Complete FPV tricopter flight in San Vito lo Capo, Sicily
ฝัง
- เผยแพร่เมื่อ 25 เม.ย. 2014
- This is the complete footage of the first flight I made there that day.
Gear:
rcexplorer style tricopter with 36cm 10x10mm hard wood arms
APM 2.5 flight controller with Arducopter 3.1 firmware
MinimOSD
T-Motor MT-2216 900kv motors
rctimer ESCs with SimonK firmware
HQProp 9x3.8 propellers
3000mah 25C 4S Turnigy Nanotech lipo
GoPro 3 Black Edition
Scherrer NR
TGY-390DMH yaw servo (fit's perfectly and has a metal case which is perfect for tie wraps) - วิทยาศาสตร์และเทคโนโลยี
Thanks! It's one of my favorite places, great flight!
Amazing.
S.vito very nice !!!
Can't believe some one gave a thumbs down. Awesome flight! San Vito lo Capo is Beautiful.
Thanks!
Your tricopter looks very very stable. I can't believe my eyes that is your video is not use gimbal.
Can you tell me about PID setting?
What servo you are using for yaw?
How about weight?
Thanks.
It is very stable, but it still has some high frequency vibrations which I think come from resonance in the wooden arms which aren't stiff enough. I used some limited (5 or 10%) Mercalli stabilization to remove them.
I can't remember what the PID settings were as I did an autotune to get them.
It's very similar to the rcexplorer 2.6 build, but with hard wooden 10x10mm arms. If the arms are too long or too flexible, then it'll shake too.
The yaw servo is a TGY-390DMH. It fit's perfectly and has a metal case which is perfect for tie wraps.
More details are in the description below the video.
***** quando ti trovi nella mia zona sicilia occidentale fammelo sapere ci facciamo un volo in fpv in formazione... !!!
dronevideo sicilia That'll be fun. When I'm there again, I'll let you know.
dronevideo sicilia I have another question: do you know if or when they clean that black seaweed out of the water? It makes the water look dirty. The video was in April, but I've been there before in June/July and it was cleaner then.
@tursurfer: I've built custom tricopter frame plates for APM/Pixhawk flight controllers to simplify/clean up wiring.
check this post on RCG: www.rcgroups.com/forums/showpost.php?p=35580911&postcount=784
and this one too: www.rcgroups.com/forums/showpost.php?p=35812928&postcount=786
Frame plates integrate Current/Voltage Sensor, PDB, OSD, 3xBECs 5v, 12v and 7.4v for servo, each 3A rated. 5v and 12v are also filtered to reduce interference form motors/ESCs.
12v has unfiltered output as well for non-sensitive accessories like LEDs and Gimbal.
if interested, let me know, I still have a few frames both populated and non-populated.
Nice looking tricopter! You must be using bigger motors and propellers than me to be able to carry a gimbal. Are you using some sort of L shaped bracket to mount the gimbal to the dampened (battery) plate?
Have you had any yaw oscillation issues with Arducopter? I had to tune one of my yaw gains very low to get it to stop wagging (which is caused by the mass of motor+prop being above it's pivot hinge).
I personally would not mount the VTX on one of the front arms as you did as turbulence in the form of horizontal gusts will cause small unwanted yawing as the yaw servo can't tilt and correct as fast as a quadcopter can.
turdsurfer i do not remember the name of param in the older firmwares (atc_y_slew_rate), but in the 3.4 its called Y axis acceleration, I've maxed that one out, yaw PID remained default. btw imho u are wrong about the mas of the motor/prop causing yaw oscillation, simply because mass of the motor is hanging off the lift disc propeller creates, its more about the change in the default settings for tri... since 3.1.5 those became mostly garbage.
motors are (iris+) 950kv with 1045 tmotor CF props hovers at just 10-11amps. 5200mah 4s multistar pack allows for 17-20min active flights.
i did not use any special brackets to mount gimbal, its just one of the cheap ebay gimbals you can buy without motors/controller without top mounting plate.
Artem Nikitin Thanks for the info. I'll take a look at those PIDs.
The lift of the propellor isn't that relevant initially. First mass is shifted sideways causing an immediate equal and opposite reaction in the frame of the tricopter which is a slight yaw in the opposite direction in which the frame wants to go. The sideways lift component of the propeller is different: It's effect isn't immediate, but a gradual increase relative to angle.
To to confirm my theory, I've also experimented with reversing the motor+prop to the underside of the frame so that they both tilt in the opposite direction, effectively pushing the frame in the direction that it needs to go. Simply doing that removed all yaw oscillations as I'd expected.
Of course the effect of the oscillations with a top mounted motor is dependent on the mass of the frame, the mass of the motors+props, the servo angular speed, and the arm lengths. If you're using a relatively heavy battery and a gimbal, you're going to notice it less.
turdsurfer "lift of the propeller is not relevant initially" that's the correct statement, however once the lift is sufficient to lift the copter, mass of the motor becomes irrelevant as that mass just adds up to the total mass being lifted. the change in yaw oscillation you've observed is due to some other reason, most likely because the mass of the copter was above the lift, also could be due to the aerodynamic of the yaw mechanism not affecting the thrust angle. (I've actually experimented with yaw being controlled by a simple fin-like a credit card sized rudder and the tail motor beeing fixed. well though the yaw authority was not as great as with the traditional serup, it was very well sufficient enough to control the copter.
i will find that parameter name and my yaw pid as soon as I can.
I disagree.
If as an example, the FC decides that a yaw correction is needed and that the yaw servo must tilt 30° to the right, then what will happen is 2 things:
A: There will be a gradual increase in the right component of the lift vector, starting from 0 until tan(30) = 0.58.
B: Simultaneously, there will be an immediate leftward force on the tail boom (as an equal and opposite reaction to the motor mass deflection to the right). This force will decrease as the motor angle increases and then suddenly stop when the servo reaches the 30° tilt.
For your theory to be true the forces represented by A and B should be mirror images of one another at any moment in time.
But that's simply not the case and 2 evident reasons prove it wrong: A has a sudden start and stop. B has a gradual increase and never stops.
It has gimbal?
no