Servers by jDrones

Pixhawk2 antenna tracker?


(Sean) #1

I’m curious if the new pixhawk2 could be used for an antenna tracker, or if the firmware can run on other boards like a navio.


(Olivier Brousse) #2

Pixhawk2 being completely compatible, it should work, although I don’t have actual experience. Not sure about Navio but can’t see why not.


(James Pattison) #3

Navio with a pi3 makes a great brain for antenna tracker. As Olivier said, cube is completely capable, but to me feels a bit of a waste keeping one on the ground :slight_smile:


(Evan Williams) #4

I had attempted to use a Pixhawk 2.1 as an antenna tracker, per my second post here.

To rehash some content from said post:

When loaded with the code directly from Mission Planner (not compiled from GitHub source), the Pixhawk 2.1 moves in a scanning motion, but seems to have difficulties regarding compass orientation. Yaw and Pitch tuning respond properly to manual adjustment. The unit itself is pitched up by 90 degrees, accounted for in the parameters (AHRS_ORIENTATION = Pitch90).
Changing various compass orientations in yaw and pitch has no effect on MP’s displayed pointing direction (but changing roll for a compass causes pitch to return as yaw?). Perhaps something to do with the three internal magnetometers of the 2.1, as opposed to the single compass internal to the original Pixhawk or APM? In general the PH 2.1 demonstrates erratic behavior, often pointing straight up and rapidly oscillating in the pitch axis . . .

In light of this odd behavior, I recompiled the AntennaTracker firmware for px4-v3 (the design for the Pixhawk 2.1).

Where I ended up with recompiled code:

  • Mission Planner (MP 1.3.51) connected to Antenna Tracker (Pixhawk 2.1, AT 1.0) via USB, tracker connected to vehicle (PixRacer, Plane 3.8.3) via telemetry radio pair.
  • Successful forwarding of vehicle telemetry through the antenna tracker.
  • Correct compass readings
  • Simultaneous location display of tracker and vehicle on MP display.
  • Antenna Tracker is limp - no response from servos regardless of mode.
  • Mission Planner shows tracker is disarmed: when attempting to arm via MP, the MAV rejects command. When manually armed via the hardware switch on tracker, it arms, but immediately goes into safe mode (MP shows (SAFE)). In addition, I get periodic messages from MP that “Failed to update home location (2)”.

I’m having a hard time debugging the behavior, as I cannot seem to convince the tracker to arm and move the servos, even in a scan motion. Servos also do not respond to tuning commands.

A secondary issue is that the stock code uploaded from MP behaves very differently from that compiled and uploaded from the GitHub.


(davidbitton) #5

Shouldn’t the Pix spit our the reason for arming failure?


(Evan Williams) #6

I’ve not been able to track anything down as such - all’s green in the pre-flight check section of MP when on the Antenna Tracker stream (except battery voltage, as AT 1.0 does not support power monitoring). My understanding of the firmware’s operation is that it requires two things: a Mavlink stream from an associated vehicle, and a physical arming button (though occasionally it springs into action the moment it sees a telemetry stream).

I’m aware that there are likely more detailed messages coming through the stream, but I’ve not found a way to explore those - if you’ve a specific place in MP I can look, that would be a great help toward tracking down the behavior. Worst case scenario would be post- op review of the dataflash or telemetry logs, not exactly great for in-situ debugging.

Tomorrow I’ll be experimenting further with an original Pixhawk, which should operate as expected. That ought to clarify whether it’s a mere configuration issue on my part (admittedly, highly likely).


(Evan Williams) #7

Another update:

The Pixhawk 1 (FMUv2) works as expected. It arms, scans if no plane is seen, and tracks (see below) once a telemetry stream is established and the arming button is pressed.

There are still some issues with the heading, though. I’m using a HERE GNSS as the GPS and external compass. I can’t figure out how to get the internal compass to point in the right direction (it gives an incorrect heading for the unit, with greater than 45 degrees error when used standalone). Would internal compasses require orientation apart from the AHRS_ORIENTATION? I would imagine not.

I can get the external compass lined up, but turning the tracker aside and back triggers runaway drift that continues until the heading is backwards.

I’ll continue experimenting this coming week.

EDIT: Perhaps this change may have something to do with the drift of the HERE GNSS external compass?


(Evan Williams) #8

Conclusion:

I never was able to get the PH 2.1 running the AT 1.0 firmware properly.

However, the Pixhawk 1 is working fine with the same firmware, though it lags noticeably when tracking.
I did discover, however, that running telemetry from the plane through the Pixhawk 1 (Running AT 1.0) strips away the ADSB data generated by the plane’s onboard ADSB receiver. To reiterate, Mission Planner 1.3.51 is connected to Antenna Tracker 1.0 via USB, which is then connected to Plane 3.8.3 via SiK-based telemetry.

I would like to see more interest in the antenna tracker firmware, and on my part I will continue experimenting via multipoint telemetry, to see whether my ADSB data may be acquired whilst using the tracker.

The Pixhawk 2.1 offers potentially high performance for tracking, and getting the firmware working for the 2.1 would likely resolve the other issues experienced along similar veins.


(Laszlo Almazy) #9

Hi Evan! I’m trying to use a Pixhawk Mini as antenna tracker. It load the firmware AT 1.0 without issues but the led turns into solid red when booting, as I describe in my post.

I have checked everything imaginable and I think it’s a compatibility problem with the PX Mini hardware… Usually, I program processors, but in another environment. It’s very complicated to recompile the AT firmware for the Pixhawk Mini?

I would greatly appreciate a guide to start studying this topic. Thanks in advance.
Regards.


(Evan Williams) #10

Hi,

I had similar behavior when loading to the PixRacer. I got a little progress by recompiling for FMUv4, but was not able to get it to arm - you may run into the same dead-end. That said, compiling the antenna tracker code is the same as compiling for other vehicles, for which I used the following guide:
http://ardupilot.org/dev/docs/building-ardupilot-onwindows10.html

There’s a small repository hiccup when getting one of the dependencies for compiling, but I’ll edit with a link to another comment of mine in which I specify my alternative repo version.

I was only ever able to get the Pixhawk 1 to work fully, but I’m equally interested in seeing the Pixhawk Mini operate an Antenna Tracker.


(Laszlo Almazy) #11

Hi Evan, I would like to apologise for delay to reply. I had to put aside the antenna tracker project for a while. Now I’m resuming it.

Thank you very much for responding and for the guidance you have provided. I really appreciate this.

I will keep you updated on the progress I make.
Best regards


(Kendall Wells) #12

Do we have an Update on Antenna Tracker?


(Pierig Le Saux) #13

I have a pixhawk 2 flashed with antenna tracker v1.0.0. It’s sort of working (I can arm and I can servo_test): I need to spend some time tuning pids, because it’s shaking all over the place when I track a plane running SITL. Is there some kind of tutorial to help me with the tracker PID tuning?


(Evan Williams) #14

That’s odd, could you share your parameters?

I was never able to convince the Pixhawk 2.1 to arm with AT 1.0.0 . . . though I’ve settled with a Pixhawk 1 for now.


(Pierig Le Saux) #15

I using the button on the Here GPS to arm. I’ve only setup orientation of the pixhawk at this point, but here’s my file.

myantennatracker.param (4.3 KB)