Home        Store        Docs        Blog

Question about arming

(WYK) #1

Today I put the bluerov in the pool for the first time.
I use qgroundcontrol to let it armed, but once it armed, it started to move in the water(I did not give any other commands and the joystick input is zero).
This problem appeared even when I did not connect my joystick.
Thanks for any help.

(WYK) #2

The mode is manual.
And, could anyone tell me what is the correct movement of bluerov when I set it armed and do not give any other commands? Will it just keep static?

(TCIII) #3


I have a 6 thruster ROV using ArduSub and QGC and I found that the Raw Servo Outputs, which are viewable in the MavLink Inspector, remain at 1500 us (neutral) when the ROV is armed and is in the Manual Mode with no joystick input.

There will be Raw Servo Output from all the channels when in the Stabilize or Depth Hold Modes, but not in the Manual Mode. You might want to use the MavLink Inspector to view the R?C inputs to see if any of the inputs are greater than 1500 us in the Manual Mode. It is possible that your joysticks are not properly calibrated.

Are you using the latest daily build of QGC that does not require the R/C Radio calibration?

Also, are you sure that your ESC are properly calibrated and will give no output when their PWM input is 1500 us?



(WYK) #4

Thank you for your reply!
Before armed, the raw servo output are all 1500. After armed in the Manual mode, the raw serve output of the 6 thrusters are 1204,1100,1419,1603,1868,1646. They change very slightly either with or without joystick connection.
Actually I calibrated both the joystick and the radio in the QGC before. After see your answer, I just download the latest version of QGC and the there is radio in the list as well.
I did not calibrate ESCs, could you please tell me how to do that?

(TCIII) #5


I believe that you are downloading the latest Stable version of QGC as my daily build version of QGC does not have a Radio Calibration Tab under the Gears Tab. It only has a Joystick Calibration Tab now.

Daily Builds can be found here under the Daily Builds heading.

You might want to recalibrate your joysticks after downloading and installing QGC.

When I set the Manual Mode and then Arm my ROV, the Servo_Raw outputs stay very close to 1500 us (neutral) and return to 1500 us when I move the joysticks either vertically or laterally and allow them to return to their center positions.

Are you sure that you are seeing the Manual Mode in the QGC header at the top of the screen when you set the Manual Mode?

Also, are you using the latest version of ArduSub for your thruster configuration?

What are you using for your navigation controller? Pixhawk or RPi/Pixhawk?

What is the model of your joystick? Logitech F310 or MS XBox Joystick?

If the ESCs came with your BlueROV1 kit, they should have come pre-calibrated.



(WYK) #6

Thank you for the instruction. Now I have the QGC without the radio calibration tab. And I recalibrate my joystick.
Yes, I’m sure that there is a “Manual” sign at the top of the screen.
I just downloaded the latest version of Arduous which was updated on 8/3.
I use RPi/Pixhawk and MS Xbox Joystick.
I attach two pictures, which shows the 6 Raw Servo Outputs. The two pictures are armed and not armed respectively, and for both pictures the joystick is NOT connected.
(The voltage is nearly zero because I do not connect the battery. It seems that the battery connection does not affect Raw Servo Outputs.)

(TCIII) #7


If you look at the battery indicator in the QGC header, you will notice that it says 100% so you are getting power from somewhere even when the battery is not connected to the navigation controller.

How are you powering the RPi/Pixhawk? Without the battery connected, the RPi and Pixhawk should not boot up and you should have no communication between QGC and your ROV. Something is definitely not wired up correctly.



(Jacob) #8
Something is definitely not wired up correctly.
That's an interesting conclusion, considering that the pixhawk is, in fact, communicating with the computer. The pixhawk may be powered by usb, or something else, it is not required to use a battery.

Anyway, how you are powering the pixhawk should have nothing to do with the issue at hand. The pixhawk is probably in “stabilize” mode, this used to show up as ‘manual’ in the top bar of qgc, things have recently changed. Wyk, can you please go to the ‘flight modes’ page in qgc and upload a screenshot? This will verify which mode you are in.


(Jacob) #9

@Wyk, additionally, can you post a screenshot of the mavlink inspector with both the RC_CHANNELS_RAW and SERVO_OUTPUT_RAW fields expanded?

(WYK) #10

Thank you two for your answer.
Yes, The pixhawk is powered by usb, and the raspberry pi is powered by another usb.
I think the wiring work is fine.
I post screenshots of the (1)fight modes, (2)RC_CHANNELS_RAW and SERVO_OUTPUT_RAW before arming, (3)RC_CHANNELS_RAW and SERVO_OUTPUT_RAW after arming.

(Jacob) #11

This is a bluerov1? Can you please go to parameters in qgc, select ‘tools’ and save parameters, then upload here.


(WYK) #12

Yes, it is a bluerov1.
Here is my QGC parameters.
Thank you again.

parameters-of-qgc-params (4.11 KB)

(Jacob) #13

Change RC_6_MAX to 2000, and RC_6_MIN to 1000, let me know if that stops the forward thrusters (servos 1 and 2) from spinning when armed. This may require a reboot after changing the parameters.



(WYK) #14

Yes, after changing RC_6_MAX and RC_6_MIN, servos 1 and 2 are very close to 1500us when armed.

(Jacob) #15

Ok, it looks like it is responding accordingly, your sticks are a little off-center. I suggest you change min and max for channels 1-8 to 1000 and 2000 in the same manner.

To clarify things in regards to your initial question about what the behavior should be when armed:

In manual mode, the pilot input is sent directly to the motors, with no stabilization.

In stabilize mode, the vehicle automatically levels itself and maintains a heading.

Depth hold (or alt hold) mode is like stabilize mode, but the vehicle also maintains its current depth with an external pressure sensor.

If you arm in stabilize or depth hold mode while the vehicle is out of water, the motors will gradually start spinning faster and faster in an attempt to get the flight controller exactly level. Therefore, it is recommended to do bench testing in manual mode.

(WYK) #16

I change min and max for channels 1-8 to 1000 and 2000, and I change the trim of channel 1-8 to 1500.
The servo_outputs seem still not correct when armed (without joystick input)

(WYK) #17

I just adjusted the TRIM number of some channels. And then the servo_outputs show like this. It seems every servo output is 1500 when armed which meets my expectation. But is this method correct?

(Jacob) #18

@Wyk, no, you should not change the trim, especially not very far from 1500 like 1000 (ch3) or 1100 (ch5)

It is reacting appropriately I think, it is just sensitive, try decreasing the input gain by configuring a joystick button to do so.


(WYK) #19

I find that if I change the trim of channel 1-8 to 1500, the servo_outputs will be far from 1500 even if there is no joystick connection. For example, servo5 will be 1900 when armed.

(Jacob) #20

@wyk I will try to test this tomorrow. Something is up.

In the meantime, a joystick must be connected, and can you tell me if the servo outputs change when you wave the pixhawk around?