New type of yaw control

[fomoco_44]

@litterbug,
Wow, the specs for the BMS-A207 look pretty awesome!….I’ll be interested to see how that goes for your build.
I think I’m going to try the TGY-313C and PDI-2004 servos and see if their increased speed makes a noticeable difference.
I’m currently trying the V7 build on my Mini Tri with F3FC integrated(DSHOT 600, 6v servo, stock hardware) with some pretty good results. Just need to get a good set of PID’s for it, as all the settings so far are for the Baby Tri!

[Chip]

Regardless of the responses I get I want to give a huge thank you to Lauka and the rest of the community that is actively involved in making these crazy contraptions fly better.
I am building a Baby Tri using the Betaflight F3 FC but am having difficulty figuring out the setup.
I have everything hooked up and working with triflight except the tilt servo. I used these commands I found to move M1 to the PPM pin and add the feedback wire to M1 on the fc.

set tri_servo_feedback = EXT1
resource PPM 1 NONE
resource MOTOR 1 B07

My question is where to solder the servo signal wire and will I have to try to remap any other resources? I tried moving motor 3 to M4 and never could get it to work. I am a handy guy but this is over my head.

Thanks Guys!
-Chip

P.S. Here is my diff?
# diff
# Triflight 0.7 Beta 2 / BETAFLIGHTF3 3.1.7 Jun 4 2017 / 13:24:55 (1698ba5)
resource MOTOR 1 B07
resource PPM 1 NONE
servo 0 1000 2000 1500 90 90 100 14
servo 5 1000 2000 1500 40 40 100 -1

feature MOTOR_STOP
map TAER1234
aux 0 0 0 1600 2100
aux 1 28 1 1400 1600
aux 2 4 1 1600 2100
aux 3 13 2 1300 2100
aux 4 31 3 1300 1700
set min_throttle = 1000
set min_command = 973
set tri_servo_feedback = EXT1
profile 0

Attachments:

1. 

   bff3-layout.png
   

[bhuism]

@Chip, see https://github.com/lkaino/Triflight/wiki/BFF3-Feedback-wire

Can u use pwm1?

[Chip]

Bhuism
I am using pwm1 for the feedback. I have pwm4 open so if I use that for the servo would I say “Resource set servo1 pwm4”?
I really struggle with the code side of this…

[bhuism]

I use:

resource SERVO 1 B09

B09 is PWM output 4 (motor 4)

So solder servo signal (to PWM4)

please read:

https://github.com/betaflight/betaflight/wiki/Betaflight-resource-remapping

cheers!

[LitterBug]

@lkaino,
Finally made it out to the field tonight and conditions were perfect with no wind whatsoever. Took both Baby Tris and TriFlight0.7Beta2NewYawControlV7. Unfortunately forgot to take my servo adjustment tool so not sure about the validity of my ground tailtune. Was eyeballing it, and my servo speed was up from where it was previously. Maybe I am getting it broken in, or maybe my angles were too low. At any rate, I did get two flights in with Baby Tri #1 which has the fast TGY-211DMH servo, and there did not seem to be the low speed oscillation that has plagued me in the past. It did do some odd things, but that could all be on the sketchy ground tailtune. Also Think I should have recalibrated the ESCs. Appears I had been playing with oneshot125 the last time out and I switched back to multishot, but didn’t recalibrate. That might explain some of the unexpected or erratic behavior, especially at low throttle and motor idle speed were erattic at 0 throttle. I redid the ground and in-air tailtunes between packs. First flight I had only eyeballed the angles and had 454 tail servo speed and 60 for the thrustfactor. Seemed to fly ok, tail oscillations on arm, none in flight, but there were some odd yaw twitches here and there that were uncommanded. Second flight, I set the servo end points to what I had saved from a previous diff. Servo speed = 449 and thrust factor of 45. Did not seem to fly quite as good, but still did not have any oscillations other than on arm.

Baby tri 2 didn’t go so well. Appears I have a Vbat or 5V power issue to the F3FC and am going to have to tear into the FC stack. Got through the ground tailtune, and one failed in air tailtune. Landed, rebooted the tri, and made a second attempt. Was going really good and got the successful tones, but when I started to land, the FC rebooted. DOH! Tried pulling the battery and plugging back in, PDB was lighting up, but was not always getting LEDs on the FC. 🙁 Hopefully just a loose connection or a piece of grass in the wrong place. Hope to be able to give them another go tomorrow. Not going to bother posting up Video or BB logs till I get accurate ground tail and ESCs recalibrated.

Cheers!
LitterBug

EDIT: running mostly stock config without adjusting PIDs or any other config.

[biggestRCEfan]

Litterbug, the behaviour you describe with baby tri #2 is similar to what I have experienced with moisture getting into the FC. After landing on wet grass FC would miss behave, solved by putting a fan on a it overnight.

Now with all my builds, I spray boards post assembly and test with a PC board lacquer spray to seal the components from see or wet grass causing short circuits.

[LitterBug]

@lkaino,
Made it out to the field a few times over the weekend with my rebuilt Baby Tri #1. Having some really bad yaw twitches that make it really uncomfortable to fly. I am going to replace the tail pivot mechanism that I reused when I rebuilt this frame. I had ordered a replacement before FliteFest, but rebuilt it using the old hardware to get it in the air quicker. I think there may be two issues going on here contributing to my yaw issues on this frame. 1) There is some slop between the servo and the motor plate. 2) There is some stick in the pivot that makes the slop hold it’s position when the servo moves. It’s most noticeable when doing the ground tailtune. If I move the servo from the right to center, it will have some offset to the right of center. If I move it from the left to center, it will have some offset to the left of center. I think the combination of slop and stick may be causing a good deal of the problems. Still need to debug my FC power issues on BabyTri #2 before testing any more with that frame. Making slow progress on the Mini Tri and HK Y-3 builds too.

Cheers!
LitterBug

EDIT: Also working with @JIhlein on the DTFc feedback mentioned previously.

[LitterBug]

Replaced tail pivot and bolt on Baby Tri #1 with the one I ordered. Servo is smooth a silk now, no more binding. Also have consistent end and center points regardless of movement speed, distance, and direction. The pivot bolt was definitely bent from when the tail got demolished previously.

Crappy wires I had used on the Logger on Baby Tri #2 had frayed, broken, and were causing an intermittent short. So looks like I just need to find time to hit the field to finally get some good V7 testing in with both frames 🙂

Gotta love it when it’s the Stupid Simple Stuff….. LOL

Cheers!
LitterBug

[LitterBug]

Been working on the HobbyKing Y-3 tricopter build with a DTFc FC. (doge) Added servo feedback to the Turnigy TGY-S306G servo that it came with and hooked that to the LED pin on the DTFc FC. (there is no RSSI pin) @JIhlein generated a version of TriFlight Beta2 NYC V7 with RSSI mapped to the LED pin. Gave it a test tonight and it was able to read the servo feedback and calculate servo speed! Speed calculated is not correct, but that is due to the limited amount of servo travel, and I have not yet measured and adjusted the angle in the TriFlight Parameters. Making Progress! Will be using the same FC on my Mini Tricopter build as well. Working parameters and resource mapping so far:

# diff
# Triflight 0.7 Beta 2 - new control test v7 / DOGE 3.1.7 Aug  8 2017 / 15:39:22 (3200a3c)
resource MOTOR 1 A10
resource MOTOR 2 A09
resource MOTOR 3 A00
resource MOTOR 4 NONE
resource MOTOR 5 NONE
resource MOTOR 6 NONE
resource MOTOR 7 NONE
resource MOTOR 8 NONE
resource SERVO 1 B08
servo 5 1226 2045 1612 40 40 100 -1

feature -RX_PPM
feature RX_SERIAL
feature CURRENT_METER
map TAER1234
serial 2 64 115200 57600 0 115200
aux 0 0 1 1450 2050
aux 1 28 1 1450 2050
aux 2 1 0 950 1050
aux 3 2 0 1450 1550
aux 4 13 2 1950 2050
aux 5 31 2 1450 1550
set min_throttle = 1030
set motor_pwm_protocol = MULTISHOT
set serialrx_provider = SPEK2048
set tri_tail_servo_speed = 559
set tri_servo_feedback = RSSI

Cheers!
LitterBug

[lauka]

@LitterBug: Good to hear you’re getting your copters sorted out :). Would like to see logs from the faster servo.

I have been doing some racing with the baby, comparing to my quad. It flies good as expected, but the delay of the servo is noticeable. Think I will also have to change the servo to a faster one.

Are the @jihlein ‘s changes available somewhere?

I would like to create new beta version of 0.7 soon with the changes from this branch. I just need more test results to see how it is working on different copters.

[bhuism]

@lkaino, I read on the betaflight 3.2 wiki this:

“Improved ITerm windup handling for tricopter”

I’m sure you knew this allready, just checking

btw: I’m considering adding a bug to betaflight for bringing bang yaw-D in the configurator and lua scripts

[lauka]

@bhuism: It’s the same as in Triflight, they copied it over to betaflight.

[bhuism]

Ha, kewl, there was no mention of this, end of the week my vacation starts, then I’ll fly the latest and greatest triflight again, keep u posted

[LitterBug]

@lkaino,
Finally made it out to the field with the new tail on BT1 with the fast servo. Conditions were very calm. Threw the Log file and raw video here:
https://drive.google.com/drive/folders/0B0qy__jJEauHenFyRXlQeFZFQmc?usp=sharing
Running default tuning with minimal changes to any other settings. Tail is still acting odd, hopefully due to no tuning specific to my build. Here’s a diff with my settings:

# diff
# Triflight 0.7 Beta 2 - new control test v7 / RCEXPLORERF3 3.1.7 Jul 30 2017 / 19:46:36 (4d12208)
name BabyTri1
servo 5 1120 1952 1539 40 40 100 -1

feature CURRENT_METER
feature BLACKBOX
map TAER1234
serial 0 1 115200 57600 0 115200
serial 1 64 115200 57600 0 115200
serial 2 128 115200 57600 0 250000
aux 0 0 1 1450 2050
aux 1 28 1 1450 2050
aux 2 1 0 1450 1550
aux 3 2 0 1950 2050
aux 4 13 2 1950 2050
aux 5 26 1 1950 2050
aux 6 31 2 1450 1550
set min_throttle = 1035
set motor_pwm_protocol = MULTISHOT
set serialrx_provider = SPEK2048
set ibat_scale = 300
set tri_tail_motor_thrustfactor = 45
set tri_tail_servo_speed = 386
set blackbox_rate_denom = 6

Let me know if you have any recommendations for changes.
Cheers!
LitterBug

[Author]

Posts