[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [Paparazzi-devel] ArduIMU and Tiny autopilot continued

From: mark.griffin
Subject: RE: [Paparazzi-devel] ArduIMU and Tiny autopilot continued
Date: Wed, 1 Dec 2010 16:05:51 +0100

Hi David,

When I tested the Arduimu on my Funjet, I found that I had to reduce the 
p_gains (ROLL_ATTITUDE_GAIN and PITCH_PGAIN) to avoid oscillations. I also kept 
the roll gyro connected. It flew fine at around 20 m/s airspeed. 

I did another test to see what would happen at higher speeds. At full throttle, 
the Funjet lost control and spiraled down to the ground. Luckily I was able to 
switch to manual and recover.

I concluded that, due to the relatively slow response of the Arduimu (~15 Hz), 
it's more suitable for slow flying non-agile aircraft.

Hope this helps, Mark

-----Original Message-----
From: address@hidden [mailto:address@hidden On Behalf Of Gareth Roberts
Sent: Wednesday, 1 December 2010 12:37 PM
To: address@hidden
Subject: Re: [Paparazzi-devel] ArduIMU and Tiny autopilot continued

Do you have a record of the G-values (are they shoved through the  
You could also fly manually with the arduimu connected to an xbee in debug  
mode, and log everything.
Flying with the arduimu flashed with a build in debug mode would be best -  
you could then see if there are any errors coming from the code (it's well  
It could also be integrated windup - in my experience with ardupilot  
(which, to be fair, was a year or so ago), it's much harder to tune  
properly so most planes may be flying around tuned fairly slack (ours  
certainly was).  The faster response rate and harder turns from a  
well-tuned paparazzi might be edging it over.


On Tue, 30 Nov 2010 23:55:11 -0000, David Conger <address@hidden> wrote:

> Hello,
> Just an update so I thought I would extend this thread for others  
> following. I was able to power all servos from the BEC. Only the PPM  
> wire from the autopilot goes to the servos. Ground and power to the  
> servos comes from the BEC now.
> So, I was able to fly. However, during sustained turns the autopilot  
> commanded a sharp roll which I was able to recover from. It is  
> consistently doing it. Also if the autopilot starts level it tends to  
> stay level. However if you engage AUTO2 while the aircraft is not level  
> it does not recover well (or at all).
> I realized after viewing this video:  
> that I am using something different. That video was using the 6DOF razor  
> which is just a sensor board. So the DCM code must be running on the  
> autopilot in this case. In my case with the ArduIMU the attitude  
> information comes from the ArduIMU. So if it's incorrect the autopilot  
> is unaware.
> Chris from DIY drones seems surprised at this behavior and claims  
> hundreds are flying with this IMU and the latest IMU running the same  
> code. I'm not sure how it's possible but believe him.
> Did I miss something? I have a video showing the attitude information is  
> correct on the ground. However in flight it simply is failing to  
> stabilize my little Merlin aircraft. I am hoping maybe there is a tuning  
> parameter I am missing someone can share with me to correct.
> I have not flown in weeks so I hope to get some new data soon.  However  
> my questions are:
> 1. Am I supposed to be tuning the IMU somehow? If so how?
> 2. The only video of a working example is not using an AHRS. Has anyone  
> a video or proof of a Tiny+ArduIMU flying? Logs that can be replayed?  
> Anything? I heard of a FunJet...can they share some more info to back up  
> the claim and help me get flying. I have a FunJet BTW as well setup to  
> go but don't dare try if the Merlin won't fly.
> Regards,
> -David
> On Nov 7, 2010, at 4:12 PM, David Conger wrote:
>> Well, all is on hold. The plane flew well in manual but I found the  
>> 3-pos switch was not working. It stayed in manual unless I turned off  
>> the radio. I am using a T7CAP Tx and the same files that worked  
>> previously so this is a mystery.
>> -David
>> On Nov 7, 2010, at 8:33 AM, Martin Mueller wrote:
>> _______________________________________________
>> Paparazzi-devel mailing list
>> address@hidden
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden

Paparazzi-devel mailing list

reply via email to

[Prev in Thread] Current Thread [Next in Thread]