PDA

View Full Version : what gps for uav


ALtitudeap
Jul 09, 2005, 09:35 PM
i dont have a nav system yet. what do you recommend for gps and nav unit. budget choice and no budget choice. thanks guys

Myron
Jul 10, 2005, 12:06 PM
You have alot of shopping to do on that one!.. Some autopilots have an board GPS included/built into the system yet others you have to use an of the shelf unit with a serial out. You need to look at your mission paremeters and then start your shopping after you know the what when where and hows...
You can get with Dave Jones at AUAV.net and he can help you out quite a bit. As for pricing, your looking at 3K and up for a full blown autopilot and GPS system... Alot of the more simple solutions use Garmin Gecko or E-Trex GPS and a device that is called a PDC-10. That device hooks to a serial GPS and it will vavigate via rudder control only. You can augment controll with flight stabilization system like an FS8 Co-pilot from FMA...

Keep reading for more info!!!

Myron

ALtitudeap
Jul 10, 2005, 02:53 PM
thanks myron. i am looking at the AP4 with the Geko 301. Quick question though. When you put it into auto mode. does the throttle lock to last known possition or how does it know how much throttle to give. in the event it is flown out of sight.

rjet
Jul 10, 2005, 04:16 PM
I use my radio's failsafe to set the throttle when I fly beyond radio range with the AP4.

ALtitudeap
Jul 10, 2005, 04:19 PM
what radio do you use?

Myron
Jul 21, 2005, 11:10 AM
I have a Futaba Super 8 PCM and a 9CAP PCM... Both have "failsafe" settings.. You can also set up failsafe functions with an FS8 Co-pilot...

Myron

TimAuton
Jul 21, 2005, 01:12 PM
The Paparazzi UAV (http://www.nongnu.org/paparazzi/) folks use a u-blox GPS (http://www.u-blox.com/products/sam_ls.html) . 4Hz update, 22g including antenna. $130 probably puts it more towards the ideal end. I wouldn't recommend a handheld GPS (Gecko etc.), it's a got a load of extra weight you don't need (screen, case, batteries...). There are many other GPS modules available for less than the u-blox that weigh much less than a handheld unit. A search for "GPS module" should find plenty. I've got a couple of these modules (http://www.rfsolutions.co.uk/acatalog/Board_Level_GPS_Modules.html) which are 25g including antenna and cheaper than the u-blox, but only update at 1Hz.

rixkix
Jul 22, 2005, 12:19 AM
The advantage of the handheld GPS is the fact that the you can program the route into it either directly or through a serial cable with most mapping programs. Makes the setup much simpler and more modular.

typicalaimster
Jul 22, 2005, 12:48 AM
Hmm for the price you are paying for a U-Blox AND the board, just go with a U-Nav Pico GPS. It's only $157 after shipping..

clolson
Jul 22, 2005, 07:05 AM
Has anyone taken a look at the etrex summit/vista which supposedly has a barometric altimeter and magnetic compass built in?

TimAuton
Jul 22, 2005, 11:00 AM
Hmm for the price you are paying for a U-Blox AND the board, just go with a U-Nav Pico GPS. It's only $157 after shipping..

The $150 U-Nav Pico GPS is just a GPS, not an autopilot. A GPS which costs $20 more than the u-blox SAM-LS, has 12 channels vs 16, 1Hz update vs 4Hz and weighs 28g vs 22g. You can find equivalents to the U-Nav Pico GPS for half the price.

The U-Nav autopilots start at $350 for rudder only and you have to add a GPS.

kd7ost
Jul 22, 2005, 12:34 PM
Are you saying the Paparazzi UAV uses a u-blox gps for 130.00 and now you have an autopilot? Tell us more if you can. That sounds great.

Dan

TimAuton
Jul 22, 2005, 02:22 PM
I was mainly just talking about GPSes, I'm (slowly!) building my own autopilot so I don't really know much about the off-the-shelf autopilots. The u-blox SAM-LS is just a (rather good) GPS, not a whole autopilot system.

The Paparazzi (http://www.nongnu.org/paparazzi/) project uses a custom autopilot board. Probably cheaper than any bought-in autopilot, but you have to like hacking electronics and code to make it work.

I'm expect something easier to implement will come from the open-source UAV community in future, but right now it's not really an option if you're not already a hardware and software hacker.

radiohound
Jul 23, 2005, 02:54 AM
Has anyone taken a look at the etrex summit/vista which supposedly has a barometric altimeter and magnetic compass built in?

The Garmin Geko 301 is much smaller and has an altimeter built in, as well as a compass. That's what I was planning on using, although the Ublox looks very nice and small indeed.

ALtitudeap
Jul 23, 2005, 03:29 AM
if i will be programming my waypoints directly into my homemade autopilot, will i be able to use something like the u blox?

TimAuton
Jul 24, 2005, 12:59 AM
If your autopilot is dealing with the waypoints any GPS will be fine, no need for the GPS itself to have programmable waypoints. All the GPS does in that situation is constantly (once or more per second) output the current position information, which every GPS with a serial port can do, u-blox included. Most, if not all, are capable of doing it in a standard format - NMEA 0183 - which is understood by everything from autopilots on oil tankers to in-car navigation systems. If you're writing your own code there should be some free libraries out there to deal with the NMEA data, which gives you loads of choices for the GPS. If you stick with NMEA it's easy to change GPS units and have it Just Work(tm), it's refreshingly simple compared to interfacing most sensors.

Most GPS units also support one or more proprietary formats in addition to NMEA, but for most people there isn't much benefit to using them.

ALtitudeap
Jul 24, 2005, 02:04 PM
TimAuton,

I think that the oem gps board is the way i want to go. which ones have you had best success with. (budget in mind)

ALtitudeap
Jul 24, 2005, 03:32 PM
will this work ?
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=5792192722&category=4673&ssPageName=WDVW&rd=1

pimmer
Jul 24, 2005, 03:47 PM
The PicoPilot form unav.com is the best all around nav unit. It is $800, though, but includes everything. GPS, altitude and plugs directly into the servo ports.

ALtitudeap
Jul 24, 2005, 03:52 PM
i am looking to make my own very specific auto plane.

moon-dawg
Jul 25, 2005, 02:04 PM
Clolson,

I looked at the Etrex Vista. The compass and barometer features are really neat for hand held use but pretty much useless for in-flight. The unit has to be level to get a compass reading. Plus if you intend to use the unit with an autopilot such as the AP4 or RCAP the compass and barometer outputs are not in the standard NEMA data streams the autopilots decode. Also, no need to pay for a GPS unit that provides more than a 1 Hz update rate because that is all that the GPS system provides.

clolson
Jul 26, 2005, 12:12 AM
Clolson,

I looked at the Etrex Vista. The compass and barometer features are really neat for hand held use but pretty much useless for in-flight. The unit has to be level to get a compass reading. Plus if you intend to use the unit with an autopilot such as the AP4 or RCAP the compass and barometer outputs are not in the standard NEMA data streams the autopilots decode. Also, no need to pay for a GPS unit that provides more than a 1 Hz update rate because that is all that the GPS system provides.

I'm going to be implimenting my own flight computer so in theory it should be no problem for me to decode additional non-standard gps sentences. I suspect that most of the time the plane will be pretty close to level and my navigation system will depend more on ground track anyway. But the magnetic compass could be interesting to help estimate wind?

Are you implying that the garmin vista can do better than 1hz, or are you referring to some other unit.

Thanks,

Curt.

moon-dawg
Jul 26, 2005, 10:33 AM
Curt,

I was refering to the u-Blox unit when I mentioned greater than 1 Hz.

In the real world, a flight control computer keeps a system state vector that is calculated at 50 Hz (even 200 Hz on digital flight control systems like the B-777). It uses inputs from inertial, radar alt, baro alt., etc. to calculate the system state vector. A Kalman Filter is used to estimate the errors in inertial at each 50 Hz interval and they are subtracted out of the system state vector. At the time of the GPS update the GPS value is assumed to be the absolute position so the GPS position is substituted for the system state vector position and the errors are zeroed out in the filter.

If you have a GPS that updates at greater than 1 Hz then your inertial will drift less between updates. In the case of the u-blox, it is running its own filter that estimates errors and provides a position value at 4 hz rate. If you are writing your own software then you can also write your own Kalman filter. So no need to use a GPS that provides more than 1 Hz update.

There is Kalman filter C source code available on-line, see http://sourceforge.net/projects/autopilot/

kd7ost
Jul 26, 2005, 12:07 PM
I emailed Garmin tech support a few weeks back about the improved altitude information in the BP enhanced GPS units. I was told that the unit does output the altitude correction improvements into the appropriate strings in any of the serial formats. I didn't ask about compass info.

These all sound like very cool projects but I have to throw in here that if you're flying model aircraft or Mini UAS that typically fly around 100mph or less, airframe dynamics and flight speeds dictate that you won't need to refresh at 50 hz. The 1 per second refresh rate from the handheld units is certainly observable as you watch it, but the plane is slow enough and dynamically stable enough that this isn't an issue for guys who wish to pursue that method of flight. You can fly to the same waypoints with a several hundred dollar system as you can with a several thousand dollar system. (There are certainly variations in how it works though) The faster planes need faster refresh so they stay tighter on course. This all presumes that you started out with a dynamically stable airframe.

However, the complexity and detail that goes into higher end or more complex systems opens up a fascinating realm of work with really fast platforms and ones that are not as stable. You can get higher performance type maneuvers in a fast refresh delta design for example.

Edited to include, I mean flying realy fast at "When flying at 150 mph or more."

Dan

radiohound
Jul 26, 2005, 01:15 PM
I emailed Garmin tech support a few weeks back about the improved altitude information in the BP enhanced GPS units. I was told that the unit does output the altitude correction improvements into the appropriate strings in any of the serial formats. I didn't ask about compass info.
Dan

The compass result is put directly into the heading string. You can select at what speed the compass is used over the gps (change in position) computed heading. So if your true speed is say less than 10 mph, it would use the compass heading.

Here are the specs that compare the Etrex Summit to the Geko. Geko is lighter, smaller and equal or better than the Summit as far as specifications go.

When you compare weight, remember that many of the OEM's will need an antenna.

The Geko also has WAAS capability, a bit difficult to recieve, but should be easier at a bit of altitude.

TimAuton
Jul 27, 2005, 12:50 AM
I think that the oem gps board is the way i want to go. which ones have you had best success with. (budget in mind)

I'm using two of these modules (http://www.rfsolutions.co.uk/acatalog/Board_Level_GPS_Modules.html), which work just fine and I've not noticed any interference from them. WAAS is a real bonus, if you live somewhere with WAAS coverage (Europe, the US and I think Japan) go for a module which supports it.

moon-dawg
Jul 27, 2005, 11:35 AM
I think that the compass heading is only in this NMEA message on the Garmin eTrex.

$HCHDG
Compass Heading

This sentence is used on Garmin eTrex summit, Vista and GPS76S receivers to output the value of the internal flux-gate compass. Only the magnetic heading and magnetic variation is shown in the message.

$HCHDG,101.1,,,7.1,W*3C

where:
HCHDG Magnetic heading, deviation, variation
101.1 heading
,, deviation (no data)
7.1,W variation

moon-dawg
Jul 27, 2005, 11:45 AM
The RCAP and AP4 GPS Autopilots only decode the NMEA $GPRMB and $GPRMC sentences so they would ignore the heading info from an internal flux-gate compass.

moon-dawg
Aug 19, 2005, 11:06 AM
I was surfing the net and found this source for a PG-31 GPS engine based on the SiRF star II e-LP chip set. The PG-31 has full WAAS and EGNOS capability. $78.95. A full evaluation kit, including PG-31, carrier board and antenna is $139.95

http://www.sparkfun.com/shop/index.php?shop=1&cart=371594&cat=63&


They also sell a cellular module that could be used for a data link. $119.95

http://www.sparkfun.com/shop/index.php?shop=1&cart=371594&cat=75&


If your looking for a new car phone - this one would look pretty cool sitting on the dash of you SUV - it is definitely not hands free

http://www.sparkfun.com/tutorial/Port-O-Rotary/portable-rotary.htm


there are also RF datalink and a wirless camera that could be used in your UAV

http://www.sparkfun.com/shop/index.php?shop=1&cart=371594&cat=62&

Peter Seddon
Aug 19, 2005, 01:17 PM
Another great design from the UK!!!

regards Peter

kbosak
Feb 13, 2008, 07:50 PM
Also, no need to pay for a GPS unit that provides more than a 1 Hz update rate because that is all that the GPS system provides.
PLS elaborate on that. This would be an important point, stating that we can do our own Kalman + temporal extrapolation to get better results than any pro 20Hz GPS for 2000USD.

DrDon
May 18, 2008, 07:58 AM
We have found that the update rate for several GPS manufacturers differ significantly on update rate but in fact that the GPS measurements are NOT limited to 1Hz. Higher quality GPSs can provide raw GPS measurements from 10-20 Hz and it is a limitation with the onboard processors/software to calculate the GPS solution. The fact that solving the GPS based position and velocity based on the raw measurements is an iterative nonlinear problem is the reason most systems only output 1-4Hz.

We have recently made some of our OEM GPS modules available if anyone is interested.

DrDon
http://www.iat-llc.com

blade thrasher
May 28, 2008, 09:27 AM
We are working on a 63" predator B UAV in hopes of having functional autonomy along with FPV and still shot from a 8mp cam. (co-pilot stabilized)

Would the use of an AP4 and Etrex be too much weight once the video, still and Tx are added to the 63" drone? We were also considering adding ailerons to the ship but this would interfere with the Co-pilots control of the V-tail.