FT Guardian Gremlin v.2 flips on takeoff

[dylanbeaudette]

The kid and I recently built the FT GG v.2 kit from the Flight Test store. The built and the follow-up programming video were great. After a couple of dead-ends, with the stick input from RX not registering on the flight controller, a firmware flash from Betaflight did the trick. I noticed that there were a couple of relevant lines in the dump file, so we ran those commands and saved. We are using the suggested 800mAh 2S LiPo battery. A single motor direction needed to be reversed in the BL Heli configurator.

Additional commands run in betaflight:
feature RX_SERIAL
set motor_pwm_protocol = DSHOT600
profile 0
rateprofile 0

As far as we can tell everything is "right":

1. motors in "X" configuration
2. accelerometers calibrated
3. motors 1, 2, 3, 4 in the correct positions as tested in betaflight
4. props installed according to the programming video
5. motors / props spinning correctly according to programming video
6. props free of obstructions
7. fully charged battery
8. prop screws appropriate length
9. controls, arm, and mode switches operate as expected in betaflight
10. flight controller installed with "forward" in the correct orientation

After all of these checks / tests, the Gremlin arms and seems fine. Yaw controls operate as expected and the quad spins around on the floor. Increasing throttle appears to work until it instantly flips to the left and lands upside down. Every takeoff attempt results in a flip to the left. Apart from a bad motor or ESC, we can't think of anything else that could be causing this behavior. Pretty disappointed at this point.

Any suggestions?
Thanks!
Dylan

Kit:
https://store.flitetest.com/ft-guardian-gremlin-v-2/

Build video:

Basic programming and motor / prop setup video:

We are using the FrSky Archer M+ Access RX:
https://alofthobbies.com/collections/frsky-archer/products/frsky-archer-m-access

Short video of the flip-on-takeoff. It always seems to flip to the left.
https://www.youtube.com/shorts/frP6Z4dwXWk

 

[Adore]

Betaflight shows your rx inputs correctly?

 

[dylanbeaudette]

Betaflight shows your rx inputs correctly?

Yes, double-checked that. Everything "seems" right, but this is all new to us--we've only built / flown fixed wing until this kit.

 

[luvmy40]

OK, I know less than nothing about quads but, it looks like it is either throttling up the right motors much faster than the left, or the left motors are cutting out at some throttle position. I am assuming the stack has one ESC that feed/controls all four motors? Has this been calibrated? Is that a thing with the quad stacks?

As dumb as it sounds, is there any chance you might have different props on the left and right as far as pitch?

 

[dylanbeaudette]

UPDATE

The gremlin is now flying like a dream, even with a giant 2200mAh battery on top.

A much more knowledgeable friend pointed out that even though the flight controller was installed correctly (arrow on the PCB pointing forward", the flight controller acted as if it were reversed. This was confirmed in Betaflight by observing the motion of the 3D model while slowing moving / tilting the drone in all directions. The flight controller was given a 180 degree rotation in Betaflight and all was well.

This thing is a ton of fun to fly with all kinds of batteries and the prop guards are great for newbies like me. It does burn through batteries: our little 800mAh 2S lipos would last about 4 minutes. It flew surprisingly well with a 1300mAh 3S battery, and even managed to get off the ground with a 2200 mAh 3S.

One of the nylon standoffs became loose in flight was thrown from the frame by the prop. Thankfully nothing was damaged and new hardware are already in the mail. Our next project is securing the motor wires safely below the FC, as they were dislodged and almost sliced in the recent incident. Still trying to find the best way of securing the cell balancing connector and RX antenna during flight to avoid accidents.

I'd highly recommend this kit!

 

[dylanbeaudette]

UPDATE 2.

Still having fun flying, but learning about the added expenses / complexity associated with multirotor aircraft...

This quad spends about as much time on the bench as it does in the air: the nylon screws used to hold the frame together have a tendency to break inside of the stand-offs, or where the threads meet the screw head. We haven't had any disasters related to frame failures in air, but we now check for broken screws after each flight. Thankfully FT provides a nylon hardware kit on the FT Store (https://store.flitetest.com/ft-guardian-gremlin-hardware-kit/) which has been incredibly useful for field repair.

Perhaps the biggest flaw in the FT GG kit is the choice of motor connectors on the flight controller (https://store.flitetest.com/ghf411-pro-bmi-40a-toothpick-f4-2-6s-aio-brushless-flight-controller/). Granted, this is simpler than soldering motor wires to surface mount copper pads on the v.1 Gremlin. Hoewver, the convenience is offset by the fragility of these sockets: a little too much pressure when installing a motor plug can detach or damage the socket. We were able to very carefully bridge a broken socket wire with solder which lasted for a while. The bridge failed during our last flight and caused to the quad to come tumbling out of the air.

Does anyone know what kind of connectors these are? I think that it would be simpler to replace an entire connector vs. attempting to "fix" a damaged one.

 

[luvmy40]

You can reinforce those connectors to the board with hot glue.

 

[dylanbeaudette]

You can reinforce those connectors to the board with hot glue.

This is a good idea. We tried that, but only after first accidentally damaging one of the connectors when installing the plug.

 

[dylanbeaudette]

Looking closer at the FC, it is possible to solder the motors directly to it using 3 copper coated notches. This worked well, and it seemed like a replacement motor was going to work.

Backing up a bit, while it had originally seemed like the connector had failed, upon closer inspection it was fine. We removed the motor and attached it to a simple ESC/RX pair and attempted to run it up. The motor behaved exactly as when it was connected to the Gremlin FC: strained rotation back / forth, and then became very hot.

Just in case, we attempted soldering a new motor to the notches on the FC, after carefully removing the bent connector. The new motor acted in the same way: strained rotation, very hot, and then some smoke. Clearly the ESC on the FC is now eating motors.

Next step is contacting FT customer support. This is a really fun quad, but nonstop difficulties.

 

[dylanbeaudette]

UPDATE

After explaining the above to FT Store Support staff, they promptly sent a replacement flight controller! How about that for great service?

I'll put the rebuild details into a new, more informative thread.