Hi,
After I did a AutoTune my drone became too responsive.
Before AutoTune flying was way more pleasant.
To rollback these settings do I have to change the PIDs to it's defauld settings?
This is my setup:
Pixhawk with Hexa Arducopter 3.2-rc14
Mission planner 1.3.15.2 (Beta)
Below the screenschots of the PIDs before and after AutoTune:
i'm not familiar with these settings and not certain what to change.
any advice is welcome!
Replies
Leanord
I have included two log files i could probably dig out the actual autotune logfile if it would be better.
Thanks
David Ardis
2014-12-04 10-55-38.bin
2014-12-04 15-30-55.bin
Hi David,
Those logs look fine. I will need to see an autotune log to get any idea what is happening.
Leonard
I found the actual autotune log
Thanks
David Ardis
2014-12-04 14-10-35.bin
Hi David,
I thought the autotune D went to 0.002, not 0.02. So it has maxed out. Sorry, brain fart.
Your autotune looks good, if anything it would benefit from higher limits. I do get high values for the X8. I suspect that it is because of the coax props and the greater inertia.
So how does it fly?
Leonard,
Are the ATC_ACCEL_RP and Yaw settings universal? Would these impact behavior during a Auto mission in MP or DP?
I am aware of the WP parameters but wondered if these could be tuned softer also for certain flights to smooth transitions in flight path for AP work?
Hi Greg,
Currently they only impact the manual flight modes but we may make the auto modes use them to smooth things out. We couldn't do that on the APM because it wasn't fast enough but the pixhawk may be ok.
That would be welcome addition. Without slowing things down substantially many waypoint transitions result in some abrupt motions.
Thanks Leonard.
Rate P is already at lowest by default (.055). Whatever the defaults are for Y6B, that's what it is. I now have ch. 6 used to adjust Rate P with min at .08 and max at .20; haven't done any experimenting yet.
My description of the copter steering in Stab mode like a Greyhound bus is about right; lumbers around, very unresponsive and obnoxiously unstable in windy conditions. The RC_FEEL parameter was at .055 (min setting) by default(?). Now with MP 1.3.16 it is at max by default apparently, so I'd assume it will be much more sensitive to stick inputs. There is no Expo used.
The wild swings occur in Loiter when it gets windier, and is much worse with the larger props (14 vs 13). Unless it's up high, I don't use Stab because it is so unstable. At least in Loiter it eventually corrects itself. The parameters are Y6 RTF, which is a much different copter than this Spy, at least that's what I presume am experiencing.
The CG is set correctly, at least by my reckoning. I use gram scales under each motor and adjust the battery so there is equal weight under each. That's the method I used for the Y6B and it worked well.
To be honest I think there are sync issues because it still does not like to lift straight up in Stab and even though I've done several ESC calibrations, the motors don't spin up the same occasionally especially at low throttle. It's hard to tell in the air. When doing CompassMot, they seemed to turn ok at 3/4 throttle, so it may be isolated to occurring at low throttle. Once high enough I switch to Loiter to stabilize it; almost too chicken to switch back.
At this point I think switching out the ESC's may be the wise thing to do before going further. Based on my understanding, these motors need higher timing, and I question if the 3DR ESC's provide that.
Still learning.
EDIT: vibrations are very good, Mag Interference is 1.7% and GPS quality good (10+ sat and 1.5 hdop the last time out).
HI DG,
Always start with Rate D. I if you start with Stab_P you just have to come back to it again after adjusting Rate D anyway. Just follow through the procedure I described. This will very quickly and safely let you know where a good starting point tune is.
Loiter, RC_Feel, are not important until you get that basic tune going and hopefully can do a good autotune.
Your method of getting your CG right should work well.
It is always a bit of a give away when the motors don't start at the same time. I would double check your esc calibration.
The 3DR ESC's are running simon K. So they will be the same as any Simon K esc. However, I don't recommend Simon K on anything bigger than 8" props. If you are going to switch our your esc, I would go with a standard ESC. I have had good experiences with Plush and MultiStar but everybody will tell you different things there.
If you go through the process I described above and you still have a very sloppy copter I would try dropping the prop size to 12". The reason I say this is the only time I have trouble tuning a solid framed copter is when the props are too large for the Motor/ESC/Battery combination. This can be fixed by increasing the cell count if the ESC and Motor can handle it. If your copter all of a sudden tunes up great on 10" or 12" props then this may be your problem.
Leonard,
You say, "Set Stab P to 4.5 (should have already been there."; as if this is a universal constant?
If so why would multiple autotunes on my dead cat keep raising this from the default 4.5? One time It went to just over 9 and the most recent to over 13.
My frame started out flying fine and ended flying fine after AT, it was just way too twitchy for AP. It tipped in fast and hard and stopped on a dime. Adjusting the parameters suggested by Rob has really helped.
Thanks
-
1
-
2
-
3
-
4
-
5
of 7 Next