M4 Ladybrid Position Hold and PH controller Questions

Info and discussion about the autonomous Micro FC, suitable for nanos, micros and mini multirotors

Re: M4 Ladybrid Position Hold and PH controller Questions

Postby OlliW » Tue Jan 06, 2015 1:26 pm

a update on a test flight this morning

I let the bird sit outside for >30min to thermalize. I double-checked my dec and inc's and did a mag calibration at exactly the RTH position. The mag magnitude was 2.0x. The acc mag was a bit higher than usual at 10.1, but that's maybe due to the temp (indoors it's 9.8x). Also, to ensure that I really do this thro-stick-center thing right I've set the deadband to a redicoulous 150us (yes, my RC does exactly 700 at stick center). The baro was covered as shown in the photos in the above. I could not find any negative effect of this on e.g. the mag calibration, RF signals etc. GPS 3D fix was within few minutes (actually missed it since it was so fast), and Hacc 0.5 m was reached within, I'd say 5-8 min. Vacc was 0.7-0.8m. Weather was nice, nearly blue sky, essentially no wind.

The BT connection seems to work very well, and I could see the telemetry data (had 10Hz) during the FULL flight in QGC's Values Grid (actually my son helped out here and had a constant eye on the data). For some reason one gets Hacc but not also Vacc. I also started a log file, however, it seems not to have recorded the full flight, just two bunches, maybe that's because I switched from Values Chart to Values Grid to see the data during flight (if this would be so I would consider this a serious bug of QGC). Also, it's not a .log file and then I reopen it with QGC it crashes. Not sure what stupid I did here. I need to retry. Anyway, the important thing is:

I/we could see the Hacc value during all flight, and it NEVER went above 0.55m, it mostly was below 0.50m. Yet, the altitude hold showed the same characteristics as before. After switching to position hold, the copter was at about 2-2.5 m height. It stayed a bit, but then declined to nearly ground, went slowly up again, etc. In one case I had to pull the sticks to prevent it hiting ground. I did a couple of RTH's, and it was always the same. I would say that the altitude hold accuracy was such that the copter stayed in a bin of 2.5m. I flew away a couple of times, pretty much as in teh video before. Here the copter had essentially perfect view on the sky. Hacc stayed nearly the same (maybe it dropped a 0.1m on average). One could see here the same thing, that it wanders up and down. It's difficult to measure, we used a marker on a pillar to "triangulate", but it also had be about 2m up and down.

This time the RTH was however much better in as much that the xy position was relatively well reached. It didn't strayed away and wouldn't get in danger to hit anything. This suggests that the earlier problems in this regards could have been simply due to wind. Since the copter reacts quite "aggresively" on sticks in PH and on the RTH command I assumed that the nav controller is tuned quite "strongly" and that it would hence make the copter to stay close to the desired path, but I might not yet understand the AQ nav controller's behavior well enough, and it seems to be not as "strong" as I thought and let the copter stray away few meters.

As a conclusion: With a Hacc of ca 0.5m this is simply what one gets.

As said earlier, I don't have the experience to compare that to anything and decide if that's impressive or not, so no comment or even judgment from my side, just a statement. Going through all the above was very helpfull to me, since I have not only a better understanding of what to expect but, more importantly, a feeling that I've got my setup now about right. BIG THANKS to all of you! :)

(don't worry, I will have some fruther questions as regards PH :lol: )

EDIT:
I did now another test, where I pulled out the GPS, since I wanted to see how the baro-only altitude hold would work outside - and - THAT's what I would have expected !!!! On the very first time it very slowly moved a bit down, I brought it back in position, and since then it just stayed there in altitude, in maybe a bin of several 10cm (I of course had to use the pitch/roll stick to keep it in xy position). THAT's how it should be!

As a conclusions: Sad that there is no mode with GPS 2D + baro altitude PH and missions ...

The current fusion of baro and GPS altitude data is not yet perfect, at least not for a beast like the M4 ladybird with dipole antenna.
OlliW
 
Posts: 60
Joined: Sat Nov 01, 2014 3:08 pm

Re: M4 Ladybrid Position Hold and PH controller Questions

Postby zhlearnaq » Thu Jan 08, 2015 3:07 am

Can anyone show me the meaning of GPS_LATENCY/UKF_POS_DELAY/UKF_VEL_DELAY? And how to choose the value of them?
zhlearnaq
 
Posts: 43
Joined: Wed May 14, 2014 10:47 am

Re: M4 Ladybrid Position Hold and PH controller Questions

Postby zhlearnaq » Wed Sep 16, 2015 3:55 am

Hi chschmid,
Sorry to disturb you, I have noticed that you said that the mags are only used at liftoff time and the values will be replayced/adjusted by GPS data afeter a short flight time, how does this happen? Do you mean that the heading value can be adjusted by GPS data? why this value can be influeced by GPS data, i am a little confulsed, could you give me some sugguestion, thanks a lot.
Alice
zhlearnaq
 
Posts: 43
Joined: Wed May 14, 2014 10:47 am

Previous

Return to AutoQuad M4 Micro Controller

Who is online

Users browsing this forum: No registered users and 13 guests

cron