Triflight 0.7 Beta 3

[juanquy]

Hi Kevin:
after setting it to 2/2, CPU load is now 5%, huge diff. I will try your advice about yaw settings tomorrow on the field to see how it behave.
question: on my old FC ( integrate naze for the mini tricopter) pid’s are perfect, what about if I copy and paste those PIDs on this new F3, it will work or not?, maybe not I guess.

Thank you
regards
JC

[hgg2k]

Hi guys,

I have finished building the Quanum Trifecta with the F3 Deluxe board.
In beta 2 I had a problem with the yaw servo turning left every time I lowered throttle,
but in beta 3 it has been 99% solved! Great work.

After the update to beta 3 I have another problem. When the servo is centered it starts
oscillating badly, but when I start turning it to the left or the right, the oscillation
goes away. During flight the problem is still there but to a lesser degree.
I have tried tuning the yaw PIDs but they had no effect.

Small clip showing the problem : https://youtu.be/ferMhxt82DA

I am using the EMAX ES09md digital servo but I have not used any feedback wire.
Do you think that this might help to solve the problem? Any other suggestions?
Is this a bug of the beta 3 release? (I’ve seen another post mentioning the same)

Thank you.

[Kevin_Erik]

@hgg2k

Please post a copy of your DIFF and possible a Betaflight log if possible.

[hgg2k]

Here is the DIFF :

# diff
# Triflight 0.7 Beta 3 / SPRACINGF3 3.1.7 Oct 21 2017 / 14:04:43 (17fa321)
name Trifecta
servo 5 965 1813 1374 40 40 100 -1

feature -RX_PPM
feature -RSSI_ADC
feature -TRANSPONDER
feature RX_SERIAL
serial 2 64 115200 57600 0 115200
aux 0 0 2 1700 2100
aux 1 1 1 1300 1700
aux 2 13 0 1700 2100
aux 3 31 3 1700 2100
set min_check = 1050
set digital_idle_percent = 3.000
set motor_pwm_protocol = DSHOT300
set serialrx_provider = IBUS
set vbat_scale = 210
set ibat_scale = 380
set gyro_sync_denom = 2
set deadband = 4
set yaw_deadband = 4
set baro_hardware = AUTO
set mag_hardware = AUTO
profile 0

rateprofile 0
rateprofile 0

I will try to post a black box log file as well.

[hgg2k]

Ok, below is the Blackbox log as well :

https://ufile.io/5fttg

Thanks.

(edited) I also have to mention two things I’ve noticed.
First is that in the black box log you can see that every time that
the yaw oscillations quite down the tricopter model shows yaw to the
right, but during the whole test I yawed both to left and to the right.

Second thing is that when I did the disarmed tail tune I moved the
stick to the left, to the right, center and then when I moved it down
for the speed calibration, the servo moved left and then to the middle.
It did not move to the right as well as I see in many videos.

I do not know if the above are relevant.

[Kevin_Erik]

Without the Feedback wire, the speed calibration is not possible. That’s why it did not complete when you pushed the stick down. Also i need the DIFF report from your Tricopter to be able to see the current PID and filter settings. (sorry didn’t see that it was on a prior post)

Question: Are your motors / FC soft mounted? (Seeing a lot of noise from them in BF Explorer)

Edit: What i did see however is that you have a small noise spike at 80hz. You could maybe lower your Gyro Low Pass Filter to 70hz. This wont prevent the wag but may make it easier to tune / fly.

Suggestion: Turn off things like Anti-Gravity, TPA or anything else that can alter your PID values while you are tuning. Otherwise your values will jump all over the place and you’ll not have a clue as to what’s going on.

[hgg2k]

Motors are not soft mounted although the Trifecta is made from a
relatively soft plastic. The flight controller though is soft
mounted on a very soft foam. Maybe soft mounted too much..

How do I create a DIFF command for only the PIDs/Filters?
Do you want me to post a screen capture of them?

[Kevin_Erik]

Did a Plasmatree for the log you just sent.

Looks like your not even able to get it off the ground due to the wagging. How does it respond (wag) when throttle is applied?

[hgg2k]

That’s the strange thing.
It gets off the ground and flies pretty stable with the yaw oscillation
to a minimum when the yaw is centered. When you punch out the throttle
is responds normally with just a small yaw turn if you lower the throttle
abruptly. Strange. Can it be a bad servo?

[Kevin_Erik]

For the Servo, look in your documentation for it as see what PWM frequency it supports. Odds are it should be good to 240-250hz. Hence set the servo_pwm_rate to 240 or 250 hz. Any Higher and it’s not going to respond to a faster input. Too low and the Servo may feel sluggish.

Can I assume your using a dedicated 5-6 vdc source for the Servo?

Side note: it’s possible to add your own feedback wire as I did here https://rcexplorer.se/forums/topic/mod-feedback-wire-for-bluebird-bms-385max/

Beyond all of that, try setting your Yaw, Pitch and Roll I-terms to “1”. Set the P-Terms to 20 and the D to 5. Then work on your Roll, Pitch and Yaw. (in that order) Start with the P-Term until you see a strong oscillation that diminishes over time. (Then back off a little) Then add D-Term to sharpen the Response. Past that you’ll want to increase / decrease P & D together so they stay relative to each over in magnitude. Finally add I-Term back so that the Copter remains at the orientation your left it.

P is mostly for gain and affects the hear and now. Too much causes overshoot and oscillations.
I is for stability and attempts to resist change. Too much causes low freq oscillations.
D is like a dampening spring (rough analogy). Too much can add noise into the entire system and overheat your motors.

Have a look at this: https://www.rcgroups.com/forums/showthread.php?2953533-Betaflight-PID-Analyzer

[Kevin_Erik]

“Can it be a bad servo?”

Doubt it if it’s new. Odds are that the Yaw limits of the Servo may of not been saved and would need to be entered manually. (40 degrees either side of center) Make sure that when you do the Tail tune (Left, Right and center Stick) that you manually “save” via your Taranis stick command. (left stick goes down and left, right stick goes down and right)

Also be sure that there is no slack in the servo and that the servo is not binding in some way. Event the motor wires can cause issues if they are too short or too long.

[hgg2k]

Servo is set to 300hz. No info on specs, but since its a digital servo,
it might support this PWM rate. (Just tried 250Hz and nothing changed)

I am using a regulated 5.0V to power the servo from the BeeRotor v2 PDB.

I think I will try to add the feedback wire.

I will try your suggestions and get back.
Thank you for your time and help!

[Kevin_Erik]

No worries, have fun!

[hgg2k]

The servo limits were saved as shown in the DIFF command
“servo 5 965 1813 1374 40 40 100 -1”

I have also checked for motor wires, servo backlash and binding and
they are all fine. Is it possible that this release does not like
servos without the feedback wire? I will try to do the feedback wire
mod and use the RSSI pad.

[hgg2k]

I’ve just finished the feedback wire mod to the EMAX ES09MG servo.
I have entered the “set tri_servo_feedback = RSSI” command and I
have soldered the feedback wire to the RSSI pad on the bottom of
the board. (SPRacing F3 Deluxe Banggood clone). I have also
enabled in betaflight configuration the analog RSSI option.

When I move the servo, the feedback pin goes from 0V to 2.5V
with respect to ground so it seems that the mod was successful.

The problem now is that the flight controller does not arm at all!
It gives a fast flashing (red led) and beeping!

I read at the beta 3 new features the following :

“Arming prevention in case of invalid servo feedback value. Arming of the copter is prevented in case the servo feedback is configured but the value is outside of calibrated range. This might happen e.g. if the feedback cable breaks. If the feedback is lost while armed, the yaw boost (motor output control based on servo deflection) is disabled and the yaw output becomes sloppier.”

..but the feedback servo has a stable voltage both when armed and disarmed.

Any ideas?

[Author]

Posts