preflight calibration

[Andrew Tridgell]

Hi Michael,

A user noticed that doing a "preflight calibration" in MP was causing a
1D accel cal. This was bad for two reasons:

1) it loses the 3D call essential for EKF

2) if the plane was not level (say it was on a catapult) the resulting
flight would be rather disasterous

I've just pushed a change to APM:Plane for the 3.1.2beta3 release so
that prefight cal works like this:

param3=1 -> baro and airspeed cal
OR
param1=1 or param2=1 -> gyro and 1D accel cal
OR
param4=4 -> radio trim cal
OR
param5=1 -> 3D accel cal

I think that is the best combination for how MP behaves at the moment.

Bill, can you check that APM Planner will work with this?

Cheers, Tridge

[Chris Anderson]

Great catch! I know our friend Brett can be a big high maintenance, but this is a great example of a demanding user working directly with the dev team to document and solve an issue. 

-Chris

--
You received this message because you are subscribed to the Google Groups "drones-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Chris Anderson

CEO, 3D Robotics

ch...@3drobotics.com

[Bill Bonney]

At he moment APM Planner send param1=1 and param 3=1 at the same time. Are they all mutually exclusive?

I can create 5 separate entries tomorrow for each type

Cheers

Bill

[Andrew Tridgell]

Hi Bill,

> At he moment APM Planner send param1=1 and param 3=1 at the same time. Are they all mutually exclusive?

the way common.xml is written they aren't, but in practice I think it is
better to do just one type of calibration.

> I can create 5 separate entries tomorrow for each type

as long as it fits with the little table in my last email we should be
OK

Cheers, Tridge

[Sarge]

Hi Tridge. I currently do a "preflight calibration" in MP before each launch after the aircraft has been powered up for a few minutes. I ensure that the pitot tube is covered and that the aircraft is kept still. For the benefit of all users, can you please clarify the current and planned behaviour (and subsequent implications) of "preflight calibration"  for different configurations e.g. :

• APM:Plane 3.1.2+ (future release)
  
• APM:Plane 3.1.1 (current release)
  
• APM:Plane 2.7X (pre 3.X release)
  
• With DMC or EKF enabled
  

Do you personally run any sort of calibration before launch on your aircraft? Is there a "best practice"? 

Kind regards,

Nick

[Randy Mackay]

For copter (master and AC3.2) we've removed the 1D accel calibration.

Michael has changed the "PREFLIGHT CALIBRATION" on the mission planner's Flight Data, Action screen to re-do the gyro calibration (i.e. param1 = 1) which is important for AC3.2 because we've added a pre-arm check that the gyro calibration has succeeded and without this button the user must reboot the board.
https://github.com/diydrones/MissionPlanner/issues/694

I've written what copter does below, basically the same as plane except it allows multiple calibrations to be done which isn't helpful so I'll make it consistent with plane.

Copter----------------------------------------------------

param1=1 -> gyro
param3=1 -> baro

param4=4 -> radio trim cal

param5=1 -> 3D accel cal

param6=1 -> compassmot

Plane-----------------------------------------------------

param3=1 -> baro and airspeed cal
OR
param1=1 or param2=1 -> gyro and 1D accel cal
OR
param4=4 -> radio trim cal
OR
param5=1 -> 3D accel cal

-------------------------------------------------------------

-Randy

[Andrew Tridgell]

Hi Nick,

> Can you please clarify the current and planned

> behaviour (and subsequent implications) of "preflight calibration" for
> different configurations e.g. :
>
>

> - APM:Plane 3.1.2+ (future release)
> - APM:Plane 3.1.1 (current release)
> - APM:Plane 2.7X (pre 3.X release)
> - With DMC or EKF enabled

It wasn't the plane behaviour that had changed - I think MissionPlanner
changed behaviour recently, possibly to accomodate a request by Randy
for copter.

The problem is that the new MP behaviour made plane do an accel cal when
you did "preflight calibration". That meant:

- the plane had to be level
- you lose the 3D accel calibration scaling factors

> Do you personally run any sort of calibration before launch on your
> aircraft? Is there a "best practice"?

I do a "calpress" in MAVProxy, which calibrates baro and airspeed sensor
(with pitot covered). I have SKIP_GYRO_CAL=1, so it doesn't do a gyro
cal on startup. I do a 3D accel calibration once for each new Pixhawk.

Cheers, Tridge

[Sarge]

Thanks Tridge!