Tarot Gimbal documentation page, Calibration, APM behavior

Archive
1

This is in regards to the documentation at copter.ardupilot.com/wiki/common-tarot-gimbal/ and perhaps the behavior of APM itself.

I just finished a month-long quest to figure out why I could not use the RC6/RC7 channels to control the tilt of my Tarot-T2D gimbal.

At some point in the history of my using APM, I had clicked on “point camera here” in the interface, and unwittingly had set MNT_MODE = 2 (MavLink_targeting).

Today after months of tinkering, debugging, breaking and fixing the control board, rewiring, etc, etc, I discovered that if I set MNT_MODE = 3 (RC_Targeting) everything works as intended.

THIS SHOULD BE DOCUMENTED ON THE WIKI!!!

Also, the process of calibrating the behavior of the tilt servo of the the Tarot-T2D should also be documented.

Here are the group of settings that got me reasonable performance with my Spektrum DX7s on AUX2:

MNT_MODE = 3 (RC_Targeting = Eureka!)
MNT_RC_IN_TILT = 7 (RC7)
RC7_Function = 0 Disabled
RC7_Max = 1851 (calibrated max)
RC7_Min = 1169 (calibrated min)
RC7_Trim = 1510 (where the mid-point “beep” is)
RC9_Function = 7 (mount_tilt)
RC9_Max = 1520 (trial-and-error max)
RC9_Min = 1070 (trial-and-error min)
RC9_Trim = 1295 (average of min/max)
MNT_ANGMAX_TIL = 0
MNT_ANGMIN_TIL = -90

Also, I had to use the Tarot windows configuration utility, and set the following:
Angle Limit Max 15
Angle Limit Min -100
Total Gain 100
RC Sensitivity 100

With these settings I can verify that the Tarot is getting a range of inputs from 0 to -90° tilt, (but I needed to open up the Tarot max/min to allow it to get there for some reason).

I’d appreciate if someone could verify “this is how its done” or point me to a better set of docs. I haven’t played with the MAVLink targeting enough to know if that’s working with these parameters yet.

Also - there might be a bug in APM Planner in the RC#_REV settings. The documentation says that 1 is for “normal” and -1 is for “inverse”, but the parameter seems to be 0 to 255. This might explain why clicking the “invert” button in APM Planner results in it unchecking when it sets and reads-back the parameter.

Once I understand the calibration and control better, I’d love to hook the “roll” up to the Pixhawk for some sort of “damped” body-fixed roll control which allows the camera to stabilize the general flight path, but keep up with the large roll angles you can get in DRIFT mode.

2

Great finding :slight_smile:

Just as a complement could you please say on which pins of the PixHawk you connected the Tarot, i.e. the Roll control to which pins and the Tilt control to which also.

I have a DYS 3 axis and currently use 3 channels straight from the receiver, but wanted to have a bit more “integration” specially concerning the ROI feature of the APM.

3

I’m using the AUX1 channel (RC9), “Minus” and “Signal” pins for the tilt control to the one of the two negative pins, and the “T” pin respectively on the Tarot.

If/when I hook up roll, I’d use the AUX2 (RC10) to the R pin on the tarot.

4

Thanks.

I used that also, but apparently that reduces the available channels (the 8 channels that current APM recognizes on the PixHawk copter)

[quote=“interplanetary”]I’m using the AUX1 channel (RC9), “Minus” and “Signal” pins for the tilt control to the one of the two negative pins, and the “T” pin respectively on the Tarot.

If/when I hook up roll, I’d use the AUX2 (RC10) to the R pin on the tarot.[/quote]

5

Yes, this should be documented! It took me hours to find this!