I’ve recently done the change from NAZA-M V2 to Pixhawk and I’m having some troubles with it…
Some Background
The Quad arms fine, gets smoothly off the ground, is super stable and very responsive while in [color=#FF0000]Position Hold Mode[/color].
Especially Throttle is working like I’m expecting it: Slowly raising the Throttle has the Motors slowly spinning faster. Slowly lowering the Throttle has the Motors slowly spinning slower.
This was tested in my bedroom with the Quad tied to the bed preventing it from crashing into anything else but the soft mattress.
Up up and away…
My house must be a bunker which is why I’m unable to test any mode relying on GPS so my test today was outdoors in the wilderness without a soul around ( and again with the Quad on a leash ).
With 18 GPS in the Skies I felt confident and started in [color=#FF0000]Loiter Mode [/color]as it’s the first mode. I slowly raised the throttle and while it started nice the thing suddenly started climbing like mad into the sky until the string stopped it with me trying to get bring it back under control by lowering the throttle which wouldn’t do a thing at all - It kept angrily screaming in the sky pulling at the string.
I managed to get it down in one piece with the help of the string attached to it but the sight of how that worked out was not a work of art, not at all. I’d actually prefer to forget all about it as it basically whacked into the ground 3-4 times before I could reel it in close enough for me to keep it on the ground and reach for the Safety Switch to turn off the motors.
Back at home…
and without the Props attached and GPS Check disabled I checked the Flight Modes and observed that both Loiter and Altitude Hold would drastically increase speed ( compared to Position Hold Mode ) when Throttle is applied but would not lower it when Throttle is taken away unless the Stick is just barely above the arming point and even then only after quite a delay.
What I’d basically like to know is:
Why was Loiter shooting up up and into the sky.
Why do Loiter and Altitude Hold Throttle not react when I’m trying to lower them.
Any suggestions what to look out for?
Thanks in Advance! =)
Stabilise doesn’t rely on gps, compass or barometer to stay in the air unlike loiter, alt hold, pos hold etc etc.
If you have a glitch in the sensors you need a flight mode that is not going to be effected by that glitch, hence why stabilise is the first mode you go to.
It also proves the uav is flyable. If it’s unstable in stabilise you have a fundamental problem that needs sorting before using an enhanced mode.
Logs are really straight forward, connect to mission planner as normal. On the bottom left box you will see some tabs including “dataflash logs”. Select that and then click on “download dataflash log via mavlink”, on the new window that pops up download all logs. They will then be stored on your hard drive in the mission planner folder under logs.
The same tab has “review a log” which lets you open up the downloaded log files so you can see what was going on.
“auto analysis” will do a check on a selected log and give you an idea of how well that flight went.
I attached a ZIP File with Log Files that matched the Date and Time of the Event, as I’m not entirely sure which one is needed I added all of them.
As suggested I took a peek into the LogAnalyzer and this is what I got:
[code]Log File C:/Users/XXX/AppData/Local/Temp/tmpFFF3.tmp.log
Size (kb) 2735.7333984375
No of lines 37309
Duration 0:04:12
Vehicletype ArduCopter
Firmware Version V3.2.1
Firmware Hash 36b405fb
Hardware Type
Free Mem 0
Skipped Lines 0
Test: Autotune = UNKNOWN - No ATUN log data
Test: Balance/Twist = GOOD -
Test: Brownout = GOOD -
Test: Compass = GOOD - mag_field interference within limits (11.24%)Max mag field length (555.70) > recommended (550.00)
Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = WARN - FENCE
Test: GPS = GOOD -
Test: IMU Mismatch = FAIL - Check vibration or accelerometer calibration. (Mismatch: 7.91, WARN: 0.75, FAIL: 1.50)
Test: Parameters = GOOD -
Test: PM = FAIL - 17 slow loop lines found, max 12.95% on line 926
Test: Pitch/Roll = GOOD -
Test: Thrust = FAIL - Avg climb rate -572.37 cm/s for throttle avg 960
Test: VCC = GOOD - [/code]
From what I’m understanding it appears that I’m looking at a Vibration issue?
Should I attempt a 30sec Stabilize test and post another Log where the Quad wasn’t desperately trying to climb while being tied to a tensioned string?
Z axis vibrations do see to be spiking quite high which could account for the altitude gain and why the stick input was ignored. 99% certain switching to stabilise would have stopped it trying to go up.
Might be worth looking at the Pixhawk mounting see if it can be improved at all.
[quote=“MarkM”]Z axis vibrations do see to be spiking quite high which could account for the altitude gain and why the stick input was ignored. 99% certain switching to stabilise would have stopped it trying to go up.
Might be worth looking at the Pixhawk mounting see if it can be improved at all.[/quote] Thanks for taking the time to help me solve this issue!
Part of the Pixhawk System also included the change from a 400kV/15x5 to a 1000kV/9.5x5 powertrain and being a tad spoiled from the superior quality of the 15x5 Props that never required balancing I grossly neglected to balance the 9.5x5 ones which I now did.
I’ve added a new Logfile where the Quad was just hovering in Stabilize Mode and while it was still tethered, because you know… t(h)rust issues, I ensured the Cable was not under tension.
[color=#FF0000]Click Image for larger Picture[/color]
It appears that either the balancing of the Props or the Cable not being under tension solved most of the alarming issues concerning Vibration?
LogAnalyzer[color=#FF0000]:[/color]
Log File C:/Users/XXX/AppData/Local/Temp/tmp4C2E.tmp.log
Size (kb) 1096.60546875
No of lines 16035
Duration 0:00:00
Vehicletype ArduCopter
Firmware Version V3.2.1
Firmware Hash 36b405fb
Hardware Type
Free Mem 0
Skipped Lines 0
Test: Autotune = UNKNOWN - No ATUN log data
Test: Balance/Twist = GOOD -
Test: Brownout = GOOD -
[color=#FF0000]Test: Compass = WARN - Moderate change in mag_field (26.93%)Max mag field length (590.07) > recommended (550.00)[/color]
Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = GOOD -
Test: GPS = FAIL - Min satellites: 0, Max HDop: 99.99
[color=#FF0000]Test: IMU Mismatch = WARN - Check vibration or accelerometer calibration. (Mismatch: 0.76, WARN: 0.75, FAIL: 1.50)[/color]
Test: Parameters = GOOD -
Test: PM = GOOD -
Test: Pitch/Roll = GOOD -
Test: Thrust = GOOD -
Test: VCC = GOOD -
[color=#FF0000]Test: Compass = WARN - Moderate change in mag_field (26.93%)Max mag field length (590.07) > recommended (550.00)[/color]
Doing a quick search it appears the first one might be a problem related to the FC being to close too the ESC’s?
While the FC is located inside the Quad ( TBS Discovery PRO ) the ESC’s are not with a ~10cm distance between them.
I also use an external GPS on a Stand located at the rear ~14cm away from the nearest ESC.
Throttle at 50-60%
[color=#FF0000]Test: IMU Mismatch = WARN - Check vibration or accelerometer calibration. (Mismatch: 0.76, WARN: 0.75, FAIL: 1.50)[/color]
Not exactly sure what the issue is with this one as the Graph and the Log appear to be within the limits of what can be considered flyable?
Managed to find a quiet spot that would also provide some soft landing in a field in case of something literally going south.
While I’ve only tested Stabilize and Altitude Hold ( I had the other Modes temporarily removed… ) it looked very promising in both modes.
With the *.bin files being larger than 2MB I’m unable to share them using the board so I’m using OneDrive: Logfile 1[color=#FF0000]:[/color] OneDrive Link to 1st Logfile AutoAnalyzer Errors[color=#FF0000]:[/color]
Test: IMU Mismatch = FAIL - Check vibration or accelerometer calibration. (Mismatch: 2.74, WARN: 0.75, FAIL: 1.50)
Test: PM = FAIL - 36 slow loop lines found, max 10.08% on line 2104
Logfile 1[color=#FF0000]:[/color] OneDrive Link to 2nd Logfile AutoAnalyzer Errors[color=#FF0000]:[/color]
Test: IMU Mismatch = FAIL - Check vibration or accelerometer calibration. (Mismatch: 3.21, WARN: 0.75, FAIL: 1.50)
Test: PM = FAIL - 38 slow loop lines found, max 8.93% on line 87508
IMU Mismatch appears to be solvable by doing ( another ) Accelerometer Calibration but what’s the Slow Loop Line thing?
If I could get a general opinion on the Logs and if there’s something to look out for I’d be grateful =)
