Can flying near radio towers interfere with Mavic Pro?

[sar104]

Any type of interference (wireless, radio, electrical) will play havoc with your drone. I live fairly close to some power lines and I always get the warnings about that every time. Sometimes it messes up my flight and sometimes it doesn’t. Go figure?

Power lines won't interfere with a Mavic. Something else is causing your warnings.

 

[Congoblue]

Feel free to disagree but you may have missed the point of my post. I was not arguing that high-power transmissions cannot interfere with aircraft communications or electronics, I was pointing out that the problem, in this case, was a very transient yaw separation on the first yaw maneuver after takeoff. It didn't persist or return, and there is no evidence to suggest that it was anything other than an initial rate gyro gain error.

If there really had been EM interference with the aircraft electronics then I would have expected to see problems throughout the flight, not just at takeoff.

That tower has microwave repeaters on it which output a very tightly focused beam of RF energy. Also, the other antennas will have directionality and will have low field strength at ground level close to the tower. So I do not think your expectation about constant problems is correct. If he flew through a microwave beam, or possibly as he got to the altitude where the main antenna radiation was, it is my opinion that could cause the temporary error that he saw.

 

[FoxhallGH]

Ask Ken Heron! Flying too close to radio towers with a drone

This is an interesting post, as it does show the wide spectrum of errors that - in this case - Ken's Phantom 4 throws up as it flies around the antennas and into the concentrated RF fields produced by those directional antennas. They have waaayyy less gain than the parabolic dishes on the Reigate tower too.

 

[Inspirephil]

Power lines won't interfere with a Mavic. Something else is causing your warnings.

Not with a Mavic it’s with my Inspire 2.

 

[BudWalker]

The problems in this flight are associated with changes in Yaw. This probably got buried as this is a very long thread. In this plot the red shows the Yaw that is computed by the FC based (mostly) in the IMU data. The plot also shows magYaw which is the heading computed by DatCon based just on the magnetometer data.

It's evident that every time the AC rotates there is a Yaw/magYaw separation. The very first Yaw/magYaw separation resulted in the YAW_ERROR. It appears that subsequent Yaw/magYaw separations were not large enough, long enough, or steep enough to cause the FC to declare a YAW_ERROR.

The plot also illustrates the FC response to a Yaw/magYaw separation. The computed YAW is held constant while the AC is rotated until the magYaw value is equal to the Yaw value.



If RF energy were coupled into the MC circuits that would show up as noise on the various signals. There were no indications of this. But, here is an example where noise can be seen on the z axis magnetometer (red plot). In this case it was caused by high current (black plot) being sent to the a motor through wires that weren't twisted enough.

 

[adam3112]

The problems in this flight are associated with changes in Yaw. This probably got buried as this is a very long thread. In this plot the red shows the Yaw that is computed by the FC based (mostly) in the IMU data. The plot also shows magYaw which is the heading computed by DatCon based just on the magnetometer data.
View attachment 53107
It's evident that every time the AC rotates there is a Yaw/magYaw separation. The very first Yaw/magYaw separation resulted in the YAW_ERROR. It appears that subsequent Yaw/magYaw separations were not large enough, long enough, or steep enough to cause the FC to declare a YAW_ERROR.

The plot also illustrates the FC response to a Yaw/magYaw separation. The computed YAW is held constant while the AC is rotated until the magYaw value is equal to the Yaw value.

View attachment 53109

If RF energy were coupled into the MC circuits that would show up as noise on the various signals. There were no indications of this. But, here is an example where noise can be seen on the z axis magnetometer (red plot). In this case it was caused by high current (black plot) being sent to the a motor through wires that weren't twisted enough.
View attachment 53108

Hi @BudWalker

Did you manage to see if there were any issues in my subsequent test flight? I posted the files the other day so it’s abit further back in the thread.

Thanks

 

[sar104]

That tower has microwave repeaters on it which output a very tightly focused beam of RF energy. Also, the other antennas will have directionality and will have low field strength at ground level close to the tower. So I do not think your expectation about constant problems is correct. If he flew through a microwave beam, or possibly as he got to the altitude where the main antenna radiation was, it is my opinion that could cause the temporary error that he saw.

It's one of those things that's hard to rule out definitively without any detailed information on the tower and its emission pattern. However, if you look at the data, it's apparent that the fusion failure was directly triggered by a yaw maneuver while in a hover - the aircraft wasn't climbing or moving laterally at that time. That suggests to me that it was not caused by entering a collimated beam.



Close to a microwave horn could, conceivably, have an effect, although it's not been reported as far as I'm aware - I've seen no flight records at all that implicated EM interference with flight control. I attended a DJI demonstration using an M210 and a Mavic for tower inspection with similar systems, and it was stated that the aircraft were not affected by microwave or cellular transmissions. Anecdotally only, I've flown around similar towers without problems. The same may not be true for tens of kW of TV or radio transmissions, but this doesn't appear to be one of those towers and they are omni-directional anyway.

 

[sar104]

The problems in this flight are associated with changes in Yaw. This probably got buried as this is a very long thread. In this plot the red shows the Yaw that is computed by the FC based (mostly) in the IMU data. The plot also shows magYaw which is the heading computed by DatCon based just on the magnetometer data.
View attachment 53107
It's evident that every time the AC rotates there is a Yaw/magYaw separation. The very first Yaw/magYaw separation resulted in the YAW_ERROR. It appears that subsequent Yaw/magYaw separations were not large enough, long enough, or steep enough to cause the FC to declare a YAW_ERROR.

The plot also illustrates the FC response to a Yaw/magYaw separation. The computed YAW is held constant while the AC is rotated until the magYaw value is equal to the Yaw value.

View attachment 53109

If RF energy were coupled into the MC circuits that would show up as noise on the various signals. There were no indications of this. But, here is an example where noise can be seen on the z axis magnetometer (red plot). In this case it was caused by high current (black plot) being sent to the a motor through wires that weren't twisted enough.
View attachment 53108

Very nice - I didn't think to look for noise on the sensor data.

 

[BudWalker]

Hi @BudWalker

Did you manage to see if there were any issues in my subsequent test flight? I posted the files the other day so it’s abit further back in the thread.

Thanks

I didn't know that you had added more .DATs. I'll take a look.

 

[BudWalker]

Hi @BudWalker

Did you manage to see if there were any issues in my subsequent test flight? I posted the files the other day so it’s abit further back in the thread.

Thanks

Where are the new .DATs?

 

[sar104]

Not with a Mavic it’s with my Inspire 2.

Makes no difference.

 

[ac0j]

post 91 and 92 @BudWalker !

 

[Congoblue]

If RF energy were coupled into the MC circuits that would show up as noise on the various signals. There were no indications of this. But, here is an example where noise can be seen on the z axis magnetometer (red plot). In this case it was caused by high current (black plot) being sent to the a motor through wires that weren't twisted enough.

That's very good, I'm convinced that you're right.

 

[adam3112]

Where are the new .DATs?

I'm not sure which dat file it is, but here's both parts from today attached.

They are here in the quoted message, but I'll re-attached them tonight when I get in from work. I would of re-downloaded them and attached to this message but my work are quite heavy on us downloading files at the minute because of a recent security breach on our servers (as I'm typing this from my work computer).

Edit: .DATs are found on page 5.

 

[BudWalker]

Hi @BudWalker

Did you manage to see if there were any issues in my subsequent test flight? I posted the files the other day so it’s abit further back in the thread.

Thanks

It seems that the MP has the same problem.


Can you do an IMU cal, then test, then a compass cal, then test? Hang in there - you might be thinking calibrate everything and get back to flying

 

[adam3112]

It seems that the MP has the same problem.
View attachment 53123

Can you do an IMU cal, then test, then a compass cal, then test? Hang in there - you might be thinking calibrate everything and get back to flying

Thanks @BudWalker I'll try and find some time later this week to do these further tests and post the results.

Just to mention again that I experienced no issues whilst doing this test. I even took the Mavic out for 20 minutes the next day and it flew like a dream.

Am I expected to see any issues like the one in question whilst doing these tests?

 

[adam3112]

One other thing worth mentioning. On close inspection of yesterday's 20 minute flight, the arrow on the map again appears to be facing slightly off to the left compared to the direction the camera on the MP is facing. It's particularly noticeable on takeoff.

Edit: Actually, looking at the Satellite view, it looks about right. Might just be the Map view is slightly inaccurate with it's layout of the park I was in.