Replies to This Discussion

Permalink Reply by Neal Kelley on May 29, 2009 at 11:56am

woooooooooooooooow no responders.. ?

• ▶ Reply

Permalink Reply by Brent West on July 7, 2009 at 7:26am

What sort of signal are the XYZ censors driving into the shield? The censors now are not becoming available until mid August.

I’m wondering if I can fake their inputs so that I can go on to testing the navigation part of the autopilot.

• ▶ Reply

Permalink Reply by Tom in ON on July 7, 2009 at 8:47am

I don't have exact numbers, but if you look at the code, I think the software is looking for 511 for each sensor on a scale of 1024. At least for the XY. That's if the sensors are perfectly level, etc. However, the reason you claibrate before flight is to establish what "level" is for your particular configuration/mounting and environment.

When I was testing the XY sensors only without all the other coding, I was getting around 505-530 when I didn't move my hand in front of it, and values ranging between 450-700 when moving my hands.

Did not test the Z sensor.

Some other may have a better and more empirical input on this question.

Please be careful, do not use these values for actual flight, only for testing.

• ▶ Reply

Permalink Reply by Brent West on July 7, 2009 at 3:51pm

Hi Tom, Thank you for responding. I’m not sure how you got the numbers that you’re talking about.
You see, I don’t have my sensors yet. I was getting some bad erratic movements from the servos when I switched to auto mode. And then it dawned on me that maybe it was because the sensor inputs were still open.
I was taking a chance that each of the sensors are putting out an analog voltage between 0 and Vcc, so I made this little header to plug into JP 10 of the shield board. Like pictured above.


Other than finding that my roll servo 1-4 in the easystar.h file needed reversed, I think I am now getting much smoother changes in direction as I walk around the yard. The true test is to see if it will guild me to my waypoints (on the ground of course).

While I have your attention Tom or anyone else?
My platform is a paraglider, so I’m using power to control and maintain altitude hold. From what I’ve read, I will need that Z sensor whenever they become available, correct?
Is there someplace else in the software that I need to look at to say I’m using throttle instead of elevator to control pitch?
Thanks for any help,
Brent

• ▶ Reply

Permalink Reply by Tom in ON on July 7, 2009 at 8:19pm

Brent,

I am by no means an expert at this, still building mine. As far as the Z sensor, you can select in the header file of 2.2.3 version that you do not have one. That should take care of that, but you must follow different calibration procedures.

I think 2.2.3 uses both throttle and elevator to control pitch (please some one chime in here), so do not know if editing out the elevator is an option.

Supplying 2.5 Volts to the analog input pins to simulate XY sensor input may work in making the software think you are at level all the time, but do not know how the plabe would react. An easier way to do this in my opinion would be to make sure that the sensor input values are set to 511 in the software where they read the analog pins. Ultimately, you may be able to use a pot the simulate pitch and roll value inputs. Of course I would not do this in the air.

AurdPilot version 1.0 uses the CPD4 copilot for stand-alone stabilization only, and ArduPilot is used for navigation. Why don't you try version one if you want navigation only for now?

• ▶ Reply

Permalink Reply by Brent West on July 8, 2009 at 9:50pm

Hi Tom,
That sounds like a good idea for now. Curious, how were you able to read the values (505-530) of your sensors, were you able to read the numbers live as you were testing them?
Brent

• ▶ Reply

Permalink Reply by Tom in ON on July 8, 2009 at 10:08pm

Brent,

Here is a program Jordi wrote for sensor testing, this is what I used to check the XY and Z sensors (sensortest).

Also, here are other programs for servo and IR sensor testing. Search in the setup blogs for ArduPilot.

Attachments:

• sensortest.pde, 1 KB
• IR_testing.pde, 2 KB
• Infrared.pde, 2 KB

• ▶ Reply