6S Tri-Baby build (Matek F405-AIO)

[Kevin_Erik]

Removed the FC and cleaned up the solder connections. Scrubbed the FC with Isopropyl Alcohol and a Toothbrush to remove any possible FOD. Thus far it looks like the card is now working again so I’ll attempt to reinstall it barring any other failures.

[biggestRCEfan]

Kevin, I spray my fc boards with laquer spray to isolate them against moisture or other fod causing shorts. Worked well for me so far. In the naze 32 days I lost a few boards due to shorts.

[Kevin_Erik]

Whats the spray called ?

[biggestRCEfan]

Bought at Jaycar in Australia

Attachments:

1. 

   IMG_20180514_075600-921x1842.jpg
   

[Kevin_Erik]

Having a hard time finding a source for this in Germany. Will have to use Electrical tape over all the unused parts for now.

[Kevin_Erik]

Link to last 6S flight and logs: https://drive.google.com/file/d/1C3KHoYxbkcHc70qTXyXGoWpFh-YqVg_f/view?usp=sharing

While the tail is stable, the propwash is pretty bad.

[Kevin_Erik]

After today I’m done with the Baby-tricopter!
Had a Servo and Motor die simultaneously while trying to hover. Didn’t even make it off the ground!(WTF)

Burned up far too much equipment in the hopes that Baby could be made 4S tunable much less 6S. Sure at 4S with 2300kv motors its flyable but any hopes of getting much more from the platform is pointless. IMO the firmware just hasn’t matured to the point where it can reliably handle a performance minded build. Maybe iNav and dRonin can fair better in some situations. However, far as I’m concerned I’m better off salvaging my gear at this point to build a standard quadcopter.

Side note: The experienced members of our community are all but absent and any progress seen in the last year feels absent in this one.

Edit: Traced the fault to the feedback wire coming undone and the copter freaking out as if the Motor had just started to Desync. The Servo however is very much burned.

[Kevin_Erik]

Ordered the Seppuku FC in the hopes that this will be viable where Triflight has not been.
Expect 3-5 weeks for delivery.

Note: The servo saver i ordered got canceled without warning. The vendor couldn’t even be bothered to contact me after the fact.

[Gajohnson]

Kevin,
Thanks for all the great work. I’ve been meaning to post for a while but have been busy. I have my BabyTri setup with an Matek F405, FlySky x6B reciever, and a Matek VTX-HV. Otherwise, I’m using the stock motors, servo (though I had to replace it), ESCs, and 4s batteries. Overall, its working well but I’m having trouble with yaw control. Sometime on a sudden throttle change, the copter will rotate counterclockwise without any yaw input from me. I also have trouble with right turns. Inspecting the copter on the bench, it looks like I get full servo deflection to the right (allowing a CCW yaw), but only get 10 degrees of servo deflection to the left (for a CW yaw). I did the tailtune procedure with the little angled block (setting the servo to 40 degrees deflection in either direction). Any idea what would be causing the non-symmetrical servo deflection?

Here is my diff statement:

# diff

# version
# Triflight on Betaflight 3.2.2 v2 / MATEKF405 (MKF4) 3.2.2 Dec 31 2017 / 15:24:22 (30f68c8) MSP API: 1.36

# name
name GAJ-Tricopter

# resources
resource SERVO 1 A08
resource ESCSERIAL 1 D02

# mixer

# servo
servo 5 1051 1927 1550 100 -1

# servo mix

# feature
feature MOTOR_STOP
feature LED_STRIP
feature AIRMODE
feature ESC_SENSOR
feature ANTI_GRAVITY
feature DYNAMIC_FILTER

# beeper

# map

# serial
serial 3 4096 115200 57600 0 115200
serial 4 1024 115200 57600 0 115200

# led
led 0 2,5::AT:0
led 1 3,5::AT:0
led 2 4,5::AT:0
led 3 5,5::AT:0
led 4 6,5::AT:0
led 5 7,5::AT:0
led 6 8,5::AT:0
led 7 9,5::AT:0
led 8 10,5::AT:0

# color

# mode_color
mode_color 6 1 2
mode_color 7 0 4

# aux
aux 0 0 0 1675 2075
aux 1 1 3 1675 2075
aux 2 2 2 1300 1700
aux 3 13 4 1300 2075
aux 4 37 2 1675 2075

# adjrange

# rxrange

# vtx

# rxfail

# master
set gyro_lowpass_hz = 100
set gyro_notch1_hz = 0
set gyro_notch2_hz = 0
set serialrx_provider = IBUS
set dshot_idle_value = 400
set motor_pwm_protocol = DSHOT1200
set failsafe_off_delay = 30
set failsafe_throttle = 1100
set failsafe_kill_switch = ON
set failsafe_procedure = AUTO-LAND
set align_board_yaw = 90
set bat_capacity = 1000
set vbat_max_cell_voltage = 44
set vbat_min_cell_voltage = 34
set vbat_warning_cell_voltage = 36
set ibata_scale = 179
set servo_center_pulse = 1520
set servo_pwm_rate = 250
set tri_tail_motor_thrustfactor = 42
set tri_tail_servo_speed = 239
set tri_servo_feedback = RSSI
set tri_dynamic_yaw_boost = 300
set deadband = 4
set yaw_deadband = 5
set pid_process_denom = 1
set osd_rssi_pos = 40
set osd_tim_1_pos = 2134
set osd_tim_2_pos = 2113
set osd_flymode_pos = 2413
set osd_throttle_pos = 2177
set osd_vtx_channel_pos = 377
set osd_current_pos = 2305
set osd_mah_drawn_pos = 2403
set osd_craft_name_pos = 362
set osd_gps_speed_pos = 218
set osd_gps_lon_pos = 82
set osd_gps_lat_pos = 65
set osd_gps_sats_pos = 51
set osd_home_dir_pos = 302
set osd_home_dist_pos = 303
set osd_compass_bar_pos = 266
set osd_pid_roll_pos = 423
set osd_pid_pitch_pos = 455
set osd_pid_yaw_pos = 487
set osd_debug_pos = 1
set osd_power_pos = 2273
set osd_pidrate_profile_pos = 345
set osd_warnings_pos = 2313
set osd_avg_cell_voltage_pos = 76
set osd_pit_ang_pos = 161
set osd_rol_ang_pos = 193
set osd_nheading_pos = 311
set osd_esc_tmp_pos = 2230
set osd_esc_rpm_pos = 215

# profile
profile 0

set dterm_lowpass_type = PT1
set dterm_lowpass = 110
set anti_gravity_gain = 3000
set p_yaw = 35
set i_yaw = 10
set d_yaw = 30

# rateprofile
rateprofile 0

set rc_expo = 80
set rc_expo_yaw = 30
set roll_srate = 80
set pitch_srate = 80
set yaw_srate = 80

#

Thanks for the help! I plan to keep testing what you share on here as I have found it helpful so far.

Thanks,
Gabe

[Kevin_Erik]

Gajohnson,

You have two options here…

Increase the Yaw I-Term or increase your Anti-Gravity settings. Beyond this, Andy-RC’s PID’s with moderate filtering appears to give the best results. (Confirmed this on my build) However, with me running 1900kv motors @ 6S, this tends to be far too powerful for how Triflight manages Yaw / Torque compensation. I ended up seeing allot of Pitch response to Yaw and vise versa.

At a quick glance i do see that your Yaw I-term is a bit low. Try Andy RC PID’s and shoot me your BF (Betaflight) log. Need one flight of at least a minute in lenght. Perform 10 Yaw, Pitch and Roll actions (Wiggle the sticks hard), then fly around fast and maybe do a few punchouts too.

Roll 45, 47, 31
Pitch 44, 74, 35
Yaw 54, 36, 20

With this I can then use that Log to create a Plasmatree plot for you. From that you can see if you need to adjust values up or down.

Beware that too much filtering will add delay and cause your Yaw to appear sluggish. Too little and you will start burning up Servo’s. Furthermore, Yaw and Servo filters should be avoided for the same reason. Far as i can see, your filter setup “looks” fine. So long as your motors don’t get hot but only warm to the touch, its ok. Unfortunately, you won’t know how your servo is doing, until it suddenly dies without warning. When this happens your copter could do anything from spin uncontrollably, to launching itself end over end as it attempts to correct for the lack of Yaw-feedback signal. While there is code in place to help prevent this, it can still sometimes happen. (had it happen to me twice)

Edit: Too tall of a build can make your Center of Gravity (CG) off center and your FC will have a difficult time compensating. Due to the way the Copter is designed, both Yaw and Pitch appear to come from the same motor. Thus if your CG is off or your battery slides back during flight, expect to see oscillations and / or tail-wag. Sticking to batteries that are 3S – 4S, 1000 – 1300 mAh, 45C helps as too much weight will screw up your CG. Idealy, you want you copter to be around 500 grams or less if possible.

[Gajohnson]

Thanks for the help Erik. I adjusted the values as you suggested but didn’t notice a significant change in flight characteristics. I was still having trouble with the yaw and ended up messing up my servo so I’ll be down for a couple of days. Here are the two logfiles I was able to capture.

My motors do get fairly hot, I would say beyond the point of warm to the touch. What should I change on my filtering?

I looked up the Plasmatree software. I’m not quite sure what I’m looking at yet, but that is a pretty cool software!

Thanks again for the help!

-Gabe

[Gajohnson]

Oops, the log file was over the size limit. Here is a google drive download: https://drive.google.com/open?id=1lcJeKStwUo3HZajLxj0V5mKsVWHSupnE

[Kevin_Erik]

Motors that are warm to the touch are OK. More than that and you should consider defaulting your filters and then stepping back a little at a time.

[Kevin_Erik]

Did a few “fun” flights late last week.
Here is one of my better ones:

Note: While the Camera mount shows in the video, it doesn’t show on the FPV feed.

[Kevin_Erik]

Updated the Tribaby to the following.

dRonin version 20190605-40bb068c
Matek F405-STD
Matek FCHUB-W
Runcam Swift 2 with TX200 transmitter (Smart Audio)

Attachments:

1. 

   Autotune-6-24-2019.jpg
   

2. 

   Roll-6-24-2019.jpg
   

3. 

   Pitch-6-24-2019.jpg
   

4. 

   Yaw-6-24-2019.jpg
   

5. 

   Wizard-6-24-2019.jpg
   

6. 

   Stab-Tab-6-24-2019.jpg
   

7. 

   System-6-24-2019.jpg
   

[Author]

Posts