Quad Freaking Out During Flips and Rolls

2

Okay, I’m dropping this here in hopes that someone has seen this before. Yesterday I smoked a brand new QQ190 Tank frame on its first flight since the rebuild.

While doing flips or rolls where my copter will stop sending stick commands to the copter and lock into a high-speed.  From what I have been able to determine this only happens if I perform a flip or a barrel roll. I have been able to fly race style with fast acceleration and hard turns, but once I go inverted all hell breaks loose.

This (was) my QQ190 Tank. I was not flying video yesterday when it happened for the last time, but this is identical. VIDEO

This has happened during my last two flight sessions. The first time this occurred I added a new profile of slightly adjusted PIDs/Rates–after making sure the profile was stable in horizontal flight I tried acro flying and my first 360º roll never ended. The crash was light so I dusted things off and went back to my original profile and continued to fly. Back on my default profile things appeared normal until it happened again this time there was frame damage and we were done.

I rebuilt the copter(steps below) and on my first battery after shaking things out she went down again in the exact same way.

 

To be clear flips and rolls will cause the craft to spin out of control on the roll axis.

Trouble Shooting:

Checked all hardware for physical damage. All wiring was meter tested for common, +volts, shorts, opens. ESCs were bench checked and calibrated all checked out. FC was wiped and reprogrammed from scratch with the latest stable TBS build and PIDs were left stock. I purchased a QQ190 Tank V2 base and reinstalled everything and checked everything again.

During the freakout, my VTX, camera, LEDs still have power and are functioning and my RX is accepting disarm and failsafe commands.

Unfortunately, the build does not have a black box so the video above is all I have.

 

SPECS:

  • Frame: QQ190 (Tank)
  • FC/PDB/VTX: TBS Powercube Elite Colibri RACE v2
  • ESC 20A Favourite Little Bee
  • RunCam Swift Mini
  • Lumenier SKITZO 2500KV
  • BetaFlight (Current TBS stable build)
  • Flight mode: Acro Airmode (only)

Sorry guys, not sure how this posted anonymously, but it’s my question.

1
Anonymous 1 Comment

Hi Jason,

Sorry to hear about the troubles with your quad. I was looking at your CLI dump and nothing stands out as a fatal in flight issue. The values and rates all line up.

I would try to flash the ESC firmware to oneshot and give it a test flight to really rule out the ESC control.

The Little Bees are a little clunky and don’t have as smooth as a resolution for control as the Bulletproofs. Running motor tests with the Little Bees has shown that their control can lock up and go a little haywire in high throttle regions. Although I have not tested the the SKIZO motors, I do recall the RK5050 and 5051 having that issue with the Little Bees and various motors. I have not seen the same issue with the Bulletproofs so they may fix the issue.

I’d imagine if you are coming from high speed and do a hard turn or a flip, the Little Bee may not be able to process the new throttle levels properly and send it into a spiral that the FC won’t be able to correct.

If it is not the ESC then it would have to be an issue with the Colibri. The gyro could be damaged and the sensor feedback could cause it to spin out of control. It would make sense since the failures occur in extreme orientation regions. Did you get that Colibri from Quad Questions?

Anonymous commented on answer June 27, 2017

Thanks for the replay. I just replaced the ESC this weekend with the Powercube Bulletproof stack. I posted an update, but I think it might be low in the thread down there. Anyway still having the issue, but the triggering of the spinout appears to be lessoned a bit which could be a result of the better ESCs and an overall drop system noise/oscillation overall and with a suspect FC that could mean bad gyro. And that sucks because it has only a hand full of flight days.

1
Anonymous 0 Comments

5051s are pretty severe- without much gain in performance, and a lot of extra amp draw.  You are welcome to check out our test data- we haven’t specifically tested the 5051 on your motor- so it’s hard to tell what it would pull, but we have tested on several other motors with interesting results-  You can dig through github to see the results- we will have posts with nice graphs published soon.

 

https://github.com/QuadQuestions/Motor-Test-Data

 

I emailed a response via our helpdesk so check there so that we can get you flying again.

1
Anonymous 0 Comments

Dang, sorry to hear about the troubles.  This looks like a common desync, and it can happen from time to time for a multitude of reasons.

We have noticed that more people seem to have desync issues with Betaflight 3.1.7- But it seems that if they roll back to an earilier version of Betaflight, it can resolve the issues.  The only thing that we can think of is that the stock D gains increased quite a bit in 3.1.7 which could be the cause.

Can you post your pid values? Better yet, send a dump?

To do so, Go to the CLI Tab in base/cleanflight, type “dump” without the quotes
push enter.

Copy the information that gets put on the screen (highlight with mouse, right click, copy, and paste into an email by right clicking and selecting paste)

Can you tell us what props you are running?

You might also want to check the ESC firmware.  We recommend updating firmware on the ESCs to the latest BLHeli Suite version.  The Little Bee ESCs have been known to have issues when being pushed hard on different prop/motor combos- at least that is what we have observed on the bench.  We have for that reason, decided to discontinue the little bee 20s in favor of TBS bulletproof escs because they don’t exhibit the same issues and have been very good performers in our tests.

It is very possible that your ESCs need be upgraded.  Check out the bulletproof TBS escs here:

TBS BULLETPROOF 25A Racing Drone ESC (BLHELI-S)

I wouldn’t expect the flight controller for now, you could try turning on expert mode in Betaflight and then observing the raw sensor data.  Move the quad all around an check for an anomaly-  You could also try disabling the accelerometer to rule that out.  But I would start with the ESCs, and the pid values/ Betaflight settings.

Anonymous deleted comment June 27, 2017
0
Anonymous 0 Comments

**SOLVED**

Well, in the end it was a bad motor. Surprising isn’t it?

After replacing EVERYTHING, minus the TBS Core Pro, it ended up being one of my Skitzo motors.

I haven’t pulled them apart, or tested them, but I tossed a pair of red bottoms on and ripped through all my batteries without a hiccup.

0
Anonymous 1 Comment

Hi all,

my QQ 190 did the same after a short roll left input.

https://youtu.be/hJCuEHptPO8

 

Before this flight i raised the rates from stock 900 to round 1200deg/sec

After resetting the rates to stock, it is never happend.

qq190 rtf with bf 3.1.0 rc6 qq from tbs agent

Anonymous commented on answer July 6, 2017

How I wish it was that easy (and inexpensive) for me. I had tried the same with adjusting rates. My rates were around 1100deg/sec then set them down to around 640 and it was still happening. I was running Betaflight 3.0.1 Colibri Race from the TBS agent. My new FC is in, but I have only test hovered. Hopefully I’ll maiden in the next couple of days and post the results.

Question stats

  • Active
  • Views12229 times
  • Answers9 answers