New version of .DAT converter

There is a field in the .DAT that I think is the IMU temperature. It always starts at the ambient temperature and then climbs at a rate of .27 degreesC/second until it reaches 65 degreesC and stays there for the remainder of the flight. This warm up can take as long as 2.5 minutes to reach 65 degreesC which means that the A/C thinks the IMU is warm enough to fly when it is still on this ramp on the way to 65 degreesC. I would have thought the IMU has to be at a particular constant temperature in order to work. Anybody know something about this?
 
Last edited:
65 C seems awfully hot. Can that be correct?
It does seem hot. This signal is actually seen in a couple of places in .DAT. One is a float. The other is 2 byte integer that is 100 times the temperature. 2234 = 22.34 C. The reason I think it's a temperature is that it always starts at the ambient temperature. Except for one time when the P3 was started after being off for 3 or 4 minutes and the starting temperature was higher.
This field will be in the next release; just trying to decide if it's labeled "IMUTemp" or "PossibleIMUTemp". I can make an experimental version of DatConverter if you want to take a look before the release.
 
It does seem hot. This signal is actually seen in a couple of places in .DAT. One is a float. The other is 2 byte integer that is 100 times the temperature. 2234 = 22.34 C. The reason I think it's a temperature is that it always starts at the ambient temperature. Except for one time when the P3 was started after being off for 3 or 4 minutes and the starting temperature was higher.
This field will be in the next release; just trying to decide if it's labeled "IMUTemp" or "PossibleIMUTemp". I can make an experimental version of DatConverter if you want to take a look before the release.

65C seems like a typical processor, cpu, or similar temperature. I have been trying to find a tear down of the phantom 3 that identifies the chips used to try to find on with that operating temperature....

That would also explain the lack of delay before you can fly...
 
More on the mystery temperature field. Below is a plot of this field. Once it climbs to 65C it looks to me like the temperature is regulated. It's always 65C. Doesn't matter what the ambient is and it doesn't vary. This implies that there must be a heater whose only purpose is to keep the temp at 65C. I agree that 65C seems hot. I suppose there could be a different way of converting this value to Celsius (the initial value is the ambient temp expressed in Celsius) but it would be easier for me to accept that this temp is actually 65C.
capture-20151213-081416.jpg
 
BudWalker, In the comprehensive CSVs from the app, these are some of the flight modes I found (2 columns).

OSD.flycState.raw OSD.flycState
1 Atti
6 GPS_Atti
10 AssistedTakeoff
12 AutoLanding

There are certainly more such as FailSafe and GoHome.
I haven't forgotten this. There is a bunch of this kind of data in the text messages that occur in .DAT. It's organized differently than the numeric data and I haven't figured out a way to make it play nice with the way the .csv is organized. I may just throw it into some columns in .csv and hope that it's useful.

What I haven't found is any kind of signal that occurs when the P3 decides it needs to switch from P-GPS to P-ATTI, i.e. not because the mode was switched on the RC or some other pilot directed action. Does anybody have a .DAT file where this has occurred? If I could get that it'd help. If you had the associated .txt from the DJI Go app that'd help even more. Thanks
 
I haven't forgotten this. There is a bunch of this kind of data in the text messages that occur in .DAT. It's organized differently than the numeric data and I haven't figured out a way to make it play nice with the way the .csv is organized. I may just throw it into some columns in .csv and hope that it's useful.

What I haven't found is any kind of signal that occurs when the P3 decides it needs to switch from P-GPS to P-ATTI, i.e. not because the mode was switched on the RC or some other pilot directed action. Does anybody have a .DAT file where this has occurred? If I could get that it'd help. If you had the associated .txt from the DJI Go app that'd help even more. Thanks
In a CSV column it'd be great.
As for the logs; doesn't this happen when you cover the P3? You could even do this when stationary I guess (engines running but not flying and then just cover the P3). Or am I mistaken?
 
In a CSV column it'd be great.
As for the logs; doesn't this happen when you cover the P3? You could even do this when stationary I guess (engines running but not flying and then just cover the P3). Or am I mistaken?
Attached to this post is a DatExp.zip. Inside that zip you'll find a DatExp.exe which is an experimental version of the next release. This DatExp will produce a .csv with 6 columns on the end flightMode, flightMode.String, flightRegime, flightRegime.String, navMode, and navMode.String. A single flyState (like in your example) doesn't make sense to me, e.g. auto takeoff can occur in both atti and gps_atti flight modes. Give it a go and lemme know what you think.
 

Attachments

  • DatExp.zip
    232.2 KB · Views: 173
I've been working with a user who has been having a lot of difficulty getting DatConverter installed so that it works for him. That got me to wondering if others have been unable to get DatConverter to work and I just haven't heard about it. I'd really appreciate any information about installation difficulties. Thank you.
 
A new version of DatConverter is being released in conjunction with the Dashware profiles that @Luap has posted P3A/P Dashware profile for for Dat Files | DJI Phantom Forum These profiles require that you use this new version of DatConverter.

The new version can be obtained at Fly.zip .Unzipping that file will produce three files; DatConverter.exe, Installing and Using DatConverter, and Release Notes.

Please report any bugs or comments here. You can also PM me.

A copy of the Release Notes follows

Dashware changes
Many of the changes since the last release are due to Dashware limitations. . This release is closely tied with the Dashware profiles that can be downloaded by clicking here. To use these Dashware profiles you must use this version of DatConverter. Invoking DatConverter with no arguments except the file name of the .DAT file will produce a .csv that is suitable for these Dashware profioles.

In a P3 flight the initial data may not be valid and will cause problems when seen by Dashware. E.g. latitude-longitude data doesn’t become valid until there are enough satellites. DatConverter can now determine which data is not valid and should not be written to the .csv and seen by Dashware.

The .csv file can become quite large causing Dashware to choke. DatConverter can now reduce the size of the .csv by using a lower sampling rate. For a more in depth discussion about sampling rate consult the accompanying Installing and Using DatConverter document.

Dashware can be painfully slow when computing new values from data it reads in the .csv. Several new computed fields have been added to DatConverter so that Dashware doesn’t have to compute these values.
Other changes
New fields - Several new fields have been discovered in the .DAT and added.
Motor speed data - commanded speed, actual speed, and load. The speed data is probably not RPM but, rather, a constant times RPM.
RC control – mode switch in addition to the existing aileron, elevator, rudder, and throttle data.
Mystery temperature data – This may be IMU temperature. At issue is that it heats to 65C which seems to hot.
HomePoint – both location and distance from.
Flight control error data – The error term for the model based flight control algorithm.
Flight Mode – atti, or GPSatti
Flight Regime – engineStart, asstTakeoff, autoTakeoff, autoLanding
Navigation Mode – goHome, waypoint, followMe, hotpoint

Text Log Output

The.DAT has a surprising amount of data that is text based. This data can be seen either by including it in the .csv, or writing another file, or both. See the description of the ‘-tlo’ command line argument in the accompanying Installing and Using DatConverter document.
 
Bud - this update is great!
I have flights of up to 29 or 30 minutes now with auxiliary batteries. I notice that the DAT files get split at 307,204 KB. So I now have 2 DAT files for one flight. They have a different filename (FLY057 and FLY058), but have the exact same time stamp.
How can these be merged and used? I can send you some examples by WeTransfer if you'd like (PM me).
 
Bud - this update is great!
I have flights of up to 29 or 30 minutes now with auxiliary batteries. I notice that the DAT files get split at 307,204 KB. So I now have 2 DAT files for one flight. They have a different filename (FLY057 and FLY058), but have the exact same time stamp.
How can these be merged and used? I can send you some examples by WeTransfer if you'd like (PM me).
Please upload both FLY057 and FLY058 to flylog.info so that I can take a look.
 
A new version of the .DAT file converter has been deployed at flylog.info. The offline version can be obtained by downloading Fly.zip from either the Phantom Pilots Wiki at Fly.zip or from Dropbox - Fly.zip. Unziping Fly.zip will produce DatConverter, DatKML, and some release notes.

Among other changes the new version provides motor speed data and control stick input data. Also, handling partial or missing records in the .DAT file is more robust.

Unless there are objections I plan to phase out the online version and rely solely on the offline version to deploy future versions.

Please report any bugs here. You can also PM me.


The files you listed do not load, 404 error, i've managed to download the dat files to my computer, install dashware with the profile , but do not know how to convert the dat file to csv... can you help?
 
The files you listed do not load, 404 error, i've managed to download the dat files to my computer, install dashware with the profile , but do not know how to convert the dat file to csv... can you help?
Sorry, but that stuff is way out of date. DatCon can now be found here

DatCon Downloads
 

Members online

No members online now.

Forum statistics

Threads
143,066
Messages
1,467,358
Members
104,936
Latest member
hirehackers