Altitude off?

[Jdruding]

Hi all,

I have a Phantom 2 Vision. I just exchanged my previous one due to a recurring controller error problem. Now, I got the new one, charged it up, fired it up and started flying. Problem is, the altitude appears to be off. When I start it up, it starts off showing about 10 ft off the ground. If i just let it sit, the altitude with drift up, up, up. Right now, it is sitting next to me on the table and it is reading 35 ft and climbing. Tried compass calibration, power on/off, advanced IMU calibration. Same problem. Anyone else experience this? Am I going to have to exchange this $%#&$ thing again? :x :x

Sincerely,

Frustrated with DJI

 

[N017RW]

Common problem, not necessarily a specific defective part.

Altitude is based on air or atmospheric pressure.

This pressure is also affected by heat which is why you see it drifting.

Most folks let it warm up then recycle power to get a 'fresh' reading at the warmer temp and then take off.

 

[Jdruding]

How long would you recommend it sit before power cycling?

 

[morph000]

Actually I'd say this is another prime example of stupidity on DJI's part.
Look at the app readout - it shows altitude to 0.1m
Yeah right...a barometric altimeter wouldn't come close to that EVER !
Even with a 20 satellite lock it'd be almost impossible (especially due to accuracy constraints imposed by the US gov't) to get within 3m.
0.1m is pure nonsense so why even bother with it ? I have no idea !
What we NEED is the ability to set a zero point. eg on the ground we hit "Zero" and that's the 0m reference point.
Maybe the coffee app ?

 

[crash1sttime]

I noticed the same thing when I got mine , not much you can do about it, me, I just fly and enjoy the view.

 

[Buckaye]

Actually I'd say this is another prime example of stupidity on DJI's part.
Look at the app readout - it shows altitude to 0.1m
Yeah right...a barometric altimeter wouldn't come close to that EVER !
Even with a 20 satellite lock it'd be almost impossible (especially due to accuracy constraints imposed by the US gov't) to get within 3m.
0.1m is pure nonsense so why even bother with it ? I have no idea !
What we NEED is the ability to set a zero point. eg on the ground we hit "Zero" and that's the 0m reference point.
Maybe the coffee app ?

Even if you set a zero point on the ground... depending on the atmospheric pressure changes and the temperature of the machine... it will still likely drift. (I have seen pretty big differences in the starting height flight to flight - so much so that I don't do anything automatic under 60 ft anymore.)

 

[Jdruding]

Thanks all. I understand that it may drift as temps or the weather changes, but I agree, it seems like when it gets a home lock, it should either reset to zero or we should have the option to reset it to zero.

 

[PhantomFanatic]

If the altitude shows 25' then if you do a RTH, then it will be trying to settle 25' below ground level. Not good!
I assume you did a successful compass calibration and you had all greens before take off. It needs to be done anytime you fly from a new location.

I wonder if a compass replacement might help your situation, or a complete swap, if you purchased from Amazon.

 

[SteveMann]

Actually I'd say this is another prime example of stupidity on DJI's part.
Look at the app readout - it shows altitude to 0.1m
Yeah right...a barometric altimeter wouldn't come close to that EVER !
Even with a 20 satellite lock it'd be almost impossible (especially due to accuracy constraints imposed by the US gov't) to get within 3m.
0.1m is pure nonsense so why even bother with it ? I have no idea !
What we NEED is the ability to set a zero point. eg on the ground we hit "Zero" and that's the 0m reference point.
Maybe the coffee app ?

I don't know which pressure chip used in the Phantom controller, but the altimeter resolution of the ($6 in quantity) HP206C barometer/altimeter module is 0.01 m. It is doable with a low-cost module. With that sensitivity temperature changes and even a slight breeze will cause some drift.

GPS altitude is not accurate due to the geometry, not any accuracy constraints. Even GPS approaches in manned aircraft are considered non-precision because of the altitude error.

 

[SteveMann]

If the altitude shows 25' then if you do a RTH, then it will be trying to settle 25' below ground level. Not good!

Not the way it works. RTH ascends to the RTH height (by the barometric altimeter) then goes to the home positing and descends until it won't descend any more. Then the motors turn off.

 

[morph000]

I don't know which pressure chip used in the Phantom controller, but the altimeter resolution of the ($6 in quantity) HP206C barometer/altimeter module is 0.01 m. It is doable with a low-cost module. With that sensitivity temperature changes and even a slight breeze will cause some drift.

GPS altitude is not accurate due to the geometry, not any accuracy constraints. Even GPS approaches in manned aircraft are considered non-precision because of the altitude error.

Impressive accuracy - pity it doesn't materialise in the V2+ eh ?
One would reasonably assume DJI used a chip of similar specs in order to achieve the illusionary 0.1m accuracy.

However...I stand by my original claims -barometric altimeters can vary by HUGE amounts (tens of meters) despite such accuracies.
I did a few checks today on my phone (Samsung Note 2) to see just how accurate things really were.
Frame of reference was government issued topographical data for my exact location. 98m ...

Phone sensor data:
(bearing in mind the weather bureau data was not at my exact location, but close)


GPS app (one of 2 but showed the same):



Accuracy with 13 sat lock:


Note incidentally how many sats have low SN ratios which probably explains why sat count drops when motors power up.

As I'm close to the beach, I went down to high water and the GPS app showed 2m, so pretty close.

So summing up - we're looking at < 5m accuracy with GPS and with phone's altimeter.

Once again - all that is needed is for DJI to add a "Zero" facility OR set it to either 0m(bird on ground) or 1.5m (shoulder height for compass cal)at power up.

 

[SteveMann]

What app are those? And how did you capture the screen?

 

[N017RW]

Definately the SNR.

In fact DJI has just improved their A2 GPS Pro 'puck' to improve SNR:

http://www.dji.com/info/releases/introd ... lus-module

 

[Jdruding]

So, to try to help me draw a conclusion...... My takeaway is that the altimeter is not exactly accurate. I also understand that drift will occur based on wind and temperature. What i don't quite understand is if my situation is still within the norm. It is currently around 30 deg F here and I'm taking the Phantom outside from my 70 deg F house. When i start it up, on the ground, it usually reads about 15 ft. If I fly it around for a few min, then land it, it usually shows around 30 ft. Restarting it seems to get it back down to 1-5 ft. Flying around again for few min and then landing, it will now be around 15-20ft. So it has continued to drift upward, but at a slower pace. You can repeat the process and the variance will be less and less, until my battery runs out.

So, is it normal for the altimeter to continue to drift upward throughout the entire flight time, when the temp is around 30 deg F, coming from a 70 deg home?

 

[doug86]

(especially due to accuracy constraints imposed by the US gov't)

Can you back up this claim of constraints by US gov't? Lets not use this forum to spread rumors, ok?

They turned off SA years ago, and the US Gov't does not do any to degrade the GPS signal: see http://www.gps.gov/systems/gps/modernization/sa/

Selective Availability (SA) was an intentional degradation of public GPS signals implemented for national security reasons.

In May 2000, at the direction of President Bill Clinton, the U.S government discontinued its use of Selective Availability in order to make GPS more responsive to civil and commercial users worldwide.

The United States has no intent to ever use Selective Availability again.

In September 2007, the U.S. government announced its decision to procure the future generation of GPS satellites, known as GPS III, without the SA feature. Doing this will make the policy decision of 2000 permanent and eliminate a source of uncertainty in GPS performance that had been of concern to civil GPS users worldwide.