Altitude reading way off

[colonel00]

I'm sure this has come up at some point but I couldn't find a relevant thread off hand. Anyway, after flying around for a while, I had the Mavic back hovering at eye level (between 5 and 6 feet) and my indicated altitude is off by 10+ feet. I can understand a little deviation but this seems to be unacceptable, especially with good GPS and 15 satellites. I know the barometer plays a role in the altitiude but it shouldn't be this far off and GPS and VPS should be adding their input to correct the issue. This is hovering right over the home point too.

 

[arcolog2]

Did you walk at all? Can't tell on map, but you are at your exact home location right?

Sent from my SM-N910V using Tapatalk

 

[Logger]

Looks about right to me. Says it is 5.2 feet AGL and 16' above your home point which is 123' or 40 metres away.

The indicated "Height" is referenced to where you took off. Not not where you are standing right now.

 

[arcolog2]

Looks about right to me. Says it is 5.2 feet AGL and 16' above your home point which is 123' or 40 metres away.

The indicated "Height" is referenced to where you took off. Not not where you are standing right now.

Yea what he said, that's what I was getting at, didn't really look at your screen data though haha

Sent from my SM-N910V using Tapatalk

 

[colonel00]

Did you walk at all? Can't tell on map, but you are at your exact home location right?

Sent from my SM-N910V using Tapatalk

Now that I think about it, I had walked down the upper path a little bit. In the photo, the path over my left shoulder. However, this path is fairly level so gain, I could accept a few feet here or there but 10+ feet seems a bit much.

 

[colonel00]

Looks about right to me. Says it is 5.2 feet AGL and 16' above your home point which is 123' or 40 metres away.

The indicated "Height" is referenced to where you took off. Not not where you are standing right now.

Right, that's what's odd is those should be very similar. Here's a photo from when I first took off a while before. Altitude and VPS are still off by a couple of feet but I guess that's better?

I'll do a more thorough test tomorrow and see if I can recreate the issue but this isn't the first time that I've see quite a deviance in the altitude reading.

 

[Logger]

Given that they are two totally different systems, using totally different methodology, to indicate height above different datums, I suspect you will struggle to find a real issue here. " GPS and 15 satellites." being totally irrelevant to the calculations as well.

Just so you understand, the H value is a pressure altimeter/barometric height as compared to take off point = 0. Whereas VPS is an ultrasonic height above whatever it can sense beneath (Up to a max of 43'/13m.)

 

[colonel00]

Why is GPS irrelevant? To me, that is the most accurate measure of altitude since the barometer can be affected by climate change and the VPS is limited in range. I guess perhaps there should be bigger concern if we are relegated to relying on a barometer as our sole reference for altitude as that obviously has potential for numerous complications.

 

[msinger]

Why is GPS irrelevant?

GPS data is not used to estimate the altitude.

 

[colonel00]

GPS data is not used to estimate the altitude.

Is there any reason as to why this is? It just seems silly to use GPS to hold position in two planes but not the third. Or at least let GPS confirm what the barometer is reading. The problem with a barometer is it is affected by more than just the altitude so, depending on where you fly or the weather conditions, the air pressure can most definitely change and skew your altitude readings. Moreover, will DJI be able to interpret these incorrect data readings when assessing if a crash should fall under a warranty?

 

[msinger]

Is there any reason as to why this is?

I'm not sure. Maybe because it's more accurate in most cases?

will DJI be able to interpret these incorrect data readings when assessing if a crash should fall under a warranty?

The warranty does not cover pilot error. So, if you (for example) fly into the ground because you're expecting the altitude to be dead accurate, DJI will likely not repair your Mavic under warranty.

 

[BudWalker]

Why is GPS irrelevant? To me, that is the most accurate measure of altitude since the barometer can be affected by climate change and the VPS is limited in range. I guess perhaps there should be bigger concern if we are relegated to relying on a barometer as our sole reference for altitude as that obviously has potential for numerous complications.

I've wondered about this also. It's the same with the P3. The gps altitude is a very noisy signal and can't be used for for navigation. That's why baroAlt is used for navigation. But, gpsAlt is accurate in the sense that a smoothed gpsAlt will almost always be closer to the actual altitude. Why can't a smoothed gpsAlt value be used to correct baroAlt? I dunno.

One complication would be an RTH at the end of a flight where there has been a barometric pressure change. I suspect that's the reason that the home point altitude is set to be 20 meters higher than the altitude at take off.

 

[FlashGJ]

GPS altitude at times will only be accurate to within +\- 15 meters (50 ft). It is an inherent weakness in the system and the accuracy will very depending upon number of satellites, signal strength, even geographic location can have an effect. That is a worse case scenario and at times it may seem to be much more accurate, but you can never fully rely on altitude displayed by GPS Units.

 

[leesmavic]

By the nature of GPS as a positioning system it was designed to accurately position 'on the ground'. GPS works by measuring distance to 3+ satellites with known coordinates. Position is identified as a crossing point of three spheres. It gives two points where one is either deep underground or very high in the sky. While GPS accuracy in horizontal pane is close to 1-2 meters, vertical accuracy is 2 to 5 times less, meaning 2-10 meters off of actual position.
Most altimeters are based on barometers as those give a way better accuracy. Radio altimeters improve on that further. VPS is there to make sure that 1 meter deviation will not be compensated for at half meters above ground.

Sent from my XT1572 using MavicPilots mobile app

 

[f3honda4me]

The barometer is what is used to determine height above/below the home point and is the primary height displayed on the app. The VPS is used by the craft to determine height when the craft is within 30 feet of the ground. (it is also displayed in smaller font on the app) As was already mentioned, GPS is not used for height at all.

 

[Logger]

Why is GPS irrelevant? To me, that is the most accurate measure of altitude since the barometer can be affected by climate change and the VPS is limited in range....

That made me chuckle - climate change occurring in Mavics 27 minutes flight time now

The problem with a barometer is it is affected by more than just the altitude so, depending on where you fly or the weather conditions, the air pressure can most definitely change and skew your altitude readings...

Lets put this in perspective for a minute. This thing has a range of 13km and an endurance of 27min at best. Typically most flights will be within a much closer range. Say 5km and 20 minutes. How much do you realistically expect the atmospheric pressure to change at your location in this sort of time window?
As an example lets imagine a massive cold front passes through with the barometric pressure falling extremely rapidly (and you are stupid enough to go flying our drone), it may drop 6mb in 3 hours. This Equates ~ 0.6 mb change in your 20 minute storm flight. This will cause the altimeter to over read by 20 feet and place the Mavic 20' lower than it should be! A highly unlikely worse case scenario and the 20' altimeter error will be the very least of your worries as you try to fly in gale force winds.

I just skimmed back through a heap of my flights that have used the same TO & LDG points. The recorded Landing height is invariably 0', -3.3', or -6.6' as compared to when it took off up to 25 minutes earlier max. This suggests to me the Baro altimeters granularity is only to the meter. In any case I see no significant changes there from Take of to LDG that would concern me.

In aviation terms, altimeter accuracy is usually ranked in the order; RAD ALT, Baro Altimeter then GPS as a last resort.

 

[colonel00]

Some interesting points and discussion.

Logger, yeah the climate change kinda sounds off but you are basically making my point of concern for me. How much do I expect the atmospheric pressure to change? Not that much really. However, something changed enough to affect the barometer to be off by over 11 feet. So based on your statements, the pressure changes should be quite small. If that's the case then what is causing the errors in altitude?

 

[Logger]

Some say the Mavic heats up and that effects the baro extremely slightly. But who cares. It is a total non issue for me. I am off to fly my drone.

 

[f3honda4me]

Some interesting points and discussion.

Logger, yeah the climate change kinda sounds off but you are basically making my point of concern for me. How much do I expect the atmospheric pressure to change? Not that much really. However, something changed enough to affect the barometer to be off by over 11 feet. So based on your statements, the pressure changes should be quite small. If that's the case then what is causing the errors in altitude?

There is no great mystery here. The height, according to the barometer, is based on the home point where the drone took off from. You were 123 feet away from the home point. The ground you were hovering over was probably higher than the ground at the home point you took off from.

 

[colonel00]

Look, I can understand a little skepticism. It wasn't a perfect test. So, I went out today to test again...

After a little flying, it was almost 10ft off. Granted, I was 8ft away from my Home Point

A little closer to the home point and its getting worse.

I can't get much closer and it's almost 13ft off.

And keeps getting worse...

By the time the battery was drained, it was almost 20ft off

I realize that this may not seem like a bit deal to some since you can still use your VPS but VPS has it's limitations and I've even had issues when flying over water so I may need to have it off.