Servers by jDrones

Nimbus Tricopter VTOL


(Greg Covey) #163

I went through the param files and didn’t find many differences. Some were expected and others may be differences between our Pixhawks (1 vs. 2). It is interesting that one uses EKF2 and the other EKF3 but it may be a Pixhawk thing because I believe they are equivalent.

I don’t use hardware handshake on my serial ports and I disable terrain following. I also only ever use a single compass. No safety button on my setup. Note that I had already incorporated some of your changes like Q_A_ANGLE_BOOST = 1 so they don’t show up here. I’m a bit puzzled why your Q_WVANE_GAIN is set to 0.

Greg’s Pixhawk 1 vs. Gordon’s V2 Pixhawk2

BRD_SAFETYENABLE,0 — BRD_SAFETYENABLE,1

BRD_SER1_RTSCTS,0 — BRD_SER1_RTSCTS,2
BRD_SER2_RTSCTS,0 — BRD_SER2_RTSCTS,2

BRD_TYPE,2 — BRD_TYPE,3

COMPASS_USE,1 — COMPASS_USE,1
COMPASS_USE2,0 — COMPASS_USE2,1
COMPASS_USE3,0 — COMPASS_USE3,1

EK2_ENABLE,0 — EK3_ENABLE,0
Slight changes here: EK2 vs. EK3?, Pixhawk vs. Pixhawk2?

GND_ABS_PRESS2,0 — GND_ABS_PRESS2,100685.2

nothing here — OVERRIDE_CHAN,0
nothing here — OVERRIDE_SAFETY,1

Q_A_RAT_PIT_D,0.004 — Q_A_RAT_PIT_D,0.015

Q_A_RAT_RLL_I,0.4 — Q_A_RAT_RLL_I,0.25

Q_A_RAT_RLL_P,0.4 — Q_A_RAT_RLL_P,0.25

Q_ANGLE_MAX,4000 — Q_ANGLE_MAX,3000

Q_TILT_MAX,65 — Q_TILT_MAX,45

Q_TRANSITION_MS,6000 — Q_TRANSITION_MS,3000
Q_VFWD_ALT,2 — Q_VFWD_ALT,0

Q_WP_RFND_USE,0 — Q_WP_RFND_USE,1

Q_WVANE_GAIN,0.2 — Q_WVANE_GAIN,0

TERRAIN_ENABLE,0 — TERRAIN_ENABLE,1

WP_LOITER_RAD,90 — WP_LOITER_RAD,60
WP_RADIUS,40 — WP_RADIUS,30


(Gordon Ottershaw) #164

Thanks Greg! I would have thought that the weathervane parameter would make a big difference during windy landings or takeoffs.


(Gordon Ottershaw) #165

Thanks for pg20 info. It would be better if the unit could detect a battery coming unplugged and switch to the remaining one.


(Graham D) #166

Well, I can confirm that the unit will fly if the battery is plugged to either A or B plugs, haven’t tested yet what happens during transition with only one battery connected but I suspect nothing will happen, the unit will constantly take power from the battery that is connected.


(Laurentiu Lazar) #167

Hi Greg,
Are you using Kris Shyi’s custom firmware or the official Ardupilot one ?


(Greg Covey) #168

I’m using the official Arduplane stable release v3.9.3.

Cheers!


(Wasim Akram) #169

Hi Greg, I was wondering how you went about tuning your Nimbus tiltrotor quad?


(Greg Covey) #170

Hi Wasim,

So far, my only tuning has been for yaw control in the hover modes. I haven’t transitioned to forward flight yet as I ran into crappy weather this fall and now it is winter. Since I am using the original Foxtech V2 Nimbus power system and my weight is similar, simply copying the forward flight PIDs should be fine. The forward flight PIDs seem to work for both V1 and V2 power systems so only the hover mode PIDs have changed.

My initial BARO and GPS mode testing was done in my backyard. I normally don’t do this due to the surrounding hazzards but I was excited and risked it. Both QHOVER and QLOITER modes worked great!

For the yaw PID tuning, I started with Graham’s PIDs and then made some changes after brief hover tests in my backyard. By leaving Mission Planner on the Full Parameter List screen, you can land, write a new PID value, and take-off again. No need to disarm or re-boot the Pixhawk so you can quickly see how the PID change has affected performance. My future plan is to do this using more aggressive maneuvers in an open area. This is the time of year (at least in NY) where patience is a virtue! :slight_smile:

Cheers!


(Laurentiu Lazar) #171

Thank you Greg for your response.


(Gordon Ottershaw) #172

Hi Greg,

Re: weathervane

Hesh (at Foxtech) almost always maneuvered the Nimbus with its tail facing him during the landing phase. The weathervane would interfere if it was enabled. Suspect this was just a personal preference.


(Greg Covey) #173

Hi Gordon,

That makes sense for manual flying. I also do my own positioning when flying manually and the weathervane feature can fight you. For autonomous flying, the weathervane feature can be a comfort when the winds are swirling or changing directions during take-off and landing. However, this has its limitations and can be problematic if your mission plan doesn’t allow for it. Been there, done that. A hovering plane in the wind can get into trouble fast, especially in QLOITER mode, if it is not pointed in the right direction. Autonomous flights can also take-off in the opposite direction from the expected mission plan.

It’s snowing again today…must be patient this time of year. :neutral_face:


(Gordon Ottershaw) #174

Hi Greg,

Thanks for your comments. I greatly appreciate your no-nonsense presentation.

Based on your experiences, it is even more apparent that sufficient skill with the system to change key parameters on-the-fly is essential. Wind and weather rarely cooperate with flight plans.

“A hovering plane in the wind can get into trouble fast…” That is so very true and I think it partially explains why I like the 2 battery setup of the Nimbus. It enforces discipline in getting up and getting on with the fixed-wing mission. There just isn’t much time to muck about in Q modes.


(Atico) #175

Hi Greg,

What FW do you used in the HOVER test? ChibiOS or NuttX?

Cheers


(Greg Covey) #176

Hi Atico,

I don’t know for sure. I am using Plane stable release v3.9.3 (7c7c5038). It does not say ChibiOS on the Mission Planner connect information at the top. Some of my other models have stated “ChibiOS” along with the version information.

Cheers!


(Atico) #177

what do you see in the messages tab?


(Greg Covey) #178

Very good, thanks!

ChibiOS: ff603d11
ArduPlane V3.9.3 (7c7c5038)


(anibal manuel marques soutinho) #179

Hi Greg .

I have a problem, I did all the conguration in manual mode all works well but, when I change to Qestabilize the motors stops.

Can you help?

I work with your parameters.

Thank you again.

Best Regards


(Greg Covey) #180

Hi,

It could be your ESC calibration or it could be one of the Q_M_SPIN parameters like Q_M_SPIN_ARM. I typically have mine set to 0 but some users prefer to see the props spin when armed for safety.

First, remove the props and then perform an esc calibration using one of the techniques below. Also, make sure that you have already done a radio calibration in Mission Planner.

Good luck!

QuadPlane ESC Calibration


(Bobby Moser) #181

Greg,

What parameter do you use to adjust the tilt servos. Mine is good or will compensate in hover mode but my left motor is a little low in forward flight.

Thanks,
Bob


(Greg Covey) #182

Hi Bob,

You need to tweak those servoX settings to fine tune the position. My recollection is that servo9 and servo10 are the tilt servos plugged into Aux1 and Aux2 of the Pixhawk. You can see where mine ended up below. I can’t remember if Min or Max was for forward flight so if you write a change and it doesn’t move, put it back and change the other setting.

SERVO9_MAX,1855
SERVO9_MIN,1065
SERVO10_MAX,1860
SERVO10_MIN,1110