Altitude incorrect

[Maddog]

In spite of all the technical jargon above, I was agreeing with you. And the point remains that GPS alt is not used in the Solo and for all of the above reasons you stated. When you are dealing with AGL altitudes of less than 400' and consumer grade GPS receivers, there is simply no place for it. As far as barometers failing, that's why there are 2 in the Solo.

Where are they located and what do they look like?
Anybody have pics?

 

[ajdelange]

The Solo doesn't use GPS for altitude. That is known. The reason is because of poor accuracy.

Even when GPS is demonstrably better? I recorded barometric pressure using a Honeywell Precision Barometer every 15 seconds for 23 Dec 2012 to 7 Jan 2013. Over that data set barometric pressure changed by an average of .3 mB in 15 minutes and by as much as 2 mB (front must have rolled through on Boxing day). On 61% of 15 minute flights made over that time period the reading of a high accurcay barometer would have changed by less than 5.5 ft but on 16% of them it would have changed by more than 11 feet, on 7% by 14 feet and, if you flew on Boxing day, by as much as 55 feet but such events are rare with more than 20 ft discrepancy between beginning and end of flight occuring less than 1% of the time in this data set.

The Solo doesn't use GPS for altitude. That is known. The reason is because of poor accuracy.

At least that's what the developers have openly discussed on diydrones and github. While everything you have said is no doubt true, I think there might be a difference between what "should" be possible in theory versus what actually works in the field.[/QUOTE] That's why I offered real, if anecdotal, data rather than theoretical arguments. The argument you should be making is that what I see in the 'lab' and what I can realize in a flying platform may not be the same.

I don't know. I do know that GPS on the Solo (and any drone) is far less accurate near the ground

No, you don't know that. A simple experiment with your smart phone should give you data better than what I got if you go outside and what I got inside was better than what I would have gotten with the altimeter on Boxing day 21012.

It would seem unwise to use any data from the GPS, even as a sanity check,

It is unwise to close your mind based on what you have read on the internet about something you don't apparently know that much about. The small amount of real physical evidence I have presented is not enough to draw a conclusion from, of course, but it should be enough to stimulate some thinking.

...when you can put 2 or 3 barometers on a drone.

You can put 40 barometers on a drone but that won't keep the barometric pressure from changing as the weather patterns do. I never would have guessed that the average change in barometric pressure over 15 minutes is as much as 0.3 mB but I guess it is and while that may not be what Ducj5150 is seeing (could be a hardware problem) it is not unreasonable to suppose that just the weather could be at least partially responsible. This suggests that we should perhaps check the glass trend before we fly! http://www.3drpilots.com/members/ducj5150.3824/

 

[ajdelange]

Where are they located and what do they look like?
Anybody have pics?

More to the point: what are their specs?

 

[ajdelange]

So now your just trolling right? Barometers are much more accurate for measuring relative changes in altitude.

I guess that's why surveyors use them instead of GPS for making contour maps. Reminds me of the old PHD qualifying exam question about naming four ways to determine the height of a building with a barometer. In the hopes of lightening up here a little I'll give the answers:
1) Take it to the top of the building and measure the pressure. Take it to the ground and measure the pressure. Stick the difference in the ICAO standard atmosphere equation and if its a standard day you have the height.
2) Measure the length of the barometer, tie it to a string and lower it to the ground counting out how many barometer lengths of string you paid out.
3) Throw it off the roof and measure the time it takes to hit the ground. h = (9.8*9.8)*t/2
4) Take the barometer to the building engineer's office and tell him that if he'll show you the drawings you'll give it to him.

 

[ajdelange]

In No. 22 I suggested that we might want to look at the barometer trends before we fly. That sent me back to the 2012 data I have (near DC which is why I am in here typing instead of outside flying). I picked out several of the trends in that data set and thought (hope) they might be of some interest WRT this question. A trend represents an appreciable period of time over which barometric pressure varies linearly with time (plots as a straight line on a curve). Each up trend is followed by downward pressure change (but it may be to jaggy for me to have considered it linear and, of course there are times when the pressure is stagnant or slowly changing too. Here are some of the data.
Trend 0.736 mB/hr ( 5.02 ft/15min); Duration of trend: 9.05 Hrs Starting at 2012-12-22 12:55:07
Trend -1.218 mB/hr (-8.32 ft/15min); Duration of trend: 3.47 Hrs Starting at 2012-12-23 10:13:28
Trend 0.306 mB/hr ( 2.09 ft/15min); Duration of trend: 15.10 Hrs Starting at 2012-12-23 17:32:58
Trend -1.081 mB/hr (-7.38 ft/15min); Duration of trend: 3.95 Hrs Starting at 2012-12-24 10:58:33
Trend 0.773 mB/hr ( 5.28 ft/15min); Duration of trend: 7.40 Hrs Starting at 2012-12-25 01:03:27
Trend 0.609 mB/hr ( 4.16 ft/15min); Duration of trend: 7.22 Hrs Starting at 2012-12-25 14:18:28
Trend -2.650 mB/hr (-18.09 ft/15min); Duration of trend: 6.56 Hrs Starting at 2012-12-26 09:51:36
Trend 1.217 mB/hr ( 8.31 ft/15min); Duration of trend: 7.54 Hrs Starting at 2012-12-27 02:24:49
Trend 0.990 mB/hr ( 6.76 ft/15min); Duration of trend: 10.46 Hrs Starting at 2012-12-27 13:44:36
Trend -1.496 mB/hr (-10.21 ft/15min); Duration of trend: 3.61 Hrs Starting at 2012-12-29 03:17:58
Trend -1.741 mB/hr (-11.89 ft/15min); Duration of trend: 3.32 Hrs Starting at 2012-12-29 10:11:23
Trend 1.325 mB/hr ( 9.05 ft/15min); Duration of trend: 3.61 Hrs Starting at 2012-12-29 17:29:37
Trend 0.901 mB/hr ( 6.15 ft/15min); Duration of trend: 5.63 Hrs Starting at 2012-12-29 21:32:43
Trend 1.385 mB/hr ( 9.46 ft/15min); Duration of trend: 4.48 Hrs Starting at 2012-12-30 05:21:15
Trend 1.057 mB/hr ( 7.22 ft/15min); Duration of trend: 2.82 Hrs Starting at 2012-12-30 14:31:58
Trend -1.321 mB/hr (-9.02 ft/15min); Duration of trend: 3.01 Hrs Starting at 2012-12-31 10:12:51

The message here is that there are fairly frequent windows of several hours duration in which we might expect a barometric altimeter reading to differ by as much as 18 feet from takeoff to landing in a 15 minute flight. But we wouldn't have been flying that day because it was pretty windy and there was some snow and rain (and there were no SOLOs).

I guess the other part of the message is that if you see or have seen inconsistent altitude between beginning and end of flight you can go to Weather History & Data Archive | Weather Underground and check the barometer readings at the relevant time and place to estimate the trend and compare to the examples here.

 

[erikgraham]

I can't read all this. The barometers are more accurate measure of relative altitude. This has been thoroughly tested and proven in the drone industry for years now.

 

[Jubalr]

I can't read all this. The barometers are more accurate measure of relative altitude. This has been thoroughly tested and proven in the drone industry for years now.

Exactly. And the reason is that it has nothing to do with accuracy over time or even '15 minutes'. For a drone it's about accuracy second by second, and the ability to maintain altitude at a given moment to moment. Imagine that you were using GPS for altitude and relying on it to maintain precise altitude.

 

[Johnny Baer]

I'm not trying to address all the ideas expressed here but I have observed that my SOLO, which I often fly in the same location, a wide open approx 30 acre level field which is kept mowed, will, as I fly away from the takeoff point, display a somewhat disconcerting tendency to lose altitude (rather slowly) as it flies away from HOME.

Say I set the altitude at approx 6' and fly outbound across the field, by the time I'm maybe 400' away SOLO will lose virtually all it's altitude. As I said, this place is level but I always must increase altitude during the flight. It isn't an optical illusion because I can see on the GCS display that the ground is getting lower. Haven't paid attention to the telemetry yet so I can't address how it relates to what I'm seeing.

I intend to do some experimentation and take measurements and screen shots to see exactly what is going on but this has been a factor for a good while now. It just hasn't been critical enough to bother with calling 3DR yet but when I saw this thread I thought I'd mention it to see if anyone else has experienced a similar situation.

 

[erikgraham]

I'm not trying to address all the ideas expressed here but I have observed that my SOLO, which I often fly in the same location, a wide open approx 30 acre level field which is kept mowed, will, as I fly away from the takeoff point, display a somewhat disconcerting tendency to lose altitude (rather slowly) as it flies away from HOME.

Say I set the altitude at approx 6' and fly outbound across the field, by the time I'm maybe 400' away SOLO will lose virtually all it's altitude. As I said, this place is level but I always must increase altitude during the flight. It isn't an optical illusion because I can see on the GCS display that the ground is getting lower. Haven't paid attention to the telemetry yet so I can't address how it relates to what I'm seeing.

I intend to do some experimentation and take measurements and screen shots to see exactly what is going on but this has been a factor for a good while now. It just hasn't been critical enough to bother with calling 3DR yet but when I saw this thread I thought I'd mention it to see if anyone else has experienced a similar situation.

People have experienced this for sure. Especially when flying fast. Having said that, 400 feet is a long way to judge flatness. There could easily be a 4 or 5 foot variance and you'd never notice it. Try walking out to the spot 400 feet away and then fly back to where you usually launch from. See if the Solo appears to gain altitude.

 

[Johnny Baer]

You're right about judging altitude at that distance. However it starts descending at no more than about 100-150' and I keep up with altitude via watching the display on my controller. You can see the grass getting closer as it heads outbound. I'm mostly experimenting with FPV by using only the tablet or phone for reference with intermittent direct visual cross-checking when I notice this. It's not particularly dangerous but it is very noticeable.

 

[erikgraham]

You're right about judging altitude at that distance. However it starts descending at no more than about 100-150' and I keep up with altitude via watching the display on my controller. You can see the grass getting closer as it heads outbound. I'm mostly experimenting with FPV by using only the tablet or phone for reference with intermittent direct visual cross-checking when I notice this. It's not particularly dangerous but it is very noticeable.

I'm not talking about judging altitude. I was referring to judging the flatness of the terrain itself.

It may look flat to you, but there could easily be a slope in the terrain. Over a distance of 400 feet, the terrain could slope up 6 feet and it would still appear flat to you.

So what you think of as a loss of altitude could be backwards. The Solo isn't descending, the ground is slowly rising up to meet it. That's why I propose walking out into that field and flying back toward your normal takeoff spot. If the Solo still appears to descend, then you've confirmed that it's loosing altitude. If it appears to gain altitude, then you've confirmed there is a slight grade to the field.

 

[ajdelange]

It may look flat to you, but there could easily be a slope in the terrain. Over a distance of 400 feet, the terrain could slope up 6 feet and it would still appear flat to you.

That's a 1.5% grade and ought to be detectable with a decent clinometer or by sighting along the edge of a carpenter's level.

 

[Maddog]

I'm not trying to address all the ideas expressed here but I have observed that my SOLO, which I often fly in the same location, a wide open approx 30 acre level field which is kept mowed, will, as I fly away from the takeoff point, display a somewhat disconcerting tendency to lose altitude (rather slowly) as it flies away from HOME.

Say I set the altitude at approx 6' and fly outbound across the field, by the time I'm maybe 400' away SOLO will lose virtually all it's altitude. As I said, this place is level but I always must increase altitude during the flight. It isn't an optical illusion because I can see on the GCS display that the ground is getting lower. Haven't paid attention to the telemetry yet so I can't address how it relates to what I'm seeing.

I intend to do some experimentation and take measurements and screen shots to see exactly what is going on but this has been a factor for a good while now. It just hasn't been critical enough to bother with calling 3DR yet but when I saw this thread I thought I'd mention it to see if anyone else has experienced a similar situation.

Are you noticing the altitude loss on the Solo app, or by visually watching Solo?

 

[Johnny Baer]

I'm not talking about judging altitude. I was referring to judging the flatness of the terrain itself.

It may look flat to you, but there could easily be a slope in the terrain. Over a distance of 400 feet, the terrain could slope up 6 feet and it would still appear flat to you.

So what you think of as a loss of altitude could be backwards. The Solo isn't descending, the ground is slowly rising up to meet it. That's why I propose walking out into that field and flying back toward your normal takeoff spot. If the Solo still appears to descend, then you've confirmed that it's loosing altitude. If it appears to gain altitude, then you've confirmed there is a slight grade to the field.

Gotcha. You are totally correct. That's why I mentioned that I must do some more precise, measurable observations on the actual field, as well as taking screen shots of the telemetry to see if there are variations there that might indicate a problem with the barometric altimeter . . .or something else that I have yet to think of. Gotta get a bit more scientific in my observations. Tomorow looks like excellent weather here so If I can squeeze in some flying I'll take some measurements and also see what a 6' level turns up.

 

[Johnny Baer]

Are you noticing the altitude loss on the Solo app, or by visually watching Solo?

See my comment to enkgraham. In short, so far it's been visual.

 

[erikgraham]

Gotcha. You are totally correct. That's why I mentioned that I must do some more precise, measurable observations on the actual field, as well as taking screen shots of the telemetry to see if there are variations there that might indicate a problem with the barometric altimeter . . .or something else that I have yet to think of. Gotta get a bit more scientific in my observations. Tomorow looks like excellent weather here so If I can squeeze in some flying I'll take some measurements and also see what a 6' level turns up.

I'll be super curious to hear your results. By all accounts the faster you go the more altitude you loose, but i thought they fixed that a few updates ago. Once your sure your not flying up a slight grade, it will be really valuable to know how much altitude is lost at different speeds. I'd like to do some high speed low passes over water but I'm a bit nervous.

P.S. I'd attempt to do some tests myself but I live in mountainous terrain. There is nowhere flat to test.

 

[Jubalr]

I'll be super curious to hear your results. By all accounts the faster you go the more altitude you loose, but i thought they fixed that a few updates ago. Once your sure your not flying up a slight grade, it will be really valuable to know how much altitude is lost at different speeds. I'd like to do some high speed low passes over water but I'm a bit nervous.

P.S. I'd attempt to do some tests myself but I live in mountainous terrain. There is nowhere flat to test.

I did some testing on this when it came up about 6 months ago. What I found was what I expected in that altitude loss was based on the amount of nose down during faster runs. That is basic physics in that as the nose pitches forward, there is a point at which there is more thrust being generated forward than up to maintain altitude. It's less noticeable in stabilize and acro modes as those allow the pilot to apply more RPM's than the assisted modes do (although those can be adjusted some in the parameters). Basically in the assisted modes, about 28deg nose down is the limit before lift is reduced and altitude is lost.
Here are a couple of runs in MP to demonstrate the results.

40mph with a limit of 28deg nose down. No loss of altitude:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

35deg nose down, rapid loss of altitude.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[erikgraham]

I did some testing on this when it came up about 6 months ago. What I found was what I expected in that altitude loss was based on the amount of nose down during faster runs. That is basic physics in that as the nose pitches forward, there is a point at which there is more thrust being generated forward than up to maintain altitude. It's less noticeable in stabilize and acro modes as those allow the pilot to apply more RPM's than the assisted modes do (although those can be adjusted some in the parameters). Basically in the assisted modes, about 28deg nose down is the limit before lift is reduced and altitude is lost.
Here are a couple of runs in MP to demonstrate the results.

40mph with a limit of 28deg nose down. No loss of altitude:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

35deg nose down, rapid loss of altitude.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Thank you. So in loiter the top speed is well under 40mph. I understand about what you are saying about the issue not being speed but aoa, but loiter is so slow I can't imagine it would pitch forward enough and he shouldn't be getting any loss of altitude right? I mean, his aoa shouldn't even reach close to 28 deg, and certainly not exceed that??

 

[Jubalr]

Thank you. So in loiter the top speed is well under 40mph. I understand about what you are saying about the issue not being speed but aoa, but loiter is so slow I can't imagine it would pitch forward enough and he shouldn't be getting any loss of altitude right? I mean, his aoa shouldn't even reach close to 28 deg, and certainly not exceed that??

I was hitting 40 with the slider at the rabbit setting. But if his slider and speed is less, then he should not be losing altitude. I think his issue is a visual deception and rising ground. I have a spot like that where I fly too. Easiest test would be to take it up 80-100', give it full forward and watch the altitude in the app or controller. It should stay within 5' of where it started the run.

 

[Johnny Baer]

I took SOLO out for a couple of batteries today. Didn't have much time to do detailed tests but the bottom line is that my auto altitude control is all over the page. Everything was done at or just above walking speed so there's no effect from AOA.

To begin with I hovered at about 5-6'. SOLO descended, uncommanded to about 18". The app still indicated 6'. The app seemed to initially register pretty accurate altitude but SOLO'S altitude would change, almost always it descended and the app reported altitude never changed.

The site WAS modestly unlevel but the odd thing seemed that if I flew toward rising terrain SOLO maintained it's altitude relative to home and appeared to be descending, as would be appropriate. As I flew away toward (definitely) descending terrain as checked by sighting down a level, it DID descend and at about 300' from home it was down to about 3' agl. I walked out to it then flew it back to home walking with it. The entire time the app indicated 10' altitude. By the time I got back to home SOLO was at 6', or head high, app showed 10'.

When I just took off and hovered at any altitude of 10' or below it's altitude would vary by as much as 5-6'.

Under normal flight conditions this wouldn't make any difference, and likely wouldn't be noticed, but I'm flying in some tight areas near trees and obstructions using fpv and I find this disconcerting. I think I'll call 3dr and see what they say.