
Hope this helps,
E
By soulchyld
#119584
Thanks, this is getting better and better, what about a program to track the vehicle, I have seen some people using google earth, have you any idea how this is done? Yes, it seems mobile network coverage is available in most parts of the country nowadays so thats a plus, so what we are saying is we can put a tracker together for about 100 bucks?
User avatar
By elevator4
#119591
Hi,
A quick way to learn what can be done with google maps is to browse http://www.gpsvisualizer.com, playing with http://www.gpsvisualizer.com/draw/. You can set marks, tracks and areas, then pan and zoom.
How is it done? It can be as simple as bringing up a browser window (using the same example site), pasting the coordinates as follows:
http://www.gpsvisualizer.com/draw/?type ... &note=This Location
Of course, the whole process can be customized and automated but takes some prior programming :|
To the cost of goods, as I said it depends on quantities. To get to 100$ - again a wild guess - you'd need to order a couple hundred pieces. For quantities to a few dozens or so, I'd recommend Sparkfun.
Also, the hardware needs to be prototyped, a PCB designed and fabbed.
I'd be glad to help where I can.
E
User avatar
By itikhonov
#119633
soulchyld wrote:itikhonov, thanks, how about if I use GSM, obviously now making a call, could I get a GSM module on the computer end and make the call from there?
Are you sure you don't want to use internet connection via GPRS? To make it cheaper you can do it this way:

1. Device waiting for call.
2. You call it
3. Device sees the call but reject it immediately so you are not charged.
4. Device turns on GPRS connection and send/receive information to/from you (via email, web, tcp, udp, etc).
User avatar
By elevator4
#119637
itikhonov wrote:Are you sure you don't want to use internet connection via GPRS?
@itikhonov
sure that's possible. I was just being conservative in my suggestion to use basic GSM because we're in a developing country, so GPRS service coverage may only be partial at best. BTW once you have the system set up with basic GSM using a module like the Telit GE865 or thereabouts, you can always upgrade its firmware or app script over-the-air at no cost.
E
By soulchyld
#119689
@ Elevator, I like what gpsvisualiser does, I guess now Im getting ambitious, you think its possible to be able to develop something that will enable us to monitor live movements? Im really excited about all this, I guess another option is to get more than one set of co-ordinates from the tracker at a time, maybe at 3 or 4 min intervals for the live movements?
User avatar
By elevator4
#119709
soulchyld wrote: [..] maybe at 3 or 4 min intervals for the live movements? [..]
Sure, you can also get instantaneous speed and direction.
soulchyld wrote: [..] I guess now Im getting ambitious, [..]
Good. Having worked on about a dozen GPS receiver projects using different chipsets, here is a very short list of caveats however:

1. Precision. While once "warmed-up", a GPS receiver delivers good precision, in a "push-to-fix" mode, i.e. intermittent operation, things may not always look as good. Fixes may be tens of metres off. Depending on the object you are tracking this could give you false indications. Now ten meters are negligible when tracking an aircraft in flight, but with eg a person, 10m make the difference between a report of "object is stationary" and "object is moving". The solution usually involves leaving the receiver powered for a bit longer, taking a sequence of fixes and averaging them, eliminating outliers etc. This however takes time and battery power.

2. Time. Don't assume your receiver will always be able to get a fix. Tough situations include indoors, under heavy tree cover, in a narrow city street lined with tall buildings, or just somebody's hand covering the antenna. In these cases and unless you program it otherwise, the receiver will just keep searching for satellite signals, forever. Searching (acquisition) is generally expensive in terms of power consumption. Remember that a GPS receiver either works acceptably, or doesn't work at all. When unable to get a position fix, as far as the receiver is concerned its location may be anywhere in the universe, so it will report "Dunno". To spare you this sort of frustration, program the tracker to report something like "No fix, but the last one was [lat/lon/alt] at 10:46PM..." when it can't see enough satellites after trying for a (predefined) time.

3. Battery power must be carefully budgeted: No answer and no report at all is even worse than "Dunno" :x

Antenna choice and placement, interference issues (eg with GSM!) are also important points.
The above is not meant to discourage you, rather to help set and frame your expectations, as well as to plan ahead and find workarounds for less-than-perfect practical situations.

To be more helpful it would obviously help to know more about your project: what are you planning to track, for how long, is the tracker unattended; size and preferred mounting location etc. Feel free to PM me.
E
By soulchyld
#119790
Thanks all this info is proving to be very useful, I want to track vehicles, fleet management is the main aim, and obviously in the event a vehicle gets nicked (God forbid), I don't think its necessary to track it all the time, (by track I mean physically see where it is) but probably it would be useful to log the data, maybe we could get it to start logging when a vehicle moves?, thats a great idea to program it to report the last time and coordinates it picked up! when you say unattended what do you mean? I guess for vehicles its best to have it hidden somewhere, that way drivers are not likely to tamper with it as well! with regard to time it would be for no longer than 10-15 minutes at a time(physically seeing where it is), just to get an idea of what drivers are up to really especially when out on long journeys.
User avatar
By elevator4
#120024
By unattended I mean whether the driver is able/meant to attend to the tracker, as for example pushing buttons for emergency or driver-initiated position reports. I see that your idea is different, as you want to prevent tampering by the driver. Ok.

Are these cars or trucks? In either case you’ll have to be creative finding a hiding place which does not attenuate the radio signals (GSM and GPS) too much. Also, some GPS antennas are better at picking up signals bounced off the ground, while the usual patch antennas cannot do that.

By ‘how long’ I mean how long does a long journey last? Because unless you wire the tracker to the vehicle battery, it needs to have its own, recharged before the trip starts.

For the rechargeable case, I did a few quick calculations and the results look encouraging: With a 1200mAh Li-Ion battery you can expect in the order of 2-3 weeks of operation at the following regime:
- two position inquiries per day over GSM
- one GPS fix (logged) every 10 minutes when moving
- one GPS fix every hour when stopped
Enclosure size, again after a few sketches, can be something like 80x40x12mm including battery. Plus some kind of mounting bracket, or magnet, or velcro strip :idea:

E