Thread Tools
Jul 07, 2010, 02:46 AM
Registered User
icebear's Avatar
Thread OP
Discussion

Gluonpilot project


I just received the Gluonpilot board from Tom Pycke described here earlier.

The PCB is proffesionaly made, quite small and includes;

80MHz Microchip PIC
ADXRS613 Z gyro
IDG500X&Y gyro
ADXL335 accel
Pressure sensor SCP1000
2MB onboard flash memory for datalogging.

An interesting feature is that it also supports Quadrocopter stabilization.

I will start testing the fixed wing stabilization and "RTL and circle" functions using my beat up but trusty SuperImpress airframe. Just waiting for some servo connectors to arrive tomorrow...

Firmware updates to include wp navigation to come - maybe a 3rd Lighthouse mission this summer?

See also www.gluonpilot.com

Flight reports to follow soon...!

/Björn
Sign up now
to remove ads between posts
Jul 07, 2010, 03:18 AM
Registered User
Hi Bjorn

Looking forward to your progress, new lighthouse mission would be an interesting test, the whole Gluonpilot system looks impressive
Jul 07, 2010, 09:50 AM
Registered User
Where and how do you buy the hardware? Looks nice and small.
Jul 07, 2010, 11:53 AM
Registered User
icebear's Avatar
Thread OP
Update:

Found some spare PCB headers so that I could connect an USB/serial FTDI cable and test the module. Everything works well so far. Tomorrow I will get the remaining headers for Rx, servos and GPS. Another interesting feature is that you can use a 'sum signal' from the rx if you have acces to it and save lots of wiring (only signal/GND/+5v for the rx connection).

@ spitfire: yes, that's the standard challenge for all my autopilots!

@ airmcn: I got my hw directly from Tom Pycke and I would recommend contacting him for pricing and availability

/Bjorn
Jul 07, 2010, 12:40 PM
Registered User
v22chap's Avatar
Looks like a neat little package ... good luck and looking forward to the results and video
Jul 08, 2010, 03:57 AM
Registered User
Paul_BB's Avatar
Yes good luck ! (A mission to and back from the Helagsfjället would be nice of course, but don't consider this choice as a moral obligation ).

- Paul
Jul 10, 2010, 07:53 AM
Registered User
icebear's Avatar
Thread OP

Installation complete!


Thanks Paul and v22!

After finally receiving some PCB headers from a local webshop I managed to bench test the Gluonpilot and install it in my Super Impress. As can bee seen in the picture this 10g autopilot PCB doesnt require much room...

Calibration is done via the PC and is a simple affair - place the PCB flat on the table and push to buttons and you are done!

I have verified control surface movements and will do a test flight with the recommended Pgain settings and +/- 35 degrees roll and +/- 15 degrees pitch.

Watch this space...

/Björn
Jul 10, 2010, 10:01 AM
Registered User
Paul_BB's Avatar
Quote:
Originally Posted by icebear
Watch this space...

/Björn
I lost two planes (the second one with an AUVDevBoard onboard) because of a Hitec 55 servo failure.
I think that the servos are stressed by the stabilization, they tend to overwork. In manual mode the plane does'nt fight against turbulence, but in auto mode it does: the servos have no rest.
The first plane was not rigged with an AP, but after exactly 137 flights one of the Hitec 55 failed (the plane was a 40" parkflyer).

-Paul
Jul 10, 2010, 11:53 AM
Registered User
v22chap's Avatar
This makes sense as we have found on heli tails that now that we have such great heading hold gyros normal servos will get hot and soon quit ..so we are running very high speed , brushless motored ,very powerfull servos on the tail rotor control now .
Jul 10, 2010, 12:08 PM
Registered User
icebear's Avatar
Thread OP
Quote:
I lost two planes (the second one with an AUVDevBoard onboard) because of a Hitec 55 servo failure.
I think that the servos are stressed by the stabilization, they tend to overwork. In manual mode the plane does'nt fight against turbulence, but in auto mode it does: the servos have no rest.
The first plane was not rigged with an AP, but after exactly 137 flights one of the Hitec 55 failed (the plane was a 40" parkflyer).
-Paul
Thanks Paul - noted - AND I agree that I really dont like the HS55s.
I normally use better servos but the Impress have had these since the start - in 2004? - and has flown numerous hours as an UAV so they have definitely served enough by now. The aileron servo is a single HS81 and I believe that is stressed more.

First mission was a succes - will post results soon!

/Bjorn
Jul 10, 2010, 12:32 PM
Registered User
icebear's Avatar
Thread OP

First flight - success!


I managed a first flight today since the winds co-operated. To make a long story short it was a succesful test.

Calibrated at home, went to the field and took off to about 80 m altitude and switched to "stabilization". With the P-gains fairly low it was evident that the turns were quite mild, but stablization definitely worked.
I switched to Auto and the plane slowly turned back and started to circle.
The Gluonpilot logs all flights and generates a nice KML-file with colours to indicate man/stabilization/auto - see attached file.

I need to increase P-gains a little but all in all a good test after such a quick integation!

/Björn
Jul 18, 2010, 03:07 PM
Registered User
Thanks for the report icebear.

I am very interested in this project and plan to buy a board for jet stabilization. I like to implement the homing function out of curiosity but stabilization is my first goal.
Jul 18, 2010, 03:58 PM
Registered User
Does it have airspeed gain scheduling? IF not then may not be a good idea for stabilization on a jet with a wide speed range.
Jul 19, 2010, 02:26 AM
Registered User
Quote:
Originally Posted by airmcn_3
Does it have airspeed gain scheduling? IF not then may not be a good idea for stabilization on a jet with a wide speed range.
I have just downloaded the source code but have not looked into it jet. I guess you can calculate an accurate airspeed from the accelerometers and implement a gain that adapts to the air speed. My first interest goes out to hovering with vector thrusting.
Jul 19, 2010, 12:55 PM
Registered User
icebear's Avatar
Thread OP
Quote:
Does it have airspeed gain scheduling? IF not then may not be a good idea for stabilization on a jet with a wide speed range.
I don't think so, and for a jet I might go for a more proven autopilot unless you (Allintheair) are into it because it is an open source project.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion RCG Project Mini/Micro Electric Helicopter Silverexpress New Forum Requests 26 May 13, 2010 02:53 AM
Sold 48" mini-shark sloper - rebuild project T_M_B Aircraft - Sailplanes (FS/W) 3 Feb 28, 2010 11:55 AM
Alert does any body know any thing about project rainbow or the Philidelphia project? b36roxs Dock Talk 17 Mar 11, 2007 08:35 PM
Discussion Father/Son Project changed to family project... ApachePilot Mini Helis 71 Dec 21, 2006 10:09 PM