Barometer way off

[RodPad]

The start of that record, when the reported altitude is dropping, looks like barometer stabilization after motor start. PhantomHelp doesn't report stick input but, judging from the battery voltages, actual takeoff looks to be at 16 s, and the altitude rises appropriately from there.

Yes, I agree.

Old but interesting,
Altitude......
To bad he never got back.

Another negative altitude.
Altitude inaccuracies during longer flight

Rod

 

[msinger]

Further, stating that they could be caused by an out-of-calibration or bad IMU appears to be an unreasonable assumption on your part unless the IMU has an influence on the barometer reading.

We really couldn't call this an assumption since it has resolved problems like this in the past.

 

[sar104]

We really couldn't call this an assumption since it has resolved problems like this in the past.

Agreed - it could perhaps resolve a problem, if there were a problem to begin with, and if the IMU calibration also temperature-calibrates the pressure sensor, but the initial observation does not even indicate a problem to be resolved since it is consistent with the pressure trend at that time and the accuracy of the sensor.

 

[msinger]

the initial observation does not even indicate a problem to be resolved since it is consistent with the pressure trend at that time and the accuracy of the sensor

I'm sure you realize all of this back and forth is meaningless since we do not have enough information. The OP needs to go out and take a few more flights to see if this issue can be reproduced.

 

[Chief Joe]

Agreed - it could perhaps resolve a problem, if there were a problem to begin with, and if the IMU calibration also temperature-calibrates the pressure sensor, but the initial observation does not even indicate a problem to be resolved since it is consistent with the pressure trend at that time and the accuracy of the sensor.

sar104, msinger is correct, and his suggestion has solved this error in the past. Also, I'm sure you realize that there really isn't a working and reporting barometer system on board the Phantom. It uses a barometric sensor to gauge the take-off point, then uses the IMU to adjust the height (among other flight parameters) indicated in the DJI GO app (or other app being used). The calculated IMU data is probably compared to the pressure sensor data at some point, but the pressure data isn't very accurate and isn't used to limit the altitude. There really isn't a constant altitude being calculated and displayed by the barometric sensor. I do not know for a fact that the GPS also contributes, but it is often used when installed on commercial and military aircraft - including helicopters, as a comparison tool. That is used to verify the IMU data and the barometric system data comparison to see which system to use. Depending on the needs of the crew, other systems can be brought into the mix to fine tune the position in the sky and over which patch of ground. The Phantom is a great aircraft, but just isn't always as sophisticated as we might like to think it is.

 

[N017RW]

...I'm sure you realize that there really isn't a working and reporting barometer system on board the Phantom. It uses a barometric sensor to gauge the take-off point, then uses the IMU to adjust the height (among other flight parameters) indicated in the DJI GO app (or other app being used). The calculated IMU data is probably compared to the pressure sensor data at some point, but the pressure data isn't very accurate and isn't used to limit the altitude. There really isn't a constant altitude being calculated and displayed by the barometric sensor.

You'll need to post a link or cite your source for the above statement before making such a declaration.

It simply contradicts all the info I have ever read regarding not only DJI MCs but all of the most common ones on the market.

Short of this I think you're mistaken.

 

[Tricky]

Did a imu calibration, on a cold machine. Altimeter reading from litchi and dji go off still ~3-5 feet.

 

[Richard R]

Did a imu calibration, on a cold machine. Altimeter reading from litchi and dji go off still ~3-5 feet.

Not surprising, IMU really doesn't have anything to do with altitude reporting. If you doubt that, just think about how the bird reacts when you lose GPS! it still holds altitude just fine, but drifts (the drifting and attitude is what the IMU controls). I doubt that DJI goes through the process of calibrating the altitude system in every bird that they build. More likely, more likely, they have a a canned table in FW that is used to relate resistance change in the barometer to changes in altitude and slight, normal variations in component values could throw those off slightly. How well does the bird hold a programmed altitude? For example, when I fly a programmed mission suing something like Maps Made Easy, i have compared the programmed altitude with that recorded in the photos and usually I'm within inches of the 'correct' altitude throughout the flight. And DJI now records the GPS MSL altitude so I pull the exif data from the ground reference photo and pics throughout the fight to determine an AGL height. if it's consistent, you can always compensate for the offset when actual altitude is that critical (maybe when shooting a POI). But if you are cutting obstacle avoidance that fine, you must really like living on the edge!

 

[msinger]

Did a imu calibration, on a cold machine. Altimeter reading from litchi and dji go off still ~3-5 feet.

3 feet isn't out of the ordinary. That's a lot better than the ~13 feet you originally reported.

 

[Tricky]

3 feet isn't out of the ordinary. That's a lot better than the ~13 feet you originally reported.

Yes, and also, I wasn't happy w the compass readings I was getting at this particular venue where the job was...as is said here often, "more testing necessary" - on the good news side, I flew all my missions without incident, just had to make allowances for the altitude discrepancies - thx all, as always

 

[N017RW]

The Baro sensor is not a simple transducer (i.e. pressure to resistance).
It is actually a digital device on a comm. or data buss.
Most likely i2C or CAN unless DJI has developed a proprietary one

 

[sar104]

sar104, msinger is correct, and his suggestion has solved this error in the past. Also, I'm sure you realize that there really isn't a working and reporting barometer system on board the Phantom. It uses a barometric sensor to gauge the take-off point, then uses the IMU to adjust the height (among other flight parameters) indicated in the DJI GO app (or other app being used). The calculated IMU data is probably compared to the pressure sensor data at some point, but the pressure data isn't very accurate and isn't used to limit the altitude. There really isn't a constant altitude being calculated and displayed by the barometric sensor. I do not know for a fact that the GPS also contributes, but it is often used when installed on commercial and military aircraft - including helicopters, as a comparison tool. That is used to verify the IMU data and the barometric system data comparison to see which system to use. Depending on the needs of the crew, other systems can be brought into the mix to fine tune the position in the sky and over which patch of ground. The Phantom is a great aircraft, but just isn't always as sophisticated as we might like to think it is.

No - that is completely incorrect. There is working pressure sensor, and it is the primary altitude diagnostic.