This is a discussion to focus the feedback and support of ardupilot on the PixRacer board from AUAV.co.
We believe that all teething problems encountered are now resolved and so we recommend people load Copter-3.3.4 which is available for all multicopters through the Mission Planner's (and other GCS's) Install Firmware screens. Note that because regular Pixhawks are still on Copter-3.3.3 the install icon may show Copter-3.3.3 but rest assured, Copter-3.3.4 will be installed (you can check the version by looking at the very top of the Mission Planner).
Replies
The purpose of ESC calibration is twofold:
in ArduCopter the PWM range for the ESCs is set by RC3_MIN to RC3_MAX. Note that these must match the throttle channel range of your transmitter.
My guess would be that your RC3_MIN does not match the minimum PWM you used when you calibrated the ESCs.
Thank you Andrew. I did a couple of quick tests to make sure RC_PWM values are correct. Here are my PX4 values
1 1 RC3_MAX 1962.000000000000000000 9
1 1 RC3_MIN 1055.000000000000000000 9
1 1 RC3_REV 1.000000000000000000 9
1 1 RC3_TRIM 1508.000000000000000000 9
APM values
RC3_DZ,30
RC3_MAX,1964
RC3_MIN,1054
RC3_REV,1
RC3_TRIM,1508
One thing may make a difference is that my throttle channel is mapped to channel 2, probably because I run Deviation with Orange DSMX receiver.
RCMAP_PITCH,1
RCMAP_ROLL,3
RCMAP_THROTTLE,2
RCMAP_YAW,4
Another observation is when I connect battery without USB connected, my ESC would start a startap sequence with two beeps and at the same time PixRacer would play tones. After two beeps the safty switch light becomes solid (i have safety disabled) and seems like Pixracer interrupts startup of my ESCs and I got continues short beeps. Does it make any sense?
One thing may make a difference is that my throttle channel is mapped to channel 2, probably because I run Deviation with Orange DSMX receiver.
ahh, I should have asked about that. In that case the critical numbers are RC2_MIN, RC2_TRIM and RC2_MAX.
Assuming RC3_* is where PX4 stores the PWM range for the ESCs, then you should set:
Hi Andrew.
There is no difference in other RC calibration values between PX4 and APM versions. here is the channel 2.
RC2_DZ,30
RC2_MAX,1964
RC2_MIN,1054
RC2_REV,1
RC2_TRIM,1054
1 1 RC2_DZ 10.000000000000000000 9
1 1 RC2_MAX 1964.000000000000000000 9
1 1 RC2_MIN 1055.000000000000000000 9
1 1 RC2_REV 1.000000000000000000 9
1 1 RC2_TRIM 1055.000000000000000000 9
When I have time, I will try to see PWM output with the oscilloscope and may see what is the difference between two during bootup. My ESC are rebranded and I can only guess that it is Xrotor 40A and they supposed to use SiLabs chipset and who knows what firmware. Looks like they are sensitive to something that APM outputs.
One or both of your compasses have a high offset value say about 500 which is bad. Caused by metal or magnet near by.
Do you have the buzzer connected? If so can you try calibrating without the buzzer connected. The buzzer needs to be placed as far away from the flight controller as practical.
No external compass?
I think you'll find a lot of us only trust external compasses, and disable the internal ones.. So, I'd start there, to at least be at the same point the rest of us are..
Then, there was, and maybe still is and issue w/ the 2nd internal compass. It would never calibrate correctly, and had to be disabled. (the joys of alpha software testing).
So for me, I disabled compass #3 before calibrating. Maybe with no external, you'd need to disable compass #2. And be sure to load the latest version of APM Master before you begin.
And, for me, things that cause compass interference, are screws or servos or buzzers or antennas near the flight controller. Or externally, I cannot calibrate anywhere in my home, and almost nowhere in my yard.... In my home I have speakers in every room.. And even being a few meters away from the speakers, the compass still points the wrong way. In my yard I have some iron fencing and a porch w/ a metal roof, and that causes trouble for about 10 meters around those areas... The magnetometers are certainly a pain to get calibrated correctly in the wrong environment!
Hey there, today I've finally "maidened" my Y3 with pixracer!
Just to clear thing out- it is a proven flyer with pixhawk and AC3.3 with more than a year total flight experience.
Surely, I was flying some dev version (about a week old) and I did not expect anything particularly good, as I've flown AC3.4 dev on the same frame with a pixhawk (early March build), with the same exactly performance.
The things I like about pixRacer - wifi telemetry, size and the new connectors.
There is one thing I like about AC3.4 - PWM RSSI, and one thing I absolutely hate - horrible alt hold.
To be clear -horrible alt hold is only with AC3.4dev both on a pixracer and a pixhawk. so I am more than sure it is AC3.4 related at least could be some setting.
With AC3.3 on a pixhawk alt hold is very good, though suffers from the known fast forward issue.
Let me define "horrible": Without Throttle mid parameter adjustment AltHold is dangerous - copter will jump up and down pretty aggressively. For the record, adjustment was not particularly large- default 500, the best value I came up in the field- 400. For comparison, with a PX4v1 on a 250 copter AC3.3 I had to lower it down to 280 to get some sort of throttle control.
With throttle mid parameter adjusted to 400 behavior is a bit "sneaky" - copter will remain "glued" to the same spot in Loiter for 40-50 secs, than all of a sudden JUMP 3-4 meters and slowly descent back without any throttle input.
Also, the fore-mentioned fast forward issue is much more pronounced than with AC3.3 on a pixhawk - could be a result of default PIDs vs. hand tuned on AC3.3
I also have a 450size quad with pixhawk and AC3.4dev (some January build), it flew very good (before I crashed it hard), at least just as good as AC 3.3 (but with pwm rssi) and did NOT exhibit any jumpiness in althold/loiter. I am attaching my logs from today's pixhawk flights, I would really appreciate someone looking into them to figure out what is causing jumps in alt hold.
p.s. pixracer in a 3d printed case with foam over baro. vibrations are pretty normal, the same airframe with pixhawk and AC3.3 flies rather smooth, but jumps with pixhawk and AC3.4dev as well.
p.p.s AutoTune got horrible Roll result and not a nearly good Pitch, didn't get to try yaw as I've crashed when trying ATT ROLL PIDs.
https://drive.google.com/folderview?id=0ByNP743-FSrrNWMxcEE5M1ptUlk...
@Artem,
It looks like your flight uses a firmware from 5 days ago, which is actually quite old by the standards of this beta! Randy has fixed a number of important ALT_HOLD issues since then.
If updating doesn't help then you might like to try lowering THR_MIN. I found the biggest factor in improving ALT_HOLD performance was to lower THR_MIN. I had mine at 130 same as you do, and found when I lowered it to 50 things improved a lot.
Looking at your log it also seems you have THR_MID a bit high. Looks like you hover around 330 not 400.Note that lowering THR_MIN may change this (Randy, is that right?)
If lowering THR_MIN doesn't help then it may be worth trying a flight with EKF1 instead of EKF2. There are really two big sets of changes that are likely to be at the bottom of your flight problems. One is the AP_Motors work (which is related to THR_MIN above) and the other is the change to EKF2. You can't easily swap back to the old AP_Motors code, but you can set parameters to use EKF1 instead of EKF2. If you find it makes it much better then we may ask for a flight with both EKF1 and EKF2 enabled to get a log comparison.
To enable EKF1 set this:
then reboot. To do a dual-EKF flight set this:
Note that some of these options are only available in the latest firmware builds.
Cheers, Tridge
Thanks a lot for looking into this!
I've just loaded 1ae85be9 and adjusted thr_min and thr_mid to the values suggested.
I will have time to burn three lipos today so should be able to try all of the settings you suggest. will be back with the logs by tonight.
-
42
-
43
-
44
-
45
-
46
of 110 Next