Compass documentation

[Mark Jacobsen]

Hi everyone,

I’ve been flying ArduPlane for a year and a half and have just recently started learning ArduCopter. I know the ecosystem pretty well, but one area I’ve always been baffled by is the compass. Judging by the many threads I’ve perused on the forums, this remains a source of confusion for many users. I believe the documentation could be much improved here, and have taken it upon myself to try making some improvements. But in order to do that, I need help with a number of questions that remain. The list is long, but these are questions that arise because the documentation is confusing and does not provide clear answers on these points. I can start inspecting code if necessary, but thought it would be much easier to start by asking here. I would be grateful for any help with these, and will hopefully be able to pay back the favor by integrating all this knowledge into a more user-friendly compass documentation page. Thank you in advance! Mark

1. Documentation uses the words “compass” and “magnetometer” interchangeably. Should this be standardized or not?

2. For APM2.5 (internal compass), APM2.6 (external compass), and Px4 (both compasses), which compass is assigned to which compass number?

3. Documentation states that Px4 auto-detects an external compass and disables the internal compass. This whole idea of automatic enabling/disabling compasses seems inconsistent with the fine degree of control over compass usage through parameters. Is this enabling/disabling done at the hardware level, or does the Px4 actually change parameters? Which parameters? is it the COMPASS_USE parameters? Can the user override this automatic configuration? Would those overrides persist, or does the Px4 change the configuration upon every boot?

4. The basic compass and advanced compass pages are inconsistent about compass orientation. The basic compass page says that the instructions for specifying a manual compass orientation ONLY apply to APM2, not Px4. I think this is a mistake, as the advanced page specifies that the manual orientation applies to either APM2 or Px4. Is that true?

5. What is the purpose of the COMPASS_EXTERN parameters? Is this strictly so the flight controller knows whether to use the AHRS orientation settings or the compass orientation settings?

6. For Px4, does the Mission Planner compass calibration routine calibrate both the internal and the external compass? I assume this is the case, since it shows two sets of offsets during calibration, but how does this square with the documentation stating that the internal compass is automatically disabled? Does Mission Planner always calibrate both compasses, or does it choose compasses based on the COMPASS_USE parameters?

7. What threshold numbers for compass offsets are considered excessive, cause “Compass Health” messages, etc.? I’ve seen a couple different numbers in the forums, none of which cite an authoritative source. Do the units correspond to anything meaningful?

8. How does ArduPilot fuse data from both GPS heading and multiple compasses? Does the compass always override GPS entirely if one of the COMPASS_USE parameters is enabled? Does it only use data from COMPASS_PRIMARY? Does data from a secondary compass get used somehow, and if so, how does this square with the documentation saying that the internal compass is automatically disabled? How does this fit with EKF vs non-EKF?

9. What checks lie behind the “compass health” message?

10. What checks lie behind the “compass variance” message? What does that even mean, if the internal compass is automatically disabled and there is only one compass providing data?

11. What happens if COMPASS_PRIMARY is set to a specific compass, but that compass’s COMPASS_USE is disabled?

12. The EKF documentation states that EKF can estimate magnetometer offsets, and there is a EKF_MAG_CAL parameter that supposedly enables this. How does this fit with COMPASS_LEARN? Is COMPASS_LEARN only for non-EKF, or do they somehow work together?

13. Given that the compass calibration routine seems to calibrate both compasses, is the same true for COMPASS_LEARN and EKF learning? Or does the parameter configuration determine which compasses learn?

[Randy Mackay]

Mark,

Some answers in line.

-Randy

-----Original Message-----
From: drones-...@googlegroups.com [mailto:drones-...@googlegroups.com] On Behalf Of Mark Jacobsen
Sent: 29-Sep-15 10:47 PM
To: drones-...@googlegroups.com
Subject: [drones-discuss] Compass documentation

Hi everyone,

I’ve been flying ArduPlane for a year and a half and have just recently started learning ArduCopter. I know the ecosystem pretty well, but one area I’ve always been baffled by is the compass. Judging by the many threads I’ve perused on the forums, this remains a source of confusion for many users. I believe the documentation could be much improved here, and have taken it upon myself to try making some improvements. But in order to do that, I need help with a number of questions that remain. The list is long, but these are questions that arise because the documentation is confusing and does not provide clear answers on these points. I can start inspecting code if necessary, but thought it would be much easier to start by asking here. I would be grateful for any help with these, and will hopefully be able to pay back the favor by integrating all this knowledge into a more user-friendly compass documentation page. Thank you in advance! Mark

1. Documentation uses the words “compass” and “magnetometer” interchangeably. Should this be standardized or not?

>>> Yes. I'd say "compass" is more easily understood by most people so we should use that.

2. For APM2.5 (internal compass), APM2.6 (external compass), and Px4 (both compasses), which compass is assigned to which compass number?

>>> on APM2.x boards there's only one compass so whether it's internal or external it's always compass #0.
>>> For Pixhawk, if two compasses are attached then external is #0, internal is #1. If no external compass is attached then the single internal compass appears as #0.

3. Documentation states that Px4 auto-detects an external compass and disables the internal compass. This whole idea of automatic enabling/disabling compasses seems inconsistent with the fine degree of control over compass usage through parameters. Is this enabling/disabling done at the hardware level, or does the Px4 actually change parameters? Which parameters? is it the COMPASS_USE parameters? Can the user override this automatic configuration? Would those overrides persist, or does the Px4 change the configuration upon every boot?

>>> Ardupilot only ever uses one compass and it makes that decision soon after start-up. It uses the COMPASS_PRIMARY compass unless it's unhealthy in which case it uses the first healthy compass. In general this means it will use the external compass if it's attached. Which compass is specified by either directly setting COMPASS_PRIMARY or by setting all other COMPASS_USEX values to "0". The subtle difference between these two methods is that setting COMPASS_USEX to "0" will mean that compass's orientation is not checked during the pre-arm.

4. The basic compass and advanced compass pages are inconsistent about compass orientation. The basic compass page says that the instructions for specifying a manual compass orientation ONLY apply to APM2, not Px4. I think this is a mistake, as the advanced page specifies that the manual orientation applies to either APM2 or Px4. Is that true?

>>> Yes, you're right in the case of the Pixhawk's external compass. The COMPASS_ORIENT parameter can be used on a pixhawk to specify that the external compass is not in its default orientation (default is pointing forward). For the Pixhawk's internal compass the orientation cannot be changed.

5. What is the purpose of the COMPASS_EXTERN parameters? Is this strictly so the flight controller knows whether to use the AHRS orientation settings or the compass orientation settings?

>>> Yes. For the APM2.x boards this parameter was necessary because we couldn't automatically detect if the compass was internal or external. It's probably also required for some of the Linux boards where we're not sure how the compass is connected.

6. For Px4, does the Mission Planner compass calibration routine calibrate both the internal and the external compass? I assume this is the case, since it shows two sets of offsets during calibration, but how does this square with the documentation stating that the internal compass is automatically disabled? Does Mission Planner always calibrate both compasses, or does it choose compasses based on the COMPASS_USE parameters?

>>> Yes. It's slightly incorrect to say that the internal compass is "disabled" because it still works, it reports values that can be seen in the ground station, etc. It's more accurate to say that it's "not used".

7. What threshold numbers for compass offsets are considered excessive, cause “Compass Health” messages, etc.? I’ve seen a couple different numbers in the forums, none of which cite an authoritative source. Do the units correspond to anything meaningful?

>>> Yes, the compass values are in milli gauss. https://en.wikipedia.org/wiki/Gauss_%28unit%29
>>> offsets cannot be higher than 600 milligauss or the pre-arm will complain, "PreArm: Compass offsets too high"
>>> magnetic field cannot be > 875 nor <185 or the the copter pre-arm will complain, "PreArm: Check mag field"

8. How does ArduPilot fuse data from both GPS heading and multiple compasses? Does the compass always override GPS entirely if one of the COMPASS_USE parameters is enabled? Does it only use data from COMPASS_PRIMARY? Does data from a secondary compass get used somehow, and if so, how does this square with the documentation saying that the internal compass is automatically disabled? How does this fit with EKF vs non-EKF?

>>> only the compass values from the primary compass are used. If the compass fails for some reason, ardupilot does not fall back to the next healthy compass. This is done on purpose because we've seen a few situations where the internal compass values are not good.

9. What checks lie behind the “compass health” message?

>>> the primary compass must have returned a sensor value within the last 0.5 seconds.

10. What checks lie behind the “compass variance” message? What does that even mean, if the internal compass is automatically disabled and there is only one compass providing data?

>>> this is a number from the EKF which is always between 0 ~ 1.0. 0 = compass data is consistent with the accelerometer and gyro data. 1 = it's inconsistent and is being completely ignored.

11. What happens if COMPASS_PRIMARY is set to a specific compass, but that compass’s COMPASS_USE is disabled?

>>> it will use the first compass that is healthy and has it's COMPASS_USEX set to "1". If no compasses can be found that satisfy that then no compass is used.

12. The EKF documentation states that EKF can estimate magnetometer offsets, and there is a EKF_MAG_CAL parameter that supposedly enables this. How does this fit with COMPASS_LEARN? Is COMPASS_LEARN only for non-EKF, or do they somehow work together?

>>> this is a question for Tridge and Paul Riseborough. I suspect that the two parameters should be combined (maybe EKF_MAG_CAL should probably be removed).

13. Given that the compass calibration routine seems to calibrate both compasses, is the same true for COMPASS_LEARN and EKF learning? Or does the parameter configuration determine which compasses learn?

>>> I can't speak for the DCM's compass-learn (which is never used in Copter) but for the EKF, it only learns the offsets for the primary compass.


--
You received this message because you are subscribed to the Google Groups "drones-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Mark Jacobsen]

Fantastic Randy, thank you! I will get to work on some documentation suggestions, and will let you know if I have further questions. Thanks for the quick turnaround.

Mark

[Mark Jacobsen]

Randy (or others),

I have a few follow-up questions. 

1. In answer #3 you write, "[ArduPilot] uses the COMPASS_PRIMARY compass unless it's unhealthy in which case it uses the first healthy compass." But in answer #8 you write, "If the compass fails for some reason, ardupilot does not fall back to the next healthy compass.  This is done on purpose because we've seen a few situations where the internal compass values are not good." Those two statements seem contradictory. Am I missing something? And if ArduPilot does not fall back to the next healthy compass, then is there any reason to have multiple compasses on board?

2. Does the statement about internal compass values not being good imply that we should advise users to never use the internal compass?

3. This might be a question for Paul, but I'm wondering if I could get a layman's explanation of how compass data can be inconsistent with accelerometers and gyros. I see how you could use/integrate data from those sensors to estimate a heading relative to a previous heading, but wouldn't that previous heading only be able to come from the compass? I can see how you could get inconsistency if the compass and gyros seem to be turning at different rates. But I can't see how "compass variance" would appear when a plane is powered up sitting on the ground, which many users have reported trouble with recently.

4. If the compass fails, does ArduPilot fall back to a GPS heading? Is this different with plane/copter?

Thanks again!

On Tuesday, September 29, 2015 at 7:48:38 PM UTC-7, Randy Mackay wrote:

[Hamish Willee]

Hi Mark

Sounds excellent. I have tried to improve the inconsistency in naming today - replacing "magnetometer" with "compass" or "compass (magnetometer)" where appropriate. There are a few places that I didn't change it - for example when referring to a product name and in the generated-from-code parameters (e.g. Rover Parameters).

You're right that there are lots of inconsistencies - for example in arming plane we see "Compass checks. All compasses must be configured and calibrated, and need to be consistent with each other (if you have more than one compass)" which does not seem to agree with Randy's point.

You obviously have a good understanding of the problems and good ideas around the solutions. Please let me know if I can be of assistance when you come to update.

Regards

Hamish

[Randy Mackay]

Mark,

 

    Thanks for trying to clean up the docs, I’m sure it’ll be helpful for many.

 

   Some more answers below..

 

-Randy

 

From: drones-...@googlegroups.com [mailto:drones-...@googlegroups.com] On Behalf Of Hamish Willee
Sent: 1-Oct-15 9:47 AM
To: drones-discuss
Subject: Re: [drones-discuss] Compass documentation

 

Hi Mark

 

Sounds excellent. I have tried to improve the inconsistency in naming today - replacing "magnetometer" with "compass" or "compass (magnetometer)" where appropriate. There are a few places that I didn't change it - for example when referring to a product name and in the generated-from-code parameters (e.g. Rover Parameters).

 

You're right that there are lots of inconsistencies - for example in arming plane we see "Compass checks. All compasses must be configured and calibrated, and need to be consistent with each other (if you have more than one compass)" which does not seem to agree with Randy's point.

 

You obviously have a good understanding of the problems and good ideas around the solutions. Please let me know if I can be of assistance when you come to update.

Regards

Hamish

On Wednesday, 30 September 2015 23:30:21 UTC+10, Mark Jacobsen wrote:

Randy (or others),

 

I have a few follow-up questions. 

 

1.      In answer #3 you write, "[ArduPilot] uses the COMPASS_PRIMARY compass unless it's unhealthy in which case it uses the first healthy compass." But in answer #8 you write, "If the compass fails for some reason, ardupilot does not fall back to the next healthy compass.  This is done on purpose because we've seen a few situations where the internal compass values are not good." Those two statements seem contradictory. Am I missing something? And if ArduPilot does not fall back to the next healthy compass, then is there any reason to have multiple compasses on board?

 

>>> the internal compass cannot be removed and the external compass tends to have less interference from the motors and power wires.  Having multiple compasses allows us to catch incorrect orientations by checking that they’re consistent.

 

2. Does the statement about internal compass values not being good imply that we should advise users to never use the internal compass?

>>> it depends upon the vehicle.  For most vehicles an external compass does better but for a small percentage an internal compass is enough.  It would be strange to fit an external compass but then use the internal compass.

 

2.      This might be a question for Paul, but I'm wondering if I could get a layman's explanation of how compass data can be inconsistent with accelerometers and gyros. I see how you could use/integrate data from those sensors to estimate a heading relative to a previous heading, but wouldn't that previous heading only be able to come from the compass? I can see how you could get inconsistency if the compass and gyros seem to be turning at different rates. But I can't see how "compass variance" would appear when a plane is powered up sitting on the ground, which many users have reported trouble with recently.

>>> it is more of a question for Paul but you can imagine that if you integrated the gyros’ reported rotation rate but the compass heading wasn’t changing then the EKF would know that either the gyros or the compass was bad and a variance would appear.

>>> I don’t know about Plane but in copter we’ve seen cases where there is a spike in the reported compass values for the first 20 seconds.  Once that’s settled down the compass values return to more sensible values but the change has caused the EKF to report a compass variance.  It’s impossible to say if this is what you’re seeing but if you have dataflash logs, especially with the LOG_BITMASK set to logging-while-disarmed we can have a peek.

 

3.      If the compass fails, does ArduPilot fall back to a GPS heading? Is this different with plane/copter?

>>> Paul will know best but for copter it can’t simply fall back to the GPS heading because the GPS reports the direction of movement of the vehicle, not it’s heading.  I think the EKF can figure out the heading using a combination of the GPS’s implied accelerations and the IMU’s accelerations.  If it’s heading estimate is correct those accelerations will be in the same direction.

[Mark Jacobsen]

Hi everyone,

In addition to working on some more updated documentation, I've taken a stab at modifying the compass configuration page in Mission Planner to better reflect the current capabilities of ArduPilot, based on Randy's answers. My sense is that the capabilities have grown through many iterations since the compass setup page was designed, which explains some of the confusion I and other users have had.

I have laid out the interface, but haven't wired up any code yet. I thought I'd run it by you all before I spend the time trying. I think the attached picture should speak for itself. Clicking any of the three "quick setup" buttons will automatically populate the parameters in an appropriate way. Note that the calibration buttons are still there, just further down out of view.

I'd welcome any feedback on this proposed design.

Mark

[iskess]

Can someone answer the question about Plane? If the external compass fails, will it fall back to the internal compass? Or is it the same as Copter without redundancy?

Thanks for helping to clarify this confusing subject.

[Mark Jacobsen]

Hi everyone!

I just wanted to let you know that I have been working all week on a new compass setup page for Mission Planner, and just submitted a pull request:

https://github.com/diydrones/MissionPlanner/pull/1021

I believe the new page will GREATLY clarify compass setup. I'm happy to chat about the work, but encourage anyone who is interested to check the new page out. A screenshot is attached.

I'm also continuing work on documentation, but will wait to push anything until I hear back on the pull request... as the documentation will need to reflect what's in the setup page. Thanks again to Randy for answering my questions, which were helpful in putting this new page together.

Mark

 

[Randy Mackay]

Mark,

 

     Looks pretty good.  Definitely nice to have more controls and information on the setup screen.

 

     A couple of suggestions:

·         I think it’s save to remove the “Enable compasses” checkbox at least for Copter (we can ask Tridge about Plane).  I was thinking of completely removing the MAG_ENABLE parameter from copter for the next release because flying completely without a compass (at least for copter) is not a supported configuration anyway.

·         There are COMPASS_ORIENT2 and COMPASS_ORIENT3 parameters so there could be orientation drops downs for compasses 2 and 3 as well.

·         I wonder if we should try and put the compass-mot calibration on that page too.

 

-Randy

 

From: drones-...@googlegroups.com [mailto:drones-...@googlegroups.com] On Behalf Of Mark Jacobsen
Sent: 4-Oct-15 1:51 PM
To: drones-discuss
Subject: Re: [drones-discuss] Compass documentation

 

--

[Mark Jacobsen]

Thanks Randy. 

- Sounds god. I'll ask Tridge about MAG_ENABLE and remove it for Copter

- The combo boxes for orient/orient2/orient3 appear automatically if "external compass" is checked, and disappear if it is not checked

- I didn't do anything with compssmot other than show the values for each compass, mainly because I've never used it before. But I'll try to get smart on this and add it.

[Hamish Willee]

Hi Mark, Randy (et al)

This allows more options but I think it might be more scary to people who have set up everything as per the instructions. 

If possible (and I have no idea if it is) one thing you might consider is having two screens:

1. Basic configuration - just has live calibration button and a link to a)the website example, b)advanced configuration page. Ideally this would get the autopilot from previous config (if we can) and make standard "assumptions about everything else".

2. Advanced Compass Config which is your new screen + compassmot.

A few questions:

1. By the time we get to this screen, don't we already know our autopilot board? If so, can we auto-select this for our settings rather than offering it as an option? Just an idea.

2. Is "Onboard mag calibration" CompassMot? Either way, isn't this only for internal compass? Is there any condition of the other settings where this (or live calibration section) could/should be hidden?

3. Randy refers to Compass #0 and #1 but I'm only seeing Compass 1 to 3 here. Is this something internal/different? I think perhaps that you'll "apply" this number based on how people select compass 1-3 and whether they are marked as internal or external.

5. What is "Automatically learn offsets"?

In terms of (existing) docs on this topic we have:

common-Advanced Compass Setup

Compass Calibration (not common)

common-3DR UBlox GPS + Compass Module 

If possible I'd like to combine the first two docs - possibly starting with a setup much like the "more simple" version. Either way, if the simple version is useful, it should be a common doc. Would be nice to link to the calibration docs rather than have duplication in the Ublox doc.

Cheers

H

--
You received this message because you are subscribed to a topic in the Google Groups "drones-discuss" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/drones-discuss/t5ggAuweWVM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to drones-discus...@googlegroups.com.

[Mark Jacobsen]

Hamish,

Thanks for the feedback! It's much appreciated.

I'm sensitive to the concern about putting too much information on page, but those three quick-setup buttons plus a calibration take care of everything. Also, the page is adaptive and hides everything that isn't relevant to a given configuration. I guess my personal bias is that displaying more information--in a logical, appropriate way--creates greater clarity than tucking information away. The rampant confusion surrounding the compass at least partly stems from the lack of information displayed on the current page. Until a week ago, I wasn't even sure which devices had internal compasses, which compass number was internal or external, which compasses were calibrated during a live calibration, what got changed by modifying the orientation combo box, etc. The new screen lays it out in a glance. I guess I'm open to having basic/advanced pages if there is consensus around that idea, but if we go that route, the basic page would still need an overhaul. 

As for your specific questions...

1. I think we want a button press requirement before configuring parameters, so user-modified parameters don't get overwritten automatically. Once we initiate the configuration process, automatic detection might work in some cases, but could be problematic in others. The one case that jumps to mind is APM2.5, which can have either an internal or external compass depending on whether the user has cut the trace for the internal compass per the documentation. I welcome inputs from others on this.

2. I don't know the answer to this and will defer to Randy, if he's reading. I've never used compassmot so am not smart on it.

3. The compasses are referred to as compasses #1, 2, and 3 almost everywhere. All the compass2 parameters have a "2" in the name, and the compass3 parameters have a "3." However, because C++ arrays are indexed from 0, developers often speak of compasses 0-2. Unfortunately that nomenclature made it into one of the parameters, which contributes to the confusion... COMPASS_PRIMARY is indexed from 0-2, and the documentation says "0 = compass1, 1 = compass2." The only way to standardize this would be to change that parameter, which would break things. But I would encourage all developers to adopt the nomenclature 1 through 3 to avoid confusion.

4. 'Automatically learn offsets' is the COMPASS_LEARN parameter, which automatically learns and updates parameters when airborne. Tridge has advised that people use this when flying ArduPlane. I'm not sure the recommendation for ArduCopter.

From my end, I'm waiting from feedback from Michael Oborne before I go any further.

Mark

[Hamish Willee]

Hi Mark

Sounds reasonable - I don't know enough about what can be "guessed" by the system to come up with a better approach. In any case, as long as we hide everything that isn't relevant based on what they select up top or what can be inferred elsewhere then this is as "scary" as things can be .... not that scary :-)

A couple more thoughts ...

1. If we can infer that we're on pixhawk we can hide those APM options and we should be able to detect the internal and external compasses on page load - so if we detect just the internal compass the other two options can be removed. For copter/plane you know that you can't be running on APM from a particular firmware version. You probably already planned that :-)

2. ">>> For Pixhawk, if two compasses are attached then external is #0, internal is #1.  If no external compass is attached then the single internal compass appears as #0. " You can add smarts to envorce that if and internal and external compass exist that the internal is #2 (assuming 1,2,3 naming convention)

Lets see what Randy says re my previous "2". 

Cheers

H

[Randy Mackay]

 

     With people mostly using external compasses at least on mid to large copters, I think compassmot is much less used than it used to be.  Personally I never use it anymore.

 

     I think Tridge and I are probably against changing the indexing to start from “1”.  I think we try to keep the documentation, parameters and coding consistent and because C++ does it’s array indexing from “0” I think it would add confusion to start number from “1”.  Better to instead handle it in the GCS with labels.

 

     Copter-3.3’s EKF actually learns compass offsets but it always does it and there’s no way to turn it off.  We should rationalise that COMPASS_LEARN parameter so that it turns on/off learning for both DCM and EKF.

Hi everyone!

 

I just wanted to let you know that I have been working all week on a new compass setup page for Mission Planner, and just submitted a pull request:

https://github.com/diydrones/MissionPlanner/pull/1021

 

I believe the new page will GREATLY clarify compass setup. I'm happy to chat about the work, but encourage anyone who is interested to check the new page out. A screenshot is attached.

 

I'm also continuing work on documentation, but will wait to push anything until I hear back on the pull request... as the documentation will need to reflect what's in the setup page. Thanks again to Randy for answering my questions, which were helpful in putting this new page together.

 

Mark

 

 

 

 

 

--

You received this message because you are subscribed to the Google Groups "drones-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "drones-discuss" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/drones-discuss/t5ggAuweWVM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Hamish Willee]

Thanks Randy

- So we don't need the learning offsets - or at least once you've "rationalised" the code. Is there an issue # for that?

Any thoughts on this query:

- 2. Is "Onboard mag calibration" CompassMot? Either way, isn't this only for internal compass? Is there any condition of the other settings where this (or live calibration section) could/should be hidden?

Cheers

H

[Mark Jacobsen]

Randy,

But my point is that most of the parameters -do- start indexing from 1. The first compass parameters are all COMPASS_USE, COMPASS_ORIENT, COMPASS_EXTERN, COMPASS_OFS, COMPASS_MOT, etc. The second compass parameters are all COMPASS2_USE, COMPASS2_ORIENT, COMPASS2_EXTERN, etc. And the third compass parameters are all COMPASS3_USE, COMPASS3_ORIENT, COMPASS3_EXTERN, etc.

I believe the only place in the compass parameters where they are indexed from 0 is the value of the COMPASS_PRIMARY parameter.

Mark

To unsubscribe from this group and stop receiving emails from it, send an email to drones-discuss+unsubscribe@googlegroups.com.