Kevin's Bi-Copter Thread

[jstremmler]

Link for setting up speedybee bluetooth: https://www.youtube.com/watch?v=7D6ZmRyOZ50
(hope with my OmnibusF4 Pro corner the command: “resource pinio 1 A03” for motor 3 will work for switching on/off the bluetooth by disarming/arming the bicopter…)

[Kevin_Erik]

I have the speedybee already… though there was something different used for dRonin is all.

[Kevin_Erik]

Here is my most recent work using Plasma Tree to tuning the Bicopters Pitch.

From the above Video there are a few steps that really stood out to me.

30-10-30

40-15-40

40-15-80

40-25-60

What I’m looking for in these plot’s is a rise time to 1.0 that’s as quick as possible while having as little overshoot as possible. Obviously overshoot with a Servo is inevitable but so long as the overshoot isn’t excessive (1.25 or >) and the plot levels out at or near “1.0” then it should be Ok. Keep in mind that there is a lot of interaction between Pitch and Yaw. Thus it may not be possible to tune in the Pitch without first improving Yaw some as well.

[Kevin_Erik]

I did a little extra work this morning.

And have come up with the following Pitch Value that seams pretty good.
40-10-55

All my other work, DIFF and resources are attached.

Attachments:

1. 

   40-10-50.jpg
   

2. 

   40-10-55.jpg
   

3. 

   40-10-60.jpg
   

4. 

   40-10-65.jpg
   

5. 

   40-10-70.jpg
   

6. 

   40-10-75.jpg
   

7. 

   40-10-85.jpg
   

8. 

   40-10-80.jpg
   

1. Diff-8-Oct-2018.txt
2. Resource-8-Oct-2018.txt

[jstremmler]

How does your bicopter fly backwards and how does it land with these PID values for pitch?

Yesterday I had a failsafe with my bicopter because of low RSSI. Maybe both 2.4 GHz antennas were in a flight position where the were right behind the CF frame??? Think I shoudn’t just let them hang our of the back of my bicopter…

Attachments:

1. 

   20181008_113434.jpg
   

[Kevin_Erik]

Went outside for a real flight and this was the plot result. (Looks clean!)

Roll needs a touch more P and maybe a little less D.

Pitch is very close and looks like it needs I & D adjustment.

Yaw, i haven’t touched yet but thats coming. The warble heard in the video during hard throttle up is from the mistuned Yaw.

Video (Should show in HD once processing is done)

[Kevin_Erik]

@jstremmler

As i mentioned before, the Bicopter is nose heavy and thus balancing the copter out will be very difficult if too much of your electronics is forward of the CG. I ended up adding two self adhesive 3-Gram weights (Ballast) to the underside of the copter. This offsets the CG where it needs to be so I no longer have to balance it by hand when attaching a battery. Later as needed i can move the ballast further back to amplify the effect without having to add more weight. Also notice where and how the receiver card is mounted. (double-sided tape) The antennas for it are run though heat-shrink thats over the winglets.

[Kevin_Erik]

Had both Servos fail together in flight. See log for details.
I figure that the gear trains for both are toast but the motors is still buzzing away.

Attachments:

1. Both-servos-seized-in-flight.zip

[jstremmler]

@Kevin,

thanks!

btw you BLF file shows slightly different PID’s than the video:

PID Settings
Name Proportional Integral Derivative Feedforward
Basic/Acro
ROLL 30 28 45 60
PITCH 40 22 50 60
YAW 90 25 5 0
Angle/HorizonStrength (Angle)Strength (Horizon)Transition (Horizon)
LEVEL 50 50 75

[Kevin_Erik]

Thats cause i am testing a large number of values and selecting the best one before moving to the next term. Its a slow and methodical process but should work given time.

[Kevin_Erik]

Was able to install the BMS-385DMAX servos as direct replacements for the 210’s. While i was at it, i rewired everything saving about 10-15 grams while rebalencing the copter better.

Tried the first flight with the Ducts installed and was able to fly abeit poorly. After a quick tune i was able to stablize it quite a bit and may attempt to do a full tune on it tomorrow.

[Kevin_Erik]

Changed out the RS2205-2300kv motors for LS2207-2550KV’s.

I defaulted the tune to something really basic like 20 / 10 / 10, attached were the results.
Note: Feed Forward, Antigravity, and I-Term Rotation was disabled.

Attachments:

1. 

   LOG00001log-3_0_response.png
   

2. 

   LOG00001log-3_0_noise.png
   

[Kevin_Erik]

Working on installing dRonin!
Attached are the current channel assignments for the Matek F405-STD.
Be aware that S7 is currently unavailable and thus the GCS may need an update before this is available.

Attachments:

1. 

   Matek-Assignments.png
   

[Kevin_Erik]

I have the updated GCS installed and all of the Servos / Motors are working.

Thanks to jihlein, Here’s the update link for the Matek F405-STD. This will allow pin 7 to be properlly addressed.

Windows:
https://1drv.ms/u/s!Am7KDquQANJl5mVX8sL99fvUkZHS

MACOS:
https://1drv.ms/u/s!Am7KDquQANJl5mQ_EHM-Xlg3THaV

[Kevin_Erik]

Today i broke the Carbon Arm after the copter didn’t recover from a loop in time.

Sigh… BF Tune is getting pretty good too. Unfortunately the Log of my last flight didn’t save so I have only the Diff and the Plasmatree from the prior flight to show.

Difference between the crash-flight and the prior is that I added Pitch & Yaw P-Term to increase it’s responsiveness. (Pich +5 / Yaw +1) Tuning this is extremely hard as too much gain results in uncontrolled oscillations.

Attachments:

1. 

   LOG00001Flight01_0_response.png
   

2. 

   LOG00001Flight01_0_noise.png
   

[Author]

Posts