Hello guest, if you read this it means you are not registered. Click here to register in a few simple steps, you will enjoy all features of our Forum.
This forum uses cookies
This forum makes use of cookies to store your login information if you are registered, and your last visit if you are not. Cookies are small text documents stored on your computer; the cookies set by this forum can only be used on this website and pose no security risk. Cookies on this forum also track the specific topics you have read and when you last read them. Please confirm whether you accept or reject these cookies being set.

A cookie will be stored in your browser regardless of choice to prevent you being asked this question again. You will be able to change your cookie settings at any time using the link in the footer.

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
4" racing drone doing death rolls and won't disarm
#1
I am asking because I can't solve this problem yet. The problem is the quad does a sudden and unexpected death roll to the right, stops responding, and won't disarm (happens 50-75% of all flights). I have failsafe on a radio switch and hit it each time as it is eating dirt and not stopping and eventually it stops (scary). I am not even sure if it stops because I hit fail safe. It just stops after eating dirt for a minute (not instantly after hitting fail safe). Pretty consistent in behavior. Scary to pick up and unplug the battery. It totally stops responding, rolls right, and eats dirt for about 10 seconds before stopping (while I try everything to make it stop). Often times it is cart wheeling across the field, but most often upside down grinding sand in the motors.

Only other symptom of something being "off" is an oscillation when I punch the throttle, and sometimes motor 1 is warmer than the rest. Every so often it doesn't punch well at all (feels weak). For example, I dive, and it barely pulls out, even at full throttle, when normally it would no problem. Inconsistent and unpredictable. For the most part it flies pretty well/okay. When it does the death roll it isn't when I punch or do a hard manuever, it is totally unexpected, often while hovering or just cruising.

I rebuilt an older drone with a bunch of old and new parts. It did this the first time around, then I rebuilt it all again, and it is still doing it. I was convinced it was a bad ESC, so I swapped that out for a more modern, much more powerful ESC. Same problem. Changed all 4 motors. Same problem. Changed firmware (from BF 3.5.7 to 4.0 to 4.1.1). Same problem. Changed props. Same problem. No wires over the gyro. No screws touching motor coils. Checked all batteries I fly and resistances are good. I am really hoping it is a Betaflight setting and not the FC or receiver. The motors are new, freshly out of the box. There is basically nothing else it could be besides FC, receiver, or Betaflight settings (????). Really hoping it is software. It is a really clean build, perhaps the cleanest I have ever done. Everything is tight, fits well, nothing loose or wobbly. Video camera and vtx are good and clear although around the time it death rolls it gets sporadic with signal perhaps indicating something.

Seriously Dodo F3 FC (RMDO target)
Betaflight 4.1.1 Joe Lucid EA3 Performance Edition

Skystars K060 60amp ESC
Cobra 2204 2300kv motors
Gemfan Flash 4052-3 Props
Matek ELRS-R24-D receiver running ELRS 3.2
Taranis X9D+ with BetaFPV ELRS 1W Module


Current BlHeli_32 ESC settings (all defaults except):
Motor timing: Auto
PWM Freq: Variable 24-48kHz
Rampup Power: 42%


Current Betaflight DIFF ALL:

Code:
# diff all
# Betaflight / RMDO (RMDO) 4.1.1 Oct 26 2019 / 12:40:34 (83ac98cfb) MSP API: 1.42 / FEATURE CUT LEVEL 6
batch start
defaults nosave
mcu_id 003600224b33571920323135

# name: Tweaker_F3
feature -RX_PPM
feature -AIRMODE
feature RX_SERIAL
beacon RX_LOST
serial 2 64 115200 57600 0 115200
aux 0 0 0 1800 2100 0 0
aux 1 1 2 1800 2100 0 0
aux 2 27 1 900 1200 0 0
aux 3 13 1 1800 2100 0 0
aux 4 26 4 1800 2100 0 0
aux 5 28 2 900 1200 0 0
aux 6 35 5 1800 2100 0 0
aux 7 36 3 1800 2100 0 0
set gyro_lowpass_hz = 300
set gyro_lowpass2_hz = 375
set acc_calibration = -8,26,245
set mag_hardware = NONE
set serialrx_provider = CRSF
set min_throttle = 1000
set motor_pwm_protocol = DSHOT300
set failsafe_switch_mode = STAGE2
set ibata_scale = 290
set yaw_motors_reversed = ON
set name = Tweaker_F3

profile 0
set dterm_lowpass_hz = 105
set dterm_lowpass2_hz = 225
set anti_gravity_gain = 3500
set pidsum_limit = 1000
set pidsum_limit_yaw = 1000
set p_pitch = 69
set d_pitch = 44
set f_pitch = 171
set p_roll = 63
set d_roll = 40
set f_roll = 162
set p_yaw = 68
set f_yaw = 162

profile 1

profile 2
profile 0

rateprofile 0
set thr_mid = 30
set roll_rc_rate = 150
set pitch_rc_rate = 150
set yaw_rc_rate = 105
set roll_expo = 54
set pitch_expo = 54
set yaw_expo = 54
set tpa_breakpoint = 1750

rateprofile 1

rateprofile 2

rateprofile 3

rateprofile 4

rateprofile 5
rateprofile 0
save
#




If there is something else from Betaflight you want to see I am happy to post. Thank you for looking this over. I am feeling a little desperate and frustated. Been working on this for weeks, and it's a miracle I haven't broken more off this quad yet. Thank you.
Reply
Login to remove this ad | Register Here
#2
Can you post a screenshot shot of the modes tab in betaflight?
One with your transmitter on and the switch in the arm position, and another with it in the disarm position?

Can you also post a screenshot shot of the failsafe tab in betaflight?

First things first, we gotta figure out why it doesn’t always disarm, because that is dangerous.

You say you have failsafe on a switch- does it disarm if you toggle your arm switch into the unarm position?
Dangerous operations.

Disclaimer: I don’t know wtf I’m talking about.
I wish I could get the smell of burnt electronics out of my nose.
Reply
#3
1. What is FC core temp doing before and around when it freaks?

2. I assume the FrSky Tx/ELRS Module combination is working with other models?


EDIT ADDED................

3. What power are you running the ELRS module at? (and is the old design X9D having to supply that or are you using a seperate battery module supply?)
Reply
#4
ALSO.............

I know you list a "Performance Edition" of 4.1.1, but since F3 boards without such cut down memory space trickery were not supported by BF 4.1.x onwards, its IMO an unwieldy cludge to get an old board with insufficient memory to work, and frankly I would not do it.
Reply
#5
I know that I advised you to install the unofficial Joe Lucid 4.1.1 performance edition of Betaflight in your BLHeli passthrough thread (HERE) but maybe that was a bad call. The first thing I would do is to flash back to the official Betaflight 4.0.6 or 3.5.7 versions and see if that eliminates your issue. It just means you will lose the bi-directional DSHOT / RPM Filtering capabilities but if it eradicates the issue you are having then that is a small price to pay to avoid a potentially serious incident occurring from an out-of-control quad.

If it was me, I would just consign that F3 FC to the scrap heap and get a more modern F4, F7, or H7 based FC.
[-] The following 1 user Likes SnowLeopardFPV's post:
  • Lemonyleprosy
Reply
#6
(06-Mar-2023, 05:25 AM)Lemonyleprosy Wrote: Can you post a screenshot shot of the modes tab in betaflight?
One with your transmitter on and the switch in the arm position, and another with it in the disarm position?

Can you also post a screenshot shot of the failsafe tab in betaflight?

First things first, we gotta figure out why it doesn’t always disarm, because that is dangerous.

You say you have failsafe on a switch- does it disarm if you toggle your arm switch into the unarm position?

I will post these things later today.

Yes, normally it disarms fine. I always use a prearm, and arming it on and off works, and disarming via failsafe works too. It's only after the death roll it stops responding.
Reply
#7
(06-Mar-2023, 08:40 AM)BadRaven Wrote: 1. What is FC core temp doing before and around when it freaks?

2. I assume the FrSky Tx/ELRS Module combination is working with other models?


EDIT ADDED................

3. What power are you running the ELRS module at? (and is the old design X9D having to supply that or are you using a seperate battery module supply?)

1) Not sure about temp. I have a limited OSD from a TBS Unify Evo vtx. I get vbat and vtx output power. That's basically it.

2) Yes, no issues with other quads.

3) I have tried a variety of power settings. Everything has been at 500hz, but testing power up to dynamic 1W, but seemed unnecessary, so I tried dynamic 100mW, and most recently I turned off dynamic power settings and just use it at static 100mW for simplicity (not doing any long range with this setup, so no point). All had the same result.

Edit: forgot to mention, yes, the Taranis and module are powered by an internal 3000mAh lipo.
Reply
#8
(06-Mar-2023, 09:00 AM)BadRaven Wrote: ALSO.............

I know you list a "Performance Edition" of 4.1.1, but since F3 boards without such cut down memory space trickery were not supported by BF 4.1.x onwards, its IMO an unwieldy cludge to get an old board with insufficient memory to work, and frankly I would not do it.

Interesting.... Well, this problem happened on BF 3.5.7 as well. But having that brought to my attention I will try to learn more about this.
Reply
#9
(06-Mar-2023, 11:49 AM)SnowLeopardFPV Wrote: I know that I advised you to install the unofficial Joe Lucid 4.1.1 performance edition of Betaflight in your BLHeli passthrough thread (HERE) but maybe that was a bad call. The first thing I would do is to flash back to the official Betaflight 4.0.6 or 3.5.7 versions and see if that eliminates your issue. It just means you will lose the bi-directional DSHOT / RPM Filtering capabilities but if it eradicates the issue you are having then that is a small price to pay to avoid a potentially serious incident occurring from an out-of-control quad.

If it was me, I would just consign that F3 FC to the scrap heap and get a more modern F4, F7, or H7 based FC.


I was having the same problem with those firmware versions as well.

I actually haven't even enable bidirectional dshot or rpm filtering yet. I have been taking this one step at a time rather than make a lot of changes at once.

For what it's worth, all my test flights have been in an abandoned community development where there is nothing but sand, weeds, fire ants, and snakes. No risk of anyone walking their dog or jogging by. I like the spot because it has bushes, some forgotten sidewalks, etc, that make for good visual landmarks, but otherwise it is barren land without a soul around.
Reply
#10
(06-Mar-2023, 05:25 AM)Lemonyleprosy Wrote: Can you post a screenshot shot of the modes tab in betaflight?
One with your transmitter on and the switch in the arm position, and another with it in the disarm position?

Can you also post a screenshot shot of the failsafe tab in betaflight?

First things first, we gotta figure out why it doesn’t always disarm, because that is dangerous.

You say you have failsafe on a switch- does it disarm if you toggle your arm switch into the unarm position?


Here we go.... I didn't think I had time this morning, but I made time.... Thanks for your thoughts. Greatly appreciated.


[Image: dxfdkF2l.jpg]
[Image: TUcUgNAl.jpg]
[Image: rgMbNVWl.jpg]



As you can see, there is a Prearm on the opposite side of the Taranis, and both switches go up towards me to arm. Failsafe and beeper are on a switch together. I haven't gotten turtle mode to work yet, but the switch is set. I use air mode and angle mode sometimes, and they share a toggle switch as well. I have been using these exact settings on other quads and used these settings with FrSky before migrating to ELRS, with some subtle adjustments.
Reply
#11
4inch 2204 2300kv does more rpms than it would on a 5inch, therefor it might matches defaults better on 5inch.

i would suggest to increase motortiming and decrease rampup, demag compensation to high. that should match better, i dont know if it could be part of the reason but i guess it should be tried anyway for good performance next to the issue - nothing to loose

wrong startuppower can cause excessive noise to the system - that would be something which would still exclude the fc as main source
[-] The following 1 user Likes hugnosed_bat's post:
  • EyeBotXander
Reply
#12
To simplify things for now, I’d suggest removing failsafe from the modes tab and simply using your arm switch to disarm in an emergency. This won’t stop failsafe from triggering if you lose rc link, it just won’t allow you to manually trigger failsafe from your transmitter.

It’s hard to tell from your screenshots, but it looks like your cpu load is at 65% when it’s just sitting there idle.

I think the death roll and failure to respond to commands might be because you’re overloading your cpu.
I don’t know what priority failsafe via switch is, but disarm should be a high priority task and hopefully take precedence over other tasks when the cpu is overloaded, hopefully allowing you to *actually* disarm after a death roll.

Can you please run the tasks command in the cli tab and either paste the output here or post a screenshot?

I’m thinking you might need to reduce that cpu load by turning off some things or setting things lower.

Edit: With betaflight 4.3+, we can go up to 75% cpu load safely, because they changed how cpu load was calculated. However, with earlier versions (4.2 and below), max cpu load needed to be significantly lower. I think it was something like 50-ish% max, with 30% being ideal. Anyone remember the exact numbers?
Dangerous operations.

Disclaimer: I don’t know wtf I’m talking about.
I wish I could get the smell of burnt electronics out of my nose.
[-] The following 1 user Likes Lemonyleprosy's post:
  • EyeBotXander
Reply
#13
(06-Mar-2023, 02:00 PM)hugnosed_bat Wrote: 4inch 2204 2300kv does more rpms than it would on a 5inch, therefor it might matches defaults better on 5inch.

i would suggest to increase motortiming and decrease rampup, demag compensation to high. that should match better, i dont know if it could be part of the reason but i guess it should be tried anyway for good performance next to the issue - nothing to loose

wrong startuppower can cause excessive noise to the system - that would be something which would still exclude the fc as main source


You think I should increase motor timing higher than 23 degrees (medium high)? Or is that a good next step to try?

Basically all of the BlHeli_32 settings I have chosen were based on Chris Rosser videos. According to what I gleaned from him, demag compensation should be increased if there are desyncs. Do you think that is what is going on? He says high is good for large quads.

Rampup power, he recommended 50% for 3 inch, and 25-30% for 5 inch.... So you're thinking closer to 30% is better for this one?

You think variable PWM with min 24khz and max 48khz is fine?
Reply
#14
(06-Mar-2023, 08:14 PM)Lemonyleprosy Wrote: To simplify things for now, I’d suggest removing failsafe from the modes tab and simply using your arm switch to disarm in an emergency.

It’s hard to tell from your screenshots, but it looks like your cpu load is at 65% when it’s just sitting there idle.

I think the death roll and failure to respond to commands might be because you’re overloading your cpu.
I don’t know what priority failsafe via switch is, but disarm should be a high priority task and hopefully take precedence over other tasks when the cpu is overloaded, hopefully allowing you to *actually* disarm after a death roll.

Can you please run the tasks command in the cli tab and either paste the output here or post a screenshot?

I’m thinking you might need to reduce that cpu load by turning off some things or setting things lower.

Edit: With betaflight 4.3+, we can go up to 75% cpu load safely, because they changed how cpu load was calculated. However, with earlier versions (4.2 and below), max cpu load needed to be significantly lower. I think it was something like 50-ish% max, with 30% being ideal. Anyone remember the exact numbers?

Ok, I will remove failsafe to simplify.

I don't know why it was so high when I plugged it in this morning. It has been sitting around 50% with the latest round of settings. This morning I did it quickly and it did catch my eye, but the number surprises me. If I lower the pid loop and gyro frequencies it goes down to 25-30%, (which I haven't really been able to tell a difference on anyways). I can probably even try Dshot 150 if I needed to. Pawel convinced me most people can't tell the difference between ESC protocols ever since Oneshot with active braking become mainstay, and I believe it.

Will run tasks in a moment..... Thanks for the info and help again.
[-] The following 1 user Likes EyeBotXander's post:
  • Lemonyleprosy
Reply
#15
This one is tasks command before making changes (Gryo/PID Loop both set to 2.67kHz & CPU load was reading ~50%):

[Image: 7oBLsoil.jpg]


This one is tasks command after lowering PID loop to 1.33kHz, and CPU load is now reading 26-30%:

[Image: VYkDVDml.jpg]




ESC protocol is still set to DSHOT600. Haven't touched that yet.
Reply


Possibly Related Threads...
Thread Author Replies Views Last Post
  Build Helios - 3.5” Sub-250g (Sub 160g possible) DJI O3 Drone – Frame/Build Thread Recursion 13 2,065 14-May-2024, 09:55 AM
Last Post: FloppyFish
  Build Aerospace Engineer + 5" Drone = EnzoDrone BigBeard 3 215 22-Mar-2024, 04:33 AM
Last Post: Bud13420
  Can anyone recommend a mini drone? gabrilesi95 9 920 05-Dec-2023, 12:32 AM
Last Post: voodoo614
  Build I've decided to build a death machine skamtastik 20 1,019 04-Jul-2023, 03:04 AM
Last Post: skamtastik
  Help my drone crashed today ... adirb 18 2,066 29-Jun-2023, 01:40 AM
Last Post: FPVme


Login to remove this ad | Register Here