Mavic 2 wants a compass calibration 30 days after last one

[BudWalker]

This morning my Mavic 2 Pro said it wanted a compass calibration. It wasn't preventing the motors from being started. I couldn't imagine why a calibration was being requested. But, I went ahead and calibrated anyway.

After looking at the .DATs it would seem that the reason was that it had been 30 days since the initial calibration. There were several of these today from the eventLog stream

-47.676 : 3918 [L-COMPASS][mag_cali_pt] date_from_last 30|

But, none in the eventLog stream from yesterday.

Also, the results of the calibration didn't change significantly. From 30 days (2018-08-29, August has 31 days) the calibration gains and offsets

-35.448 : 4531 [L-COMPASS][scale cali(0)] estimation error:[15.0]
-35.448 : 4531 [L-COMPASS][scale cali(0)] succeed! bias:238.9 -148.1 -1314.6 scal:3.239 3.169 3.356|

From today after the calibration

-106.155 : 11073 [L-COMPASS][scale cali(0)] estimation error:[16.8]
-106.155 : 11073 [L-COMPASS][scale cali(0)] succeed! bias:235.1 -137.2 -1277.7 scal:3.210 3.154 3.280|

I received my M2 Pro very early. It'll be interesting to see if others start to see their Mavic 2 requesting a calibration after 30 days.

 

[Former Member]

This morning my Mavic 2 Pro said it wanted a compass calibration. It wasn't preventing the motors from being started. I couldn't imagine why a calibration was being requested. But, I went ahead and calibrated anyway.

After looking at the .DATs it would seem that the reason was that it had been 30 days since the initial calibration. There were several of these today from the eventLog stream

-47.676 : 3918 [L-COMPASS][mag_cali_pt] date_from_last 30|

But, none in the eventLog stream from yesterday.

Also, the results of the calibration didn't change significantly. From 30 days (2018-08-29, August has 31 days) the calibration gains and offsets

-35.448 : 4531 [L-COMPASS][scale cali(0)] estimation error:[15.0]
-35.448 : 4531 [L-COMPASS][scale cali(0)] succeed! bias:238.9 -148.1 -1314.6 scal:3.239 3.169 3.356|

From today after the calibration

-106.155 : 11073 [L-COMPASS][scale cali(0)] estimation error:[16.8]
-106.155 : 11073 [L-COMPASS][scale cali(0)] succeed! bias:235.1 -137.2 -1277.7 scal:3.210 3.154 3.280|

I received my M2 Pro very early. It'll be interesting to see if others start to see their Mavic 2 requesting a calibration after 30 days.

Many of the M2P's are requesting a calibration after every flight or every few flights. @Robert Granata you are the ironman, can you tell us if you have to calibrate very often?

 

[sar104]

Yes there is something strange going on. Mine is giving the same result today; magnetic interference, move aircraft or calibrate compass. And yet the interference reading in the sensors section of the GO app is way down in the green and the aircraft heading is correct.

33.131 : 1849 [L-NS][AHRS] wmm dec: 8.590796​

33.131 : 1849 [L-NS][AHRS] wmm inc: 62.866161​

34.367 : 1922 [L-COMPASS][mag_cali_pt] date_from_last 30|​

34.522 : 1931 [L-FMU/LED]action changed. compass out of range:need_cali(1)​

It's been thirty days since I first received and calibrated the compass, not because it was asking but because the interference reading was up in the yellow. I'm not sure where the lat/long values came from - they are obviously wrong.

145.088 : 8407 [L-COMPASS][scale cali(0)] fill num:[139]​

145.088 : 8407 [L-COMPASS][scale cali(0)] estimation error:[13.3]​

145.088 : 8407 [L-COMPASS][scale cali(0)] succeed! bias:-602.9 -463.6 97.1 scal:3.374 2.905 3.290|​

145.088 : 8407 [L-COMPASS][save data] app cali all success​

145.088 : 8407 [L-COMPASS][save data] in user index mode​

145.088 : 8407 [L-COMPASS]mag cali pos and time saved success!​

145.088 : 8407 [L-COMPASS][mag_cali_pt]lat:0.626607, lon:-1.855615​

145.088 : 8407 [L-COMPASS][mag_cali_pt]height:2256.3, date:20180829​

 

[Eugenious99]

nothing special here , after interference with metal net or screen or metal reinforcement on concrete driveway it always asking for calibration

 

[BudWalker]

.......I'm not sure where the lat/long values came from - they are obviously wrong.
.......

145.088 : 8407 [L-COMPASS][mag_cali_pt]lat:0.626607, lon:-1.855615​

.....​

They are expressed in radians. Now we know where you do your flying.

 

[Robert Granata]

Many of the M2P's are requesting a calibration after every flight or every few flights. @Robert Granata you are the ironman, can you tell us if you have to calibrate very often?

No, but I calibrated it when I got it and tomorrow will be 1 month, so we'll see what happens.

 

[sar104]

They are expressed in radians. Now we know where you do your flying.

How on earth did I forget that? Thanks.

 

[BudWalker]

How on earth did I forget that? Thanks.

Yeah, I was kinda wondering.

 

[gnirtS]

The manual has changed its compass description from previous models.
It now recommends calibration if its unused for over 30 days, travelled over 31 miles or if prompted by the app.
Maybe its detecting this.

Mine appears to be prompting to recalibrate every time it moves a large distance (1000 miles or so) which the M1 never did.

 

[Former Member]

The manual has changed its compass description from previous models.
It now recommends calibration if its unused for over 30 days, travelled over 31 miles or if prompted by the app.
Maybe its detecting this.

Mine appears to be prompting to recalibrate every time it moves a large distance (1000 miles or so) which the M1 never did.

Glad you took the time to post this. Maybe that is what is happening.

 

[sar104]

The manual has changed its compass description from previous models.
It now recommends calibration if its unused for over 30 days, travelled over 31 miles or if prompted by the app.
Maybe its detecting this.

Mine appears to be prompting to recalibrate every time it moves a large distance (1000 miles or so) which the M1 never did.

It's detecting the 30 days and prompting. It does not appear to be detecting change in location and prompting based on that, or at least mine isn't.

As to why they are doing this, my assumption would be that compass issues (even if user-induced) are responsible for many lost and damaged aircraft, and that they are trying to cover themselves as much as possible in terms of responsibility.

It's also worth considering that in all the hundreds of logs that I've examined both here and on PhantomPilots, I can only remember maybe one or two where a compass calibration problem was even a possible suspect.

 

[kilomikebravo]

SAR...You really and truly know your stuff. The question that comes to my mind is, why now the 30 day detection?

 

[sar104]

SAR...You really and truly know your stuff. The question that comes to my mind is, why now the 30 day detection?

I don't know. I talked to @BudWalker about that earlier this morning and we agreed that there are a couple of possible reasons.

One is the "just to cover themselves" approach that I've mentioned previously, given that maybe the majority of crashes/losses reported, at least on this forum, are due to magnetic interference. Not that calibration fixes that, or even has anything to do with it.

Another is that these magnetometers do appear to drift a bit, as might the magnetic state of the aircraft, and so regular calibration may be a good idea.

We also discussed whether there might have been any hardware changes that make the compass non-linear with respect to field strength, which could be related to prompts to calibrate if the location changes significantly, but that seems like a long shot. That doesn't mean, however, that they might not have add a location check and a requirement to recalibrate, even if there is not a technical reason to do that.

It would be interesting if anyone who has had a calibration prompt that they think is related to location could post the mobile device DAT file for that flight. Since the event log explicitly includes notification of the 30 day check we might expect it to make a note of the location reason if that happens.

 

[kilomikebravo]

It'll be interesting to see if others start to see their Mavic 2 requesting a calibration after 30 days.

Coincidence? Probably not. Had a break in the rain today, fired everything up, and lo and behold,

“Calibrate compass or move.”

I also checked and as an early recipient as well, I was right at the 30-day mark. So, me thinks there’s something to this 30-day clock.

 

[Former Member]

Coincidence? Probably not. Had a break in the rain today, fired everything up, and lo and behold,

“Calibrate compass or move.”

I also checked and as an early recipient as well, I was right at the 30-day mark. So, me thinks there’s something to this 30-day clock.

It makes sense to just move to a different spot

Mavic 2 Compass Calibration requested?

 

[BudWalker]

Coincidence? Probably not. Had a break in the rain today, fired everything up, and lo and behold,

“Calibrate compass or move.”

I also checked and as an early recipient as well, I was right at the 30-day mark. So, me thinks there’s something to this 30-day clock.

If you can provide the tablet .DAT we could take a look and see if it was due to the 30 day clock. Did you calibrate the compass 30+ days ago? Look here to see how to retrieve the tablet .DAT.

 

[BudWalker]

It makes sense to just move to a different spot

Mavic 2 Compass Calibration requested?

Moving to a different spot won't matter if the calibration request is due to it being 30+ days since the previous calibration.

 

[sar104]

Moving to a different spot won't matter if the calibration request is due to it being 30+ days since the previous calibration.

Agreed. In my case there was no interference at all and the compass sensor interference reading was firmly in the green. The log is clear that it is a 30-day check.

 

[kilomikebravo]

If you can provide the tablet .DAT we could take a look and see if it was due to the 30 day clock.

Bud: Up to this point, I've never needed to look at .DAT files so I'm a rookie with DATcon and CSView but first, it now seems like there are two folders on an iOS device, "FlightRecords" and "FlightLogs." I see only the encrypted .TXT files in FlightRecords and only .DAT files in FlightLogs but these data files are very small as in less than 1k.

However, there's a subfolder under FlightRecords that also contains .DAT files and these are probably what you want since they're much larger than the tiny .DAT files mentioned above.

In \..\FlightRecords\MCDatFlightRecords\ I have five .DAT files with yesterday's date. I don't know why the forum software will not accept the .DAT file extension for uploads but I have archived the first two (chronologically) .DAT files in the attached .ZIP.

Yes, the aircraft requested a calibration upon unpacking but has not asked again until yesterday. That initial cal was on 8/28 so having it pop up on 9/29 (I didn't fly on Day 30 due to rain so it was one day beyond 30 when it asked.)

If neither of these provides the info you need, holler and I'll zip up the other three. I did take a look at the exported .CSV for these two attached files and only see "False" in the different compass error fields.

TIA,

KB

 

[BudWalker]

Bud: Up to this point, I've never needed to look at .DAT files so I'm a rookie with DATcon and CSView but first, it now seems like there are two folders on an iOS device, "FlightRecords" and "FlightLogs." I see only the encrypted .TXT files in FlightRecords and only .DAT files in FlightLogs but these data files are very small as in less than 1k.

However, there's a subfolder under FlightRecords that also contains .DAT files and these are probably what you want since they're much larger than the tiny .DAT files mentioned above.

In \..\FlightRecords\MCDatFlightRecords\ I have five .DAT files with yesterday's date. I don't know why the forum software will not accept the .DAT file extension for uploads but I have archived the first two (chronologically) .DAT files in the attached .ZIP.

Yes, the aircraft requested a calibration upon unpacking but has not asked again until yesterday. That initial cal was on 8/28 so having it pop up on 9/29 (I didn't fly on Day 30 due to rain so it was one day beyond 30 when it asked.)

If neither of these provides the info you need, holler and I'll zip up the other three. I did take a look at the exported .CSV for these two attached files and only see "False" in the different compass error fields.

TIA,

KB

Looking at both of the .DATs you submitted it doesn't appear that the calibration request was due to the 30 day clock. I.e., there were no entries in the eventLog stream that looked like
34.367 : 1922 [L-COMPASS][mag_cali_pt] date_from_last 30|

I think what happened was that the M2 was sitting on the concrete driveway close to some rebar which caused the "compass calibration or move" request.That was in FLY086. The magMod value was very low indicating a geomagnetic distortion.

Roughly 60 secs later the M2 was turned off then back on which created a new .DAT file, FLY087. After roughly 200 secs the M2 was picked up to perform the calibration. At, this point the magMod value took on a normal value. That was because the geomagnetic distortion likely extended only 1 foot at most above the surface of the concrete. After the calibration the M2 was placed on the concrete which then caused the magMod to once again be unrealistic, this time very high.


The crucial part of this incident is that the M2 was sitting on the concrete in the geomagnetic distortion prior to launch. Why, then was there not a fly away seen so many other times in this scenario. It's because the M2 was being held out of the distortion for roughly 150 secs during the calibration. It's not because the calibration was performed. If you had just held the M2 above the distortion for 150 secs the effect would have been the same.

After motorStart the M2 sat on the concrete for 9.62 secs. As the M2 ascended it left the distortion and the magMod value returned to normal.


You were lucky. If the M2 had sat longer on the ground the effects of the distortion would have had time to affect the Yaw value, possibly causing erratic flight and/or a fly away.