Smelly motor - how dangerous was this?

[BudWalker]

.................
Something puzzles me: I looked at the flight log, and I can't find an error on the flight where the prop got blocked. It's not recorded at all. Is there a separate error log I can find? Shouldn't there be something there with an error message? Also I had a flight indoors two days ago, and that flight isn't recorded in the log at all. Is that correct?
..................

That info will be found in the .DAT which is a file on the P3 itself. Have a look here:
GUI Version of DatConverter to see how to retrieve that .DAT and look at the info.

I'm quite interested in the .DAT for the flight where your motor overheated. If the .DAT has information about motor temperature that .DAT should make it easy for me to figure out where that info is. If you could retrieve that .DAT and post it via DropBox I'll analyze that flight for you.

Retrieving a .DAT from the P3
1. Go to the DJI GO App and connect to the P3
2. MC Settings --> select Advanced Settings Advanced Settings --> Enter flight data mode
3. wait until you get a dialog telling you that the P3 is in flight data mode.
4. Connect a USB cable from the USB plug in the nose of the P3 to your computer. This is not the USB connector on the side of the camera. Your P3 should have come with the required USB cable
5. The data drive should mount and be available.
6. Use the date on the FLYXXX.DAT files to pick the one you want and copy it to your computer

 

[frostacaust]

I'm flying again today for 11 time ..will let you later today if gravity wins.

 

[Mastiff]

Creative way of testing it, with the broom handle! I didn't dare to do something like that because I was afraid the drone could break free from the legs! Is it safe to strap it down only through the legs?

 

[Mastiff]

BudWalker, didn't see your post there. I will try to get that out when I come back from my beach trip.

 

[frostacaust]

Well now....I'm only letting it idle and give it a little throttle once in awhile ..listening the motor , checking for heat build up ...etc...

 

[Mastiff]

BudWalker, I now have done the full procedure, and I have a project in DashWare. I have some motor graphs like this. Is that what you were looking for? I saw a battery temp gage, but not a graph showing the temp of the engines. Maybe the red curve is it? The left rear engine shows a rather different curve at the end then the other ones!

 

[N017RW]

There are no sensors for motor temp.

The red graph above is labeled as CMD or [throttle] command.

 

[Mastiff]

Aha, OK. I guess that means commands from the ESC to the motor, right? Or else it would have been equal between the motors.

 

[BudWalker]

View attachment 43961 BudWalker, I now have done the full procedure, and I have a project in DashWare. I have some motor graphs like this. Is that what you were looking for? I saw a battery temp gage, but not a graph showing the temp of the engines. Maybe the red curve is it? The left rear engine shows a rather different curve at the end then the other ones!

Oops, sorry, I should have provided more of an explanation. I developed DatCon, the app you used to convert the .DAT into a .csv. (@Luap developed the Dashware profile to convert the .csv to a Dashware presentation).I did that by reverse engineering the .DAT file. If I knew what kind of signal to expect then I would look through the .DAT with tools I've developed to find data that would match that expected signal. If you could provide that .DAT then I'd look to find a motor temp signal if it exists. If found, then I'd add that field to DatCon.

 

[KyleMaxxUAV]

Hey Bud,
Is that process pretty resource hungry ? What I'm wondering is, I have an older laptop that I want to use solely for running your & luap's setup on but it's older and let's just say "not the quickest" lol. Will it run just fine or do I need a fairly decent computer to run all that through ??


Sent from my iPhone using PhantomPilots mobile app

 

[BudWalker]

There are no sensors for motor temp.

The red graph above is labeled as CMD or [throttle] command.

Are you saying this because it's evident from the hardware? I've never been inside the P3. If there were a temp probe on the motor there would have to be a wire (or wires).

The motor cmd value isn't derived from the throttle setting.

 

[BudWalker]

Hey Bud,
Is that process pretty resource hungry ? What I'm wondering is, I have an older laptop that I want to use solely for running your & luap's setup on but it's older and let's just say "not the quickest" lol. Will it run just fine or do I need a fairly decent computer to run all that through ??


Sent from my iPhone using PhantomPilots mobile app

Not too bad. On my machine it can take 30 seconds to convert a 300 MB .DAT. Whatever machine you have it's almost certainly to be faster running DatCon instead of uploading to flylog.info. Also, DatCon has more features.

 

[KyleMaxxUAV]

Are you saying this because it's evident from the hardware? I've never been inside the P3. If there were a temp probe on the motor there would have to be a wire (or wires).

The motor cmd value isn't derived from the throttle setting.

There are no extra wires for a temperature probe coming from each motor. Just the 3 power wires.


Sent from my iPhone using PhantomPilots mobile app

 

[KyleMaxxUAV]

Not too bad. On my machine it can take 30 seconds to convert a 300 MB .DAT. Whatever machine you have it's almost certainly to be faster running DatCon instead of uploading to flylog.info. Also, DatCon has more features.

Perfect!!! I'm going to give this a shot !!

Thanks again Bud !!!



Sent from my iPhone using PhantomPilots mobile app

 

[alokbhargava]

If the motor was blocked why would the esc be fried? Doesn't make any sense, esc is not a moving part. It just sends a signal to motor to run. If the motor cant run why would the esc be damaged.

ESC does not send signal to motors, it actually supplies current to motors. In case of overload and short-circuits, its expected to limit the current and limiting of the current depends on the line inductance. If its a dead short, even an ESC might burn before it could limit the current.

 

[Mastiff]

Sorry, BudWalker, I misunderstood. Corrected in PM.

alokbhargava, that board was what I first feared was damaged, but it's not. Not at all. Luckily. Because if I damaged something for several hundred dollars in my first week with the drone my wife would have been a widow...by her own hand!

 

[Luap]

View attachment 43961 BudWalker, I now have done the full procedure, and I have a project in DashWare. I have some motor graphs like this. Is that what you were looking for? I saw a battery temp gage, but not a graph showing the temp of the engines. Maybe the red curve is it? The left rear engine shows a rather different curve at the end then the other ones!

Try converting from motor start to recording stop then check the motor graphs again.
In this case it doesn't matter if its not Dashware compatible. Not Dashware compatible means it will not show a proper map.

 

[BudWalker]

Thank you @N017RW and @P3P Canada for pointing out that a motor temp value in the .DAT is unlikely. @Mastiff had provided the .DAT so I looked at the flight anyway. Very strange. It would appear that after the pilot did a CSC and the motors were commanded to stop that the LeftBack motor continued to run for another 20 seconds. Two possibilities come to mind; 1) it's a sensor error and it didn't really happen that way, or 2) the motor obstruction damaged the ESC and it stayed on.

By default DatCon sets the TimeAxis parameters so that the upper end of the time interval coincides with motorStop. To see this event set the upper end to Recording Stop

I also set the lower end to 100 seconds to just produce data that focuses on this event.

The eventLog shows an ESC stall at 108.50 which I am supposing coincides with the motor obstruction. The pilot initiated the CSC at about the same time.
108.500 : 101622 : 8269 [esc_is_stall] status changed: last(0x00000000) != current(0x00000002)
109.600 : 102282 : 8324 [esc_is_stall] status changed: last(0x00000002) != current(0x00000000)
109.818 : 102413 : 8335 [M. Stop]REQ_RC_NORMAL
109.818 : 102413 : 8335 Motor Start 1 Total 109.88
109.818 : 102413 : 8335 CTRL reset all by motor stopped
121.820 : 109614 : 8935 Battery power off
122.820 : 110214 : 8985 Battery power off
123.900 : 110862 : 9039 Battery power off
125.420 : 111774 : 9115 Battery power off
126.060 : 112158 : 9147 Battery power off
127.140 : 112806 : 9201 Battery power off
128.118 : 113393 : 9250 [ACT.msg] ACT_AUTO_LANDING can't exec auto-landing while motor off
128.218 : 113453 : 9255 [LED] changed: battery error!

The following plot shows that the LB motor continued to run even though it was being commanded to stop. It also show the load greater than 0.

 

[Mastiff]

BudWalker, I can assure you that all engines stopped at the same time! The locked up motor did not run any longer then the other ones, and this I know because the squealing in agony stopped as soon as I did the CSC. So I'm guessing possibility 1. Also there are no after effect, and no error messages from the ESC after four "missions", so that's not at all likely.

 

[BudWalker]

BudWalker, I can assure you that all engines stopped at the same time! The locked up motor did not run any longer then the other ones, and this I know because the squealing in agony stopped as soon as I did the CSC. So I'm guessing possibility 1. Also there are no after effect, and no error messages from the ESC after four "missions", so that's not at all likely.

OK, I guess it must be data and how it gets generated. We've used the speed and load data on several occasions and it has always been consistent with other measurements and the pilot's description of events. Don't know what happened here.