Very weird behaviour

RobotBee · May 3, 2020

sar104 said:
Right - that is what is happening - spurious throttle input with no elevator input.

Your analysis is really helpful - I'll pass the info on and see if I can get a 3rd controller...

Thanks a lot

RobotBee · May 3, 2020

dawgpilot said:
I've had this same issue pop up sporadically too. I'm also on my 2nd controller after DJI replaced the original one, and still see it. Unfortunately I'm out of warranty now so further warranty replacements aren't an option for me anymore.

I'm upgrading to the Mavic Air 2 next week, so hopefully these ghost inputs don't follow me to the new model!

That's not good! Did you get the drone replaced too, or just the controller?

Are you selling your Air 1 or keeping it when you upgrade?

Thanks

dawgpilot · May 3, 2020

RobotBee said:
That's not good! Did you get the drone replaced too, or just the controller?

Are you selling your Air 1 or keeping it when you upgrade?

Thanks

Just the controller was replaced, about 1 year ago. The issue doesn't always affect my flights so I still find the drone OK to fly most of the time.

I think for now I'll still keep my Air 1, especially for travel purposes since it fits in a smaller volume than the Air 2. If I find that traveling with the Air 2 isn't too bad, I might get rid of the Air 1 though.

Deleted member 94047 · May 3, 2020

This is indeed weird, and even more so considering you had the RC replaced. If it is still in warranty, return it - it is a disaster waiting to happen. May be third time is the charm

Deleted member 94047 · May 3, 2020

RobotBee said:
Yes, but what I mean is that during that time I'm pretty sure I wasn't touching either of the sticks, yet the left stick is still registering downwards slightly and causing the drone to drop slowly...

Thanks

A phantom input on one of the sticks could be because of a stuck stick, and is not that uncommon. A remote controller stick calibration can fix that. But I am not sure this is the case with yours because a lot more is going on. The cross talk between the left and right sticks especially is indicative of a faulty controller. I don't think any amount of stick calibration can fix that. By the way, you can go in to the remote controller calibration settings and see how the sticks respond without powering on or connecting to the aircraft. Power on the remote controller, connect your phone, and launch GO4. On the home screen, tap on "Enter Device", and on the next page tap on "Enter Camera View". Once GO4 enters camera view, tap on the three dots at the top right corner to open up settings, tap on the remote controller icon, and then "remote controller calibration". You will see this

Don't tap calibrate, just push the sticks every which way and see what inputs they register and see if any of the sticks is registering phantom inputs while they are centered and at rest. The cross-talk issue should also exhibit itself here when you push on the sticks; if it doesn't, it may suggest an issue with the AC too. When done with your testing, simply exit the app.

Good luck!

RobotBee · May 4, 2020

Doppler said:
A phantom input on one of the sticks could be because of a stuck stick, and is not that uncommon. A remote controller stick calibration can fix that. But I am not sure this is the case with yours because a lot more is going on. The cross talk between the left and right sticks especially is indicative of a faulty controller. I don't think any amount of stick calibration can fix that. By the way, you can go in to the remote controller calibration settings and see how the sticks respond without powering on or connecting to the aircraft. Power on the remote controller, connect your phone, and launch GO4. On the home screen, tap on "Enter Device", and on the next page tap on "Enter Camera View". Once GO4 enters camera view, tap on the three dots at the top right corner to open up settings, tap on the remote controller icon, and then "remote controller calibration". You will see this

View attachment 100440

Don't tap calibrate, just push the sticks every which way and see what inputs they register and see if any of the sticks is registering phantom inputs while they are centered and at rest. The cross-talk issue should also exhibit itself here when you push on the sticks; if it doesn't, it may suggest an issue with the AC too. When done with your testing, simply exit the app.

Good luck!

Thanks. With the previous controller, I did this, and I could see the sticks doing weird things - registering an input without being touched.

With the current controller, I've only entered this screen when I've been doing the calibration and haven't tried to replicate the phantom inputs without calibrating. They seemed normal when I was calibrating, but as the problem is intermittent anyway that doesn't mean anything I guess.

I have asked the retailer to try sending me a new controller before a new drone so we'll see what they say. I'm sure they'll say it's DJI's decision, so more waiting I guess.

Seems unbelievable that I've got the same issue with two controllers and two drones, but looks like that's the only explanation!

Thanks for the reply.

RobotBee · May 4, 2020

Drone and controller are both going back to the retailer so we'll see what happens next. Thanks guys for the help.

Deleted member 94047 · May 4, 2020

RobotBee said:
Drone and controller are both going back to the retailer so we'll see what happens next. Thanks guys for the help.

Good luck and keep us updated.

slup · May 5, 2020

No gigantic smart watch on your wrist during these weird ghost inputs... other thing's of magnetic material close to the sticks?

RobotBee · May 5, 2020

slup said:
No gigantic smart watch on your wrist during these weird ghost inputs... other thing's of magnetic material close to the sticks?

Someone suggested that, and I do wear a smartwatch, but I tried it with the smartwatch nowhere near it and it was still doing the same thing...

I used to use one of those magnetic micro USB plugs in the side of the controller, but since the controller was replaced I haven't used it, so that's not the issue either....

Thanks for the suggestion though!

sar104 · May 5, 2020

slup said:
No gigantic smart watch on your wrist during these weird ghost inputs... other thing's of magnetic material close to the sticks?

The correlation between elevator and throttle inputs in some of those instances would seem to rule out random interference.

RobotBee · May 5, 2020

Doppler said:
Good luck and keep us updated.

Thanks, will do!

Nostris · May 13, 2020

Could it be interference of the Wi-Fi signal? It’s really strange, for sure.

RobotBee · May 16, 2020

Nostris said:
Could it be interference of the Wi-Fi signal? It’s really strange, for sure.

Doesn't seem to be, cos it does it in different locations....

Anyway, the retailer has sent me their own controller to use to test it....

RobotBee · May 16, 2020

sar104 said:
Oh yes - that is very strange. There is clearly some kind of crosstalk going on between elevator and throttle on several occasions, and it isn't just random interference on the throttle signal - it is well correlated with the elevator input. It doesn't happen every time you apply throttle though:

View attachment 100419

I've never seen that before - my guess would be that you have a faulty RC. Are you sure that it was replaced?

What software do you use to view that graph?

I've got another controller to try with, so would be good to be able to check the flight logs afterwards...

Thanks a lot

sar104 · May 16, 2020

RobotBee said:
What software do you use to view that graph?

I've got another controller to try with, so would be good to be able to check the flight logs afterwards...

Thanks a lot

TXTlogToCSVtool to convert the txt log, and Wavemetrics Igor Pro to process and view the data.

slup · May 16, 2020

RobotBee said:
What software do you use to view that graph?

I've got another controller to try with, so would be good to be able to check the flight logs afterwards...

Thanks a lot

sar104 said:
TXTlogToCSVtool to convert the txt log, and Wavemetrics Igor Pro to process and view the data.

You can also use CsvView which is free ... download here --> CsvView Downloads

1. Click here & point to a local stored .TXT or .DAT log
2. Click on "Empty" to get up an empty chart
3. Chose the RC signals (stick inputs)
4. View them in the chart area

Use the mouse scroll wheel to zoom in/out in the chart ... or click-draw a box in the chart location you want to look closer into.

RobotBee · May 16, 2020

sar104 said:
TXTlogToCSVtool to convert the txt log, and Wavemetrics Igor Pro to process and view the data.

slup said:
You can also use CsvView which is free ... download here --> CsvView Downloads

1. Click here & point to a local stored .TXT or .DAT log
2. Click on "Empty" to get up an empty chart
3. Chose the RC signals (stick inputs)
4. View them in the chart area

Use the mouse scroll wheel to zoom in/out in the chart ... or click-draw a box in the chart location you want to look closer into.

View attachment 101515

Thanks a lot guys.

What are aeleron, elavator, throttle and rudder in terms of left/right stick up/down?

Thanks

slup · May 16, 2020

RobotBee said:
Thanks a lot guys.

What are aeleron, elavator, throttle and rudder in terms of left/right stick up/down?

Thanks

With the default stick mode 2 it is ...

Also note that CsvView still say for instance Elevator (Right stick up/Down) if you use another stick mode than 2 ... so with mode 1 or 3 the Elevator graph actually show Throttle (AC Ascend/Descend)

RobotBee · May 16, 2020

slup said:
With the default stick mode 2 it is ...

View attachment 101552

Also note that CsvView still say for instance Elevator (Right stick up/Down) if you use another stick mode than 2 ... so with mode 1 or 3 the Elevator graph actually show Throttle (AC Ascend/Descend)

View attachment 101553