Fool unmasked?

[Donnie Frank]

Well, clearly it is working very well

There’s a place at the Salton Sea where I fly that my homepoint is about -220ft below sea level, and it does report accurate negative elevations from there.

Yep. I think what we can take away from this conversation is that a "table" isn't going negatively impact any drone flight and/or landing....<;^)

D

 

[AMann]

Yep. I think what we can take away from this conversation is that "table" isn't going negatively impact any drone flight and/or landing....<;^)

D

You are right, I’m sure there’s a lot of variables and other issues with these sensors that rely on mechanical input.

By the way, New Mexico! Drove a bunch of back roads there to avoid the highways last winter on my way to Texas, you live in some beautiful country my friend!

 

[Donnie Frank]

You are right, I’m sure there’s a lot of variables and other issues with these sensors that rely on mechanical input.

By the way, New Mexico! Drove a bunch of back roads there to avoid the highways last winter on my way to Texas, you live in some beautiful country my friend!

Agreed. Assuming our local government doesn't implode our state, I plan on staying here the rest of my life. It's a great place to be based out of.

D

 

[mnix]

Think about it- If your bottom sensors are not working and your drone is auto lending, the barometer/altimeter will tell the drone that it’s at ground level at the elevated takeoff height, and the motors will shut off too soon making your drone fall the remainder of the distance. It’s sort of like what happened to the Mars Polar Lander.

No it wont - when landing the drone keeps the motors powered for a short time after it detects that the controlled descent has stopped. Why do you think it shuts off the motors after a hand catch?

 

[mnix]

So far I've had no in-flight issues the few times I've launched from the steel checkerplate deck on my "MLRP".
My M2P does generally report mag interference before launch though - that is likely a good thing if the IMU realises the compass has issues it likely just treats the sudden correct reading as "it's working now" rather than getting confused.
I land manually as it's a small area - only issue so far has been getting too close to a stick that was in there - the M2P just shut that motor down early and completed the landing without issues!
They really have got the motor control right on these things - Testing a payload dropper out the other day it somehow sucked the string to the payload into the rear prop and wrapped it round the motor during takeoff.
It actually shut down that motor, landed from a few inches up and idled the other props until I gave the shutdown command. Needless to say I hand launch if there's a payload now (and yes it actually gives a "payload detected" message).

 

[zamboni]

Late to the party but i do have some input. I live in an RV, towed by an F250 pickup. I also live entirely on solar power. That means i need to know where South is to aim the panels. I keep a compass on the dashboard, and when i get to a new campsite, i look for South so i know which way to park.

When i first started doing this, I was lazy and didn't get out of the truck. One night i got up for a piss and noticed the moon was rising...to the Left of the trailer. I had parked the whole rig aimed NORTH, based on my compass reading from the drivers' seat. Since then, I've gradually decreased my laziness by first opening the door, then standing next to the truck, then walking a short distance away...

These days, i walk at least 20 feet from the truck to check a compass reading before parking. And i walk even farther when i'm pulling out the bird.

 

[r.gwynn]

1. This takeoff pad position is (a) a good idea, (b) the last time this drone will ever fly (c) obviously the idea of a beginner and its good he didn’t really try it.

View attachment 74527

C

 

[MEFfly]

MP Platinum.
I've rarely gotten the mag interference error, but the worst instance I was on the edge of a large rock outcrop "mountain"... it took me a good 5 minutes to find a place far enough away from the rocks to get the error to clear so I could actually take off... So I assume rocks with high levels of iron bearing minerals are also an issue?

 

[Thomas B]

MP Platinum.
I've rarely gotten the mag interference error, but the worst instance I was on the edge of a large rock outcrop "mountain"... it took me a good 5 minutes to find a place far enough away from the rocks to get the error to clear so I could actually take off... So I assume rocks with high levels of iron bearing minerals are also an issue?

Absolutely

 

[Ricapoody]

Tried to launch from a table in a park. Metal frame. Timber panels secured with bolts.
Mavic said no. Moved one metre to cement base around table - no problems. I guess the metal frame was more of a problem than the rebar in the slab.

 

[Thomas B]

Tried to launch from a table in a park. Metal frame. Timber panels secured with bolts.
Mavic said no. Moved one metre to cement base around table - no problems. I guess the metal frame was more of a problem than the rebar in the slab.

Or it was reinforced concrete and had no rebar.
Good thing is you moved it instead of trying to force the bird to fly.

 

[Cloudhopper]

Late to the party but i do have some input. I live in an RV, towed by an F250 pickup. I also live entirely on solar power. That means i need to know where South is to aim the panels. I keep a compass on the dashboard, and when i get to a new campsite, i look for South so i know which way to park.

When i first started doing this, I was lazy and didn't get out of the truck. One night i got up for a piss and noticed the moon was rising...to the Left of the trailer. I had parked the whole rig aimed NORTH, based on my compass reading from the drivers' seat. Since then, I've gradually decreased my laziness by first opening the door, then standing next to the truck, then walking a short distance away...

These days, i walk at least 20 feet from the truck to check a compass reading before parking. And i walk even farther when i'm pulling out the bird.

This ... Exactly. A car's built in compass is calibrated to subtract the car's own magnetic field. A pocket compass on the dash cannot be calibrated and can never be trusted to show magnetic north.

 

[AnzacJack]

This ... Exactly. A car's built in compass is calibrated to subtract the car's own magnetic field. A pocket compass on the dash cannot be calibrated and can never be trusted to show magnetic north.

Yep, this is why we do compass swings and calibrations on aircraft so that a deviation card can be made and applied to that craft in that configuration. When there is an engine change or equipment changes, this compass swing must be redone.

 

[zamboni]

I agree the compass cal is meant to factor in local magnetic deviations, but those deviations ought to be the large-scale undulations of the earth's magnetic field; not the presence of some hunk of steel nearby the launch site. The compass cal is used throughout the flight so whatever deviations you're calibrating to should also be present the whole time. Once you take off and grab a few dozen feet of altitude, the truck is no longer part of the picture so you wouldn't want it to be factored into the calibration. That'd be like setting your aperture and f-stop indoors then going outside and taking pics.

Now, I don't know how "good" the DJI calibration is. Maybe it's smart enough to filter out a geologically-tiny local object. Maybe not. One can fly very close to a vehicle or steel building and the drone doesn't freak out instantly -- so obviously this is not a hypersensitive issue. Drones launched from pickup trucks probably aren't going to explode and fall out of the sky for no reason. But there's a tiny chance that it matters, and from where I stand, every tiny chance to reduce risk is worth it.

 

[Meta4]

I agree the compass cal is meant to factor in local magnetic deviations, but those deviations ought to be the large-scale undulations of the earth's magnetic field; not the presence of some hunk of steel nearby the launch site.

Compass calibration has nothing to do with magnetic deviation or the earth's magnetic field.
It is only to measure the magnetic fields that are part of the drone itself so they can be discounted, allowing the compass to measure the earth's magnetic field properly.

Drones launched from pickup trucks probably aren't going to explode and fall out of the sky for no reason. But there's a tiny chance that it matters, and from where I stand, every tiny chance to reduce risk is worth it.

Every week crash reports come in from flyers that launched from reinforced concrete surfaces or other places where their compass was close to steel.
Their problem is yaw errors caused by initialising within the distorted magnetic field of the steel item.

 

[zamboni]

Always happy to be corrected when I've got the details wrong...I didn't realize how the calibration was meant to account for the bird's own internal fields. However it sounds like it uses the earth's magnetic field to do this, as that is the "reference" component which does not move while you're spinning and turning the drone. So the point is not to know where North is, but to have frame of reference for evaluating those internal fields. Local distortions would impact that reference.

So I got the "why" wrong but I think the takeaway here is still on-point: The bird should be away from metal stuff when calibrated. If the truck skews the drone's own internal fields, then once you fly away from the truck that influence is gone and the calibration is less accurate. And if there are /weekly/ crash reports related to this, then the chance of error isn't "tiny" like I said -- it's substantial -- and this level of caution is justified.

This does get me thinking, though. I'm not trying to argue -- want to learn: Why is calibration advised when changing location? If the drone figures out its own fields by subtracting the earth's field from its measurements, that tells it what the internal fields are. Now it knows. If you move the drone 1000 miles, the local magnetic field is different...but have the drone's internal fields actually changed, too? Why repeat the cal? It sounds like the best thing to do is calibrate your drone out in the middle of the desert where there isn't any steel for miles, and then never recalibrate again, since the existing cal would already be very "clean". Do the drone's internal fields drift?

 

[Thomas B]

Always happy to be corrected when I've got the details wrong...I didn't realize how the calibration was meant to account for the bird's own internal fields. However it sounds like it uses the earth's magnetic field to do this, as that is the "reference" component which does not move while you're spinning and turning the drone. So the point is not to know where North is, but to have frame of reference for evaluating those internal fields. Local distortions would impact that reference.

So I got the "why" wrong but I think the takeaway here is still on-point: The bird should be away from metal stuff when calibrated. If the truck skews the drone's own internal fields, then once you fly away from the truck that influence is gone and the calibration is less accurate. And if there are /weekly/ crash reports related to this, then the chance of error isn't "tiny" like I said -- it's substantial -- and this level of caution is justified.

This does get me thinking, though. I'm not trying to argue -- want to learn: Why is calibration advised when changing location? If the drone figures out its own fields by subtracting the earth's field from its measurements, that tells it what the internal fields are. Now it knows. If you move the drone 1000 miles, the local magnetic field is different...but have the drone's internal fields actually changed, too? Why repeat the cal? It sounds like the best thing to do is calibrate your drone out in the middle of the desert where there isn't any steel for miles, and then never recalibrate again, since the existing cal would already be very "clean". Do the drone's internal fields drift?

The drone should have a compass recalibration if moved to a take-off spot > 50 km from the last one, if the drone indicates a compas calibration error, or if it has not been calibrated for 30 days per the full DJI manual on page 65.

https://dl.djicdn.com/downloads/Mavic_2/20190417/Mavic_2_Pro_Zoom_User_Manual_v2.0_en.pdf

The drone will likely force a compass calibration every 30 days.
Easier than chasing a fly away!

 

[Meta4]

Why is calibration advised when changing location? If the drone figures out its own fields by subtracting the earth's field from its measurements, that tells it what the internal fields are. Now it knows. If you move the drone 1000 miles, the local magnetic field is different...but have the drone's internal fields actually changed, too? Why repeat the cal? It sounds like the best thing to do is calibrate your drone out in the middle of the desert where there isn't any steel for miles, and then never recalibrate again, since the existing cal would already be very "clean". Do the drone's internal fields drift?

There really is no need to recalibrate after travelling any magic distance since it's only measuring the fields that are part of the drone and they don't change (unless you modify the drone).
The whole silly idea comes from DJI's bad documentation which suggested it.
DJI finally got their compass calibration text (almost) right with the manual for the Phantom 4 pro in late 2016.
But DJI confused everyone again with the Mavic 2 manual which goes back to recommending recalibrating when:

1. Flying at a location farther than 31 miles (50 km) away from the location the drone was last flown. ​

2. The aircraft has not been flown for more than 30 days​

Although there is no physical reason for this, they have confused people further by making some Mavic 2s force this.

Easier than chasing a fly away!

All that unnecessary recalibration of compasses does nothing to prevent what some flyers might call a "flyaway".
Those are caused by other features, notably the lack of understanding how to fly safely.