Tello Edu prone to unexpected rotation behavior -- experiences?

I have 3 regular Tellos (white) 2 Edu Tellos and 2 TT models.

I am using the Tello SDK to program them using Python and other means. But the weird behavior I am seeing with 1 of the Edu models happens even when I am using the official iOS app.

I can get the Edu model to take off OK but then it often (not always) will start to yaw without any commands to do so. Sometimes the yaw is extreme and I am worried it will spin out of control. When I give a cw or ccw command, sometime nothing happens at all. Up and Down commands work ok generally. rc commands that have yaw component will also not work from time to time.

The other Edu model and the TT models do NOT exhibit this behavior.

Happens when the room is well lit and when the floor surface is not a uniform color too.

If I turn the Tello on and off and try again, this bad behavior sometimes is gone for a period of time – I can fly the Tello normally.

I am running the latest firmware for the Edu model. I have tried IMU calibrations and Center of Gravity calibrations and they may seem to help for a couple of restarts but then this misbehavior always comes back after a time.

I suspect a hardware glitch of some kind but returning to China for a repair does not seem like a path to recovery.

Anyone with experiences along these lines? Any remedial steps that I might try?

I have been using cages for my Tellos – the Bubble frame that I saw being used in one of the DroneBlocks Python courses. For one of the Edu models, no issues resulted when the cage was in use.

I ordered and just received a flatter cage design that I saw in one of the schools being shown on a local TV news segment I linked to from a DroneBlocks web page.

So far, this cage when enclosing the problem Edu model has not been associated with the problematic flying.

Anyone with Tello cage stories to tell: models, recommendations, etc.?

What firmware version are you using and is it consistent with the firmware on the “working” Tello drones? Let me know and I can do some code testing with my Tello drone this weekend.

I have not run into any issues other than decreased battery life due to the Tello cage. I know @pat.ryan prefers the 360 degree cage as they call it. We did have one customer that request the flatter cage because they were using Tellos inside a netted area for a competition. Let us know what your experience is after you do more testing.

I use the cage because I just dont know everyones flying environment and I have one more than one occasion caused the Tello to fly up and hit the ceiling. :slight_smile:

Given that most people are new to programming and the Tello I expect the Tello to get away from people. The cage has done a good job in protecting the Tello all of the things I have run into.

The only issue I have seen with the cage is trying to do flips. I absolutely do not recommend doing flips. While it has been successful for me, it has also ended badly.

1 Like

I have done some more testing and the white (original) tello’s (I have 3) and one of my Edu tello’s do quite badly with the 360 degree cage. Bad behavior while not 100% repeatable is quite common. Most common symptom is unplanned rotation once in the air. The one other Edu model is very stable with the 360 cage but the Talents are a little twitchy. I can leave the LED matrix attached inside the 360 degree cage though.

So far with the flatter cage I am seeing NO issues. I now have 4 of them and they are enclosing my two Tello Edus and Talents. I needed to get a new top shell for the TT models in order for the lego dimples to not impact the top support used in the flat cage and these are working well.

With my ESP32 controller, the white ones do well with the standard prop guards attached. But in a classroom, I want more protection than that provided by the prop guards.

All Tellos are using the latest firmware for the respective model.

I do see reduced battery life as well.

The brands I have tried are PYGTECH (the spherical cage) and CYNOVA (the flat cage).

I use the CYNOVA (Flat) cage and have had no problems - looks more like a drone, and kids in the class are safer without it looking “un-aerodynamic” (!?) protects them beautifully when crashed into walls, ceiling fans etc

In a classroom I am running up to 10 tello drones. I have found that because the tellos all use the same IP address, even if you turn them on one at a time to avoid channel interference, the running of coded programs can result in weird drone behaviour. As far as I can see, the interference caused by multiple drones affect the drone flight. With a camera feed it seems to get worse. Drones take off and land, totally ignoring a looped shuttle run. If you abort a flight midway, when you try the flight again, it seems to start from where it was aborted, so it looks like you have to turn the drone off and on again to drop the program. I have used the chrome extension which is quite good but has no camera utility. The windows beta version of droneblocks runs well but not with multiple drones. We are using the
I tested all the drones one by one and could not replicate the flight errors. As soon as we added other drones, we had errors where drones turned when there was no turn programmed or simply took off and landed- ignoring lopped shuttle programming. I’m not sure what the answer is other than to use a joystick.

Janine, would you mind reading through the post below and giving it a try? Specifically please try toggling off the video stream and see if it leads to improved behavior. There is a new build linked in the bottom of the post.

HI Dennis,
In our session yesterday we used the new build and the response of the drones was much improved. We did toggle off the cameras unless flying, and this function seemed to keep the wireless noise down. The drones were reading loops correctly and so their grid pattern flight was much more successful. We only had once or twice where random flights occurred, but I think some cameras were turned on at the time. I would say the noise is the factor. I had LED lights installed in the classroom to increase the room’s brightness, which has also helped. Going to complete the mapping today and have a look at the resulting images of the mapping. Mission pads will be next week.

1 Like

Thanks Janine! We’ll have an update coming soon where students can save and retrieve their Tello mission code just like we have on the DroneBlocks mobile apps. Great to hear things are working better.