Despite the very few examples of APM antenna trackers being used for FPV in the field I am going to make one with the Servo City PT785 (3.8:1). Has anyone else used this?
I will try to document and report back with how I am doing. If anyone else has used this admittedly rather pricey unit from Servo City please let me know how you did. After using a MFD with sloppy chattery servos I just decided to get a better unit with more torque. It looks fabulous. Let’s hope it performs well.
Also, APM AAT does not support continuous 360 yet does it? That would seem to be a good idea.
It would be nice to see more action around Antenna tracker and OSD areas of Ardupilot.
Chad. I have this unit working as a Pitlab AAT right now. I plan to fire up an Ardupilot AAT with the same thing and make it dual use. Not done it yet. I hope I can document some success or lack thereof soon. Sorry I cannot be of more help now.
I have the same problem that Pixhawk is not driving any servos in testing. Pls see my posting under " Testing button not working in Antenna Tracker Extended tuning page".
But when I use APM 2.6 as AT controller, it drives the servos in testing. Weird.
Posting on this thread because I’m building an AT with the PT785-S pan/tilt set up as well. I’m using a pixracer as the control board. After having a few issues with the set up process and getting through that, now I’m having difficulty tuning the system with the multi-rotation servos on the PT785 (HS-785HB servos). The system will yaw and pitch no problem with the test button in the extended-tuning page, but I cannot send PWM signals <1100 or >1900 so there’s no way to test the multi-rotation. It also doesn’t rotate a full 360º, just about 330º. When the tracker is in Auto mode it will yaw fine until it reaches the end of its first rotation, then it just gets stuck there shaking back and forth, it never reverses back around the other way at all. When this happens it pitches over past 90º (despite the limit I have set in the tuning page) and points at the drone upside down.
I think it’s an issue with the way the servos respond to the PWM output of the pixracer but I don’t know enough about any of this to work out how to fix it. I’ve tried tuning the PID values and this reduces the amount of shake but doesn’t send the yaw back in the other direction, nor continue around past about 330º.
I’m hoping someone out there has been successful with the PT785-S and can help me out, especially as it’s one of the systems suggested on the antenna tracker page.
Ok so I went in and used the Servo Output page to manually enter the min, max and trim values. I got min and max from the servo spec sheet (600 - 2400) then I used the extended tuning ‘test’ button to find the trim value. Doing this now allows the yaw to turn almost 2 full turns using the test button, and it also does this in Auto mode. Doing the same with the pitch servo stopped it from over pitching beyond 90º. This might be obvious to experienced users, but it wasn’t something covered in the AntennaTracker Documentation page.
But… there is still a very strange yaw oscillation that happens when the tracker starts tracking a moving target. It begins moving OK but then will get into a position where it is facing 90º to the target and it just sits there yawing back and forth. I can reduce the amount of yaw with PID tuning, but still can’t seem to get it to move properly to point at the target.
Ok, following up on previous posts. I got the tracker to function correctly, not quite tuned 100% but that’s easy enough from here. The things I did (additional to what I already described):
the yaw servo had to be set as ‘continuous’ even though it is not a continuous rotation servo. I have no idea why that worked, but it did.
the pitch servo had to be set as ‘position’, which is the correct setting for this type of servo.
Once I have the tune sorted out I will post my param file for anyone else using this pan/tilt mechanism. It’s been a frustrating journey, but I now have a working tracker.
Have you had any success in tuning your PIDs? I have been working on a tracker using the PT785-s system and a Pixhawk 1 for a few months but keep shelving the project after getting frustrated. I am also only able to get about 330º of rotation on the yaw servo but am unable to set my PWM max param >2200. For minimum it will allow as low as 500. In testing the servos however, I am unable to test outside of the 900 and 2200 values so I suspect this is why I am not able to get as low as 600. Have you upgraded your tracker to use the latest firmware?
Hi Derek - yes I got mine working great after a long trial and error process. I’ll have to fire it up and check what firmware I am using but I’m fairly sure it’s on the latest release. I did post my param file in one of the other threads, that may be useful to you?
Thanks for the quick reply Adam! I found the parameter file and will take a look. I seem to be having more success today with the Pixhawk 1 board than I was with the more expensive Pixhawk 4, but I’ll chock that up to a clear head after taking a break from the project. Do you have any photos of your tracker? I’m still trying to land on a mounting solution for mine.
Still fighting both pixhawk boards with different issues on each.
On the 3DR Pixhawk 1, I wasn’t able to get both compass modules working 100%, and the GPS quality was not great. I can get it to track a SITL target with some luck, but always a few degrees off. It seems like it would track fairly well if the target was within about 30 degrees of north, but after that it was consistently off target on Yaw. Pitch seems OK from what I can tell. If I opened up the tuning page it would look as if the source and target graphs were aligned, however this was not the case in the real world. i shelved it for now and moved on to the pixhawk 4.
On the Pixhawk 4, The compass and GPS are much better, with accurate headings and less drift. I got the servos tuned as best as I could for both minimum and maximum values and went through the full parameter list to review differences. PID settings required between the two are different. The Pitch seems much more sensitive to P gains and with anything above 0.01 I see oscillations. At this point I’m at a loss for what else needs change. I am still not quite sure i understand how the PID values are working.
For now everything is mounted and my vehicle and antenna tracker and quad-rotor are communicating VIA my IP mesh network, I just need to get the tracker working a bit better.
I’m attaching my config files for both if someone with a working tracker has the time to take a look and point out any glaring issues.Pixhawk 4 Antenna Tracker 22072020.param (8.5 KB)