6 August, 2018 at 19:17 #58669
Been out doing yard work this morning which means my real brain (subconscious) has been working on a solution. I’ve been going about this a bit wrong by testing in leveling mode. That can’t work on the pitch axis since there is no pitch authority possible but rather the weight hangs on the props and swings freely. So at best, the FC could do roll leveling and/or rate control, Pitch is full manual, and Yaw could be rate mode or manual. (Rate = Acro) So I set up a custom flight mode “stabilized1” that has attitude (leveling) for roll, manual for pitch, and rate for yaw. Will give it another go later today.
Also been perusing YouTube for other BiCopters and found a few very well flying KK2 based ones like this https://youtu.be/nkzap_DiHcg
Certainly if it could be done on the KK2, we can do this with more modern tech. 😀
LitterBug6 August, 2018 at 22:47 #58672Kevin_ErikModerator
Watching that video, one thing i noticed right off is that the props are not level but instead appear to be angled 5-10 degrees. Furthermore there are Oscillation but its not as obvious due to the slow speeds with which the copter was flying.
Edit: The tail reminds me of a dragonfly. Could the tail act to dampen oscillation?6 August, 2018 at 23:02 #58675
I’ve made a bit of progress today, still not dialed in but got rid of the uncontrolled oscillations. Cut default PIDs in half and got rid of leveling on pitch axis. Using Rate(acro) on pitch it will eventually get into a pitch oscillation. Lowering the battery, or adding some horizontal stabilizers might fix that, but I’m going to lower the pitch mix and try to get it to work with manual pitch mode. Currently with pitch in manual mode, it tries to flip over forward or backwards and roll across the ground. Really need to get some video so you can see all the fun I’m having. Haven’t killed or hurt a prop yet even with plenty of grounding the props.
LitterBug12 August, 2018 at 18:13 #58739
Here is the latest diff from the original prototype:
# Betaflight / RCEXPLORERF3 (REF3) 3.4.0 Jul 8 2018 / 04:27:31 (cce8a59be) MSP API: 1.39
resource MOTOR 1 B00
resource MOTOR 2 A04
resource MOTOR 3 NONE
resource MOTOR 4 NONE
resource SERVO 1 A07
resource SERVO 2 A08
mmix 0 1.000 1.000 0.000 0.000
mmix 1 1.000 -1.000 0.000 0.000
servo 2 900 2100 1585 100 -1
servo 3 900 2100 1420 100 -1
servo 4 1050 1950 1500 100 -1
smix 0 2 2 100 0 0 100 0
smix 1 2 1 -100 0 0 100 0
smix 2 3 2 100 0 0 100 0
smix 3 3 1 100 0 0 100 0
serial 1 32 115200 57600 0 115200
aux 0 0 0 1300 1700 0
aux 1 1 1 900 1300 0
aux 2 2 1 1300 1700 0
aux 3 28 0 1725 2100 0
set gyro_lowpass_hz = 90
set acc_calibration = -12,17,47
set mag_hardware = NONE
set baro_hardware = NONE
set rssi_scale = 41
set rc_interp_ch = RP
set min_throttle = 1020
set use_unsynced_pwm = OFF
set motor_pwm_protocol = ONESHOT125
set motor_pwm_rate = 2000
set align_board_yaw = 90
set current_meter = ADC
set ibata_scale = 336
set ibatv_scale = 336
set servo_pwm_rate = 250
set deadband = 5
set yaw_deadband = 6
set dterm_lowpass_type = BIQUAD
set dterm_notch_hz = 260
set anti_gravity_gain = 1000
set setpoint_relax_ratio = 100
set dterm_setpoint_weight = 0
set iterm_windup = 50
set yaw_lowpass_hz = 50
set throttle_boost = 0
set p_pitch = 60
set i_pitch = 45
set d_pitch = 60
set p_roll = 45
set i_roll = 20
set d_roll = 15
set p_yaw = 40
set i_yaw = 20
set d_yaw = 60
@LitterBug Cool idea to try stabilization on only select axis. The pitch is by far most difficult to control in acro mode. It’s doable but very intense. Would be cool to try just having that in stabilized mode to see how the copter feels.14 August, 2018 at 21:55 #58800
Been working on a second EZbi with my spare RCExplorer tilts, Baby PDB, and F3FC Racing. Trying to get it in the air without chopping any wires so I don’t have to extend anything once the RCE kit gets here. I’ll test dRonin with the F3FC on this build until the Kits arrive. :-D… Working with @jihlein to figure out the best PWM mapping that also allows for dual servo feedback.
LitterBug15 August, 2018 at 16:47 #58847
How about PA7 or PA4? That’s the same that’s used for the current and voltage sensor. Should be good not having to start another ADC.
We could move the motor output to PB1 or something instead.15 August, 2018 at 17:15 #5884815 August, 2018 at 18:43 #58852
The way the dRonin software is setup, thru the GCS I can already turn off the current sensing function, and use the current sensing pin as the triflight feedback pin, so there’s no changes required to get a 2nd feedback pin on the F3Fc racing target.
That said, I’ve not come up with any cool ways to use the feedback yet. Got some other things I want to try first…..
16 August, 2018 at 10:48 #58860
- This reply was modified 4 years, 5 months ago by jihlein.
I switched the props on mine to the super light 5x4x3 HQ props (the ones in the store) And it made a huuuuuuge difference. It now flies awesome! It only wigs out when you fly it hard backwards now! Much much easier to fly now. Freaking awesome!
Also it would be great to still have the current feedback. I use it all the time on all my models to keep track of the battery. My batteries vary grately from very saggy to very high C and it’s nice to know if the battery is running out or just sagging.
I’ll try to see if I can compile betaflight with the pins swapped around.
16 August, 2018 at 12:02 #58862Kevin_ErikModerator
- This reply was modified 4 years, 5 months ago by RCExplorer - David.
Those 5x4x3 HQ props have more of a blunted end. I’ve found that the “Trapezoidal” or blunted style of props are better for stability and work best with Ducted vehicles. Granted they’re not alway the most efficient but they work.
My favorites that I’ve flown.
* HQ 5×4.5×3 (no longer carried on RCE, Fiberglass)
* HQ DP-5X4.5X3V3-PC ( Polycarbonate )
* DAL T5044 Freestyle props ( Polycarbonate, my current propeller )
Note: Even the DAL 5045’s are great so long as your not using the “C” variant.16 August, 2018 at 15:46 #58863
FWIW, I’m putting together a build of dRonin that allows the user to choose F3Fc Racing Pad R2 to be either the Uart2 Rx input or the 4th ADC input for second servo feedback. Uart2 Rx is the default setting.17 August, 2018 at 18:19 #58873
@jihlein Awesome! Any particular reason you chose Uart2 rx?
Just trying to understand the thinking behind it 🙂17 August, 2018 at 20:50 #58874
Combination of what pins could support the ADC function and also what would be the least disruptive. With S.Port telem on UART2 TX, kind of leaves the RX pin hanging, so figured why not put it there. Not real scientific…..
That said, can’t get the code to work – yet. Think I forgot to initialize the interrupt handler, will debug more tonight.17 August, 2018 at 21:20 #58875
I have EZbi #2 pictured above all wired up, configured with @jilein’s latest F3FClite branch, tested the mix, and ready to maiden. Just need to get rid of all this rain. Looks like Sunday at the earliest…
Pretty much configured the same as EZbi #1, but with more stock RCExplorer parts. 😀
Baby PDB @6V for servos
F3FC Racing FC
Current/voltage monitoring configured
PWM1 Right motor
PWM2 Left Motor
PWM3 Right Servo
PWM4 Left servo
UART1 RX -> R-XSR SBUS
UART3 TX -> R-XSR S.Port telemetry
Left UART2 open in case we use it for feedback at a later date.
Will move all the electronics over when my RCE BiCopter kit gets here.
EDIT: So the sun popped out while I was grillin’ up dinner which made me a bit antsy to give it a maiden. Threw caution to the wind and put it immediately in the dirt. Forgot to adjust the board orientation in the config for 90 degree rotation in anticipation of the RCE frame. Stick mix was right, Gyro mix was off 90 degrees, Which means I have some cleaning to do before the next dry spell. LOL
21 August, 2018 at 02:55 #58908
- This reply was modified 4 years, 5 months ago by LitterBug.
I’ve got the 2nd fdbk pin working on the F3FC R2 pad in dRonin.
- The forum ‘Everything about the Bicopter’ is closed to new topics and replies.