Developer

Indoor fun with a tiny ArduCopter quadcopter



Do you have an Ardupilot Mega and shield and want some fun for little money? check this tiny quad... my last drone in family...



Some specs:
- Size : 26.5cm (10.4 inches) from motor to motor
- Motors : 18-11 200kv 10 gram motors
- Props : 4x2.5 (standard CCW props)
- ESCs : Turnigy 6A
- Battery : 3S610 (or 3S800)
- Sonar : Maxbotics LV-EZ0 [optional]


The nice part about this frame is that it use standard kite parts, so it´s very easy to construct and it´s perfect for some radical tests.
Itried to find a bit different design that also make the quad less simmetrical (it´s very easy to loss the orientation with a quad) and I wanted to do as small as posible to flight indoors...
At this size and weight it´s very safe and robust also. The props use prop savers and this help a lot in crashes...
I am not very good writing... so I took a lot of photos during the build phase and I created a build log:
https://docs.google.com/document/d/1fIZL-Ca2fx2RPhBQLzoOdqdCuikXRKePwKbxDlJWiss/edit?hl=es

This code supports a sonar for altitude hold and also has a battery alarm (2 alarms).
Thisquad works with ArduCopter code, but because how main board is mounted, actual code doesn´t work. This is the code that I used with this quad :
http://code.google.com/p/arducopter/source/browse/#svn/branches/Arducopter_indoor
(as zip format:
Arducopter_indoor.zip )
note that this will be integrated on main Arducopter code. Now is temporarily in a branch.
ThePID gains are different than in standard ArduCopter, so you need to Initialize EEPROM on Configurator or put this ones manually:
-Stable mode: P : 4.8 ; I : 0.12 ; P_rate (D in configurator) : 0.4 ; Yaw : (4.5 ,0.15 ,2.8 )
-Acrobatic mode : P : 0.75 ; I : 0.1 ; D : 0 ; Yaw : (2.4 ,0.2 ,0 )

OK, but was all so perfect like in the video? I don´t believe that...
Yes, you are right, but fortunately this design is very crash resistance... and this part is also fun...


After all this battles I only needed a few drops of ciano in the crash with the lamppost :-)

What´s the real size?


On center, standard arducopter frame, on left, old ArduIMU quadcopter part III, on right-top this project.

And what about the automatic aerobatics on video?
I am starting to test a new function to do automatic aerobatics (Automatic Aerobatic Procedure). The idea is that you enable this function with a switch in the Tx to make an automatic predefined aerobatic and then automatically recover to a stable mode.
Thisis actually a proof of concept but works very nice. One limit here is the maximun rate of the gyros, in our case 500deg/s. If we exceed this, the IMU get confused...
An interesting point to note is that the IMU doesn´t loose the orientation in this hard movements...
Thestory is that I have never feel too much confidence doing flips manually with quads. Then I start thinking that with an automatic code we can get a precise timing and controled movement so maybe this could help to do better aerobatics. I started testing this function outdoor, on an open field but I was improving the code and got more tightly maneuvers, so tight, that I ended up doing flips with a quad inside my room!!!


Here is the code I was testing : (AAP function:
AAP.txt )
I implemented this functionality as a state machine where there are three kind of states:
- fixed commands for a limited period of time. (example : Apply a throttle command)
- fixed commands until we reach some IMU state (example : roll until we reach a certain angle)
- And a mix between this categories and a control (example : Throttle command + stable mode on roll, pitch, yaw)
There are many things to test and improve here but the basic concept works very nice...

If I can get a bit of help from a friend, I will probably made a carbon fuselage for this prototype...

Enjoy this “Science & Fun” mix...

Jose Julio.
Arducopter Team
E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • 1. exaserbate Definition and used in a sentence:
    Ivan totally exaserbated me by correcting my spelling of exacerbate!!
  • i highly suggest trying a new battery first, if a full pack under load goes that far down on voltage during initial takeoff its probably got a bad cell. bigger motors may exaserbate the problem.
  • Yeah, I think my ESC's cut of is like 3v, so very near when 3 cells are 9.5v....

    Another problem now, she tend to fall left seriously. Difficult to tune.

    I ordered bigger motors from HK.

  • tero, saw your video, seems your battery voltage was dropping way too low under load. imbalance of rpm to the props occour and loss of stability and down she goes... it happens to my brothers parrot on just one of his three battery packs. try a new battery
  • Developer

    Great Ralph!

    a) If you like to tunning PID, I recommend this : The main param to adjust is P_rate (that is the D param in the configurator). The second param to adjust is P and you usually don´t need to move I (you can try to use a bit lower values...)

    b) I had no problems with sonar sensor. I assume you are using the same sensor (LV series). In  this sensor (as the manufacturer says) you need to take into account that in the startup (when you connect battery) there would be no objects in about 50cm) in order to initialize properly. In altitude hold it mantains in a 5-10cm range...

     

    @Eric: Is not easy to get CCW and CW props in all sizes. In 4x2.5 size I don´t find CW props...

    Jose.

  • Hi Jose Julio,

    this weekend I was able to finish my TinyArducopter and can report that your design worked for me very well - thank you! I would have two questions:

    a) I think I would like to try different setting of the PID-control algorithms - do you know if there is a systematic way to find the "right" values or is this more or less trial-and-error?

    b) I replaced the original (longer-range) ultrasonic sensor with an (shorter-range) infrared sensor. The main reason for this is that I was not able to get a reliable reading from the ultrasonic sensor (even with the added capacitor. Could you say something about the readings you get from your sensor? How much is the arducopter moving if you engage altitute hold?

    thank you

    Ralph

     

  • Mine "crash test" series, maybe OSD system weights too much (80g), or battery can't stand -10 C

    https://www.youtube.com/watch?v=G56HwS4FKVw

  • Pizolq - I was wondering the same thing.  If you look closely, Jose has a single white signal wire per channel going from the receiver and directly soldered to the APM.  Only one channel has all 3 wires connected so it can power the receiver.  He only has enough 3 pin angle headers to connect the ESCs.

     

    You can see it clearly on his linked build log on Google docs.

     

    Looks like nice way to get rid of the nest of wires from the reciever to APM.  I have to be careful they don't get chopped up by the props on my mini.

     

     

     

  • Jose. Is clear for me that between ESC and APM output we have to remove the red cable, but what about between APM input and receptor ¿how many cables de you connect for each chanel? In the photos it seems to me that just one...
  • Awesome Jose'! From your build log, it says that you used 4 CCW props and mounted 2 of the opposing motors at 8deg to  counter-act the yaw torque caused by all 4 motors turning in the same direction.. Just to be sure I'm not missing some point of theory, is there any reason you didn't use  2 each CCW and CW running motors? Just a matter of personal preference?
This reply was deleted.