Question about HDOP.
Where I live is reasonably hilly so often I get a HDOP which is higher than the default (Which is around 2m I beleive) Im generally getting 3.5m or so.
Ive looked but cant find where I can increase the min HDOP value. so
1) Is there a parameter to change somewhere and
2) what would absolute safe min be? Im thinking between 3-4m.
Its usually whilst still on ground. Once up 20m signal improves, so it generally happily finds its way home, then I guide it to the ground if needed.
I currently have to disable the full preflight check (as it wont let me just set ignore GPS for some reason,keeps reverting to ignore throttle?) to let me arm.
I am in NZ and Have a reasonable Gps signal using std ublox 6 series unit ,, my Hdops are normally around 184.108.40.206 on most of my machines,
A piece of tin metal wheel under the GPS antenna often helps a lot on reception. Like 15 x 15 cm, maybe smaller is okay too. There is lots of experience on this to be googled.
"wheel", why did I, or autocorrect, type that? SHEET.
Ok I'll experiment, thanks all.
Its a stock standard Iris with the pixhawk position adjusted as they provided the template.
I have a second Octo ready to test with a GPS mast, so might be useful for comparing the 2 in the same location.
You should be getting HDOP < 2. If there's nothing blocking a large portion of the sky (buildings, trees, you), I'd suspect it's a noise issue. Even though noise isn't directly related to HDOP, it can jam some satellites which lowers the number of satellites detected and consequently your HDOP.
From my experience, the 3DR GPS is very sensitive to noise and is easily jammed. Examples of great jammers are: Beaglebone Black with HDMI enabled; the internal oscillator of the STM32F4 (I haven't checked PX4 firmware but I hope this is disabled!); Wii IR camera. You can literally turn these on nearby, lose GPS lock, turn it off and get lock back. The MediaTek MT3329 V1.0 didn't seem to have this problem as it was probably less sensitive.
Other sources of noise (like DC-DC converters and some hard drive motherboards), can affect it just enough to lose a few satellites and reduce HDOP. To gauge noise levels, the 3DR GPS actually gives you metrics called "horizontal accuracy, vertical accuracy and speed accuracy" (in meters and m/s), I'm not sure why these aren't used as feedback on APM or PX4 (as far as I know). I've done some experiments with this and it's not simply proportional to HDOP. You can see the accuracy decrease when noise is introduced, but HDOP remains constant.
I think I should write a proper blog post about this when I get a chance, with more detailed findings.
As a disclaimer: I don't have APM or PX4 but I use the 3DR GPS and my own STM32F4 based autopilot with similar components.
It's in the Full Parameter List. Click the Config/Tuning Icon. I checked with AP2 and AC3.15 and it's there for me.
Maybe Im blind, but I couldnt work out where to find the values in APM planner 2 (osx). I did a search for HDOP on the page but it didnt show.
maybe it's not only the mountains? Do you have an electromagnetic shield under your GPS antenna? Sometimes electronics close to the antenna can also reduce SNR and thus increase HDOP.
Its not only the mountains, its also the geometry of the sats available, have a look for programs that predict the best time of day to fly.