I have spent most of today setting up my gaui 330 with a Mega and IMU. I got everything on the frame and connected, esc's calibrated and motors spinning the correct way.
I hooked up to the configurator after having to spend far to much time trying to get the configurator and software install from the ardupirates wiki working.... anyway after finally getting it working I init the eeprom and calibrated my TX. Checking the sensors seemed to show everything was ok.
Next I armed the motors and ran the pre-flight checks and all the motors responded correctly. Next I had an issue in working out which was the DIP switch for stable/acro different pages and user comments on the ardupirates wiki conflict each other some say switch 1 some say switch 3 :( anyway after sussing that out I proceeded to PID tuning.
To start with it was difficult to follow the wiki because I have no reference as to what very stiff as far as correction goes or what is an oscillation when shaking it is difficult to sense an oscillation when the motors are spinning up and down to compensate the shake motion. Anyway I guessed the same kind of feedback that the gaui controller gave in acro mode would probably be what I was looking for. The defaut PID values seemed to work best so I unplugged and tested. To start with there was some drift forward and right and also a clockwise yaw so I landed and trimmed and then tried again. After a while I seemed the pitch roll sorted but I cant get rid of the yaw one way or the other. Also I was finding that after about 10 seconds of reasonable stable flight I would start getting some wild yaw movements and in trying to manually correct that would kick off a uncommanded pitch or roll movement. Any ideas? It also seems the rudder stick or throttle stick inputs are some how acting on each other in other words throttle up gives me clockwise yaw, lower throttle counter clock wise, which PID do I need to modify to correct this?
I also tried to play with the D value the wiki suggested starting with a -5 increment when I tried this the motors would hardly spin up even at max throttle so I assume the default D value of zero is the lowest anyway, wiki could do with making this clear seeing as an initial setup has a zero D value. So I had it flying but not 100% and I decided just to try some stable PID settings.
To start with the Wiki says to use the acro settings to begin then in the very next sentence it says to use 4.0 as a start, I tried my acro PID for stable and it just wanted to flip so thats no a good start so I used the defaults and it flew but I had no real idea of what the quad was doing or if it was even trying to stab in any way seemed my poop acro settings hands off were better that stable.
Does anyone have any tips to get started here or even some sample PIDS for a 330x so I can compare to see if I am even going the right way with this.
Thanks,
Mark
Replies
Guys a few updated on my progress,
Firstly the motor 1 spinning up on mega boot up I have noticed only occurs when the usb is connected, I don't see why it should do this.
I checked the heading value in the artificial horizon and it still does not seem to report a proper compass heading and does not show 0 at north, instead it only seems to report a value between 20ish - 170ish. Am I supposed to see a proper compass heading on this screen? however in stable mode the mag seems to correct the yaw correctly, however I have no Idea yet if it will work in gps hold mode to make the proper heading adjustment to correct position.
I had another successful flight last night but I still get more drift than I would think I should get in stable mode and its still a little jerky and not so smooth I will try and post my Acro and Stable PIDs later and maybe you can let me know if the look like they are in sensible ranges, I need to do more tuning here.
The issue I had were stabilisation not being applied after some reboots does not seem to have returned, maybe I wasn't letting the quad fully boot before flying? or should the quad not allow the motors to arm before it is fully up and running?
I did try altitude hold again and it seems to hold altitude ok, it is still a bit jerky and pulses heavily when it is correcting altitude, can this be smoothed out with the sonar PID's, are there any tips here on how I should adjust?
Nice write up John and helpful especially the part about getting rid of the bullet connectors. After flipping my quad a few
times I carefully desoldered them and after close inspection I found that 6 of the male ends were faulty [not the greatest
of quality] Its amazing that I managed to get into the air a few times without flipping. Since I have now soldered all the
wires I have had no problems with the motors. Also they are stationary after the initial arming and only start spinning
when I increase the power. I have crashed several times now and am pleased to say I've only managed to break 1
propeller and the top circular piece of plastic. I love it and can't wait for the return to home feature to be implemented.
Keep at it Mark when you do get to come right [I also battled and was extremely frustrated in the beginning] you'll
be happy you made the effort. Regards.
Hi Mark:
Are you using a magnetometer? If so, it has to be enabled in the code, also in the configurator, and then also through the CLI. If the mag isn't working right, it can cause yaw problems.
Also, make sure to fully calibrate your transmitter through the configurator, and have your radio trims set to zero during this process. You can then trim them manually once you are flying. You can make sure you have done this correctly by making sure your neutral stick points are in the green zones on the transmitter calibration window in the configurator. Only after making sure of this, should you then add manual trim.
Also, when you do pitch or roll, you shouldn't get any unnecessary yaw. The PID's will never add yaw to a pitch or roll command. However, the magnetometer is affected by pitch and roll. The magnetometer offset calibration through CLI mode makes sure that the mag reading stays true even as your mag is pitched and rolled.
Finally, make sure you don't have any unintentional mixing enabled on your transmitter. Some fancier radios in helicopter mode or other modes can mix different channels. Probably not happening, but doesn't hurt to check.
By the way, a little drift forward and back is normal. You can get it to hover reasonable well, but the wind (ground effects if you are flying indoors) contributes to some movement here and there. However, that random yaw movement makes me suspicious of your magnetometer.
Hope that helps some.
John