added magnetometer models

[RomanLut]

This PR adds 3D models of various magnetometers and magnethometer chips to Alignment tool tab.

[Jetrell]

@RomanLut Very nice rendering of the GNSS/mag hardware models.. It looks to work much better.
It even allows XYZ work-arounds, which it did not before.. What I mean is. If you know how, there are a couple of ways to setup the hardware orientation, and still have it work correctly.. I think its great that it can account for this 👍.

But I did notice the alignment tool would occasionally lock-up the configurator after a save/reboot.
This more often occurred when the XYZ model was used, instead of the magnetometer model selections.. Maybe because it was experiencing changes it didn't agree with... But after a power cycle and restart of the configurator. Those less orthodox settings I loaded still worked correctly.
But it maybe a bug, so I thought it worth mentioning. But it may also be related to the electron compiler build. I was using WIndows OS.

[samsab1]

sounds good :-)

[RomanLut]

@Jetrell But I did notice the alignment tool would occasionally lock-up the configurator after a save/reboot.

Do you mean that after "save and reboot" configurator may fall into never endling loading cycle? Unfortunatelly it happens with other tabs with current Master. Alignment tool is not exception here.

[Jetrell]

@RomanLut

I forwarded this build to a friend who was having trouble setting up his quad. Which has both the GNSS/mag module and FC board rotated from the forward default position.
These changes didn't make it any easier for him, even though it provided a magnetometer model of his combine module.
He has no idea how to reference the XYZ axis rendered image.. All he understands is making the image of the GNSS module in the tool, look the same as it does on his hardware. And hoping that will provide him with the settings he needs to make it work.

I'd like to give you some history.
Matek systems were one of the first companies to release combine GNSS/mag modules back in 2018. And they provided the information required for setup. e.g. Orientation Preset CW270FLIP. And the mounting direction reference arrow, or plug facing the opposite direction to it.
But over the last couple of years. There have been a flood of Chinese manufactures releasing these combine modules, that have totally screwed up the whole orientation setup procedure. By not including these default settings.

The procedure was simple. By default, the arrow on the FC and the arrow on the GNSS module should face the direction the aircraft will travel. With the Orientation Preset being the offset of the magnetometer chip, with respect to the arrow they provide, or the plug facing the opposite direction to the arrow (rear of the model)

Can I ask a question respectfully. I'm under the impression that you misunderstand what the Orientation Preset or CLI align_mag setting is for ?

I ask because it seems you have integrated align_mag or orientation preset with the align_mag RPY slider settings.
However it should not be this way.

• The first is a default offset, based on the manufactures board design (mag chip) and mounting recommendation. It can be
  likened to a zero calibration offset.
• While the second is for the user to make their own installation adjustments.

Look at the example below. This is what the tool has always shown when CW270FLIP is loaded as the Orientation Preset.
The orientation preset is crossed out. And the align_mag_RPY sliders are altered to suit.. But it shouldn't be this way. The orientation preset has already taken care of that offset.

Instead, it should look like this with the arrow facing forward, or the plug facing the rear of the aircraft.. The Roll, Pitch and Yaw sliders should be zero after the orientation preset (offset) is applied.

If all three sliders are not at zero, after the orientation preset is loaded. CW90, CW180 and CW270 with flip, will never allow the slider settings to line-up with what the user sees on their hardware, when comparing the degrees offset, to the front of the aircraft as a reference.
align_mag_yaw has already been taken care of by CW270 . And align_mag_pitch was taken care of by the FLIP function.

The image below shows how the tool does not work as the user would expect.
With slider align_mag_yaw = 0. The GNSS/mag arrow should be facing forward, as it would be on their aircraft hardware.

The only one that is not effected is CW0 or CW0FLIP.. Because a module with that preset, will already have a correct reference to the front of the aircraft.

[RomanLut]

---

@Jetrell All he understands is making the image of the GNSS module in the tool, look the same as it does on his hardware. And hoping that will provide him with the settings he needs to make it work.

Yes he has to do this way. This is the whole idea of the tool. If he can provide photos of his setup, we may try to understang what is wrong.

---

The are two methods to provide orientation of the mag in inav:
a) with align_mag preset
b) with align_max_pitch, align_max_roll, align_max_yaw angles.

If all (b) are zero, then align_mag is used. if any of (b) is non-zero, then (b) is used and align_mag is ignored.

(a) defines rotation relative to FC. (b) defines absolute rotation.

This is how settings work in inav. This is not going to be changed.

---

On this picture tool shows that align_mag is ignored - dropdown is crosed out. Orientation is set with (b) - defined with sliders.
Dropdown can have any value here, it is just a current value of align_mag setting from cli which is ignored.

---

On this picture, tool shows that align_mag=cw270flip is used (dropdown is NOT crossed out).

At the same time, sliders show values for (b) which would provide the same orientation as align_mag=cw270flip with current FC rotation. If user desides to define orientation using (b), he can touch sliders. Dropdown will became crossed out. This very usefull if mag is mounted slightly off. User can rotate it a little bit at any time, tool will switch to (b).

---

The image below shows how the tool does not work as the user would expect.
With slider align_mag_yaw = 0. The GNSS/mag arrow should be facing forward, as it would be on their aircraft hardware.

Arrow is not facing forward because align_mag is ignored.

I thought that crossed out dropdown would be enough to understand that preset is not used. If it is not the case - I can add explicit bold text somewhere.

---

If you want to define orientation with (a) - select preset in dropdown, do not touch sliders, save and reboot.
If you want to define orientation with (b) - setup orientation using sliders. Dropdown is crossed out once sliders are toched.

---

You might wonder what will happend if user set sliders to zero and save setting - will it revert to cw270flip preset and screw up orientation? No, tool will save align_max_pitch=0, align_max_roll=0, align_mag_yaw=1. Orientatino will still be defined by angles consistently.

[Jetrell]

The are two methods to provide orientation of the mag in inav:
a) with align_mag preset
b) with align_max_pitch, align_max_roll, align_max_yaw angles.

I agree. And I understand your view point on this matter.
However.. I'm speaking from the perspective of others who will not understand. Those who have never used a magnetometer before.. We often see this on Discord.

(a) defines rotation relative to FC. (b) defines absolute rotation.

I believe the above explanation is somewhat confusing for new users.
In reality, both (a) and (b) are relative to the flight controller.
While (a) is also relative to the magnetometer chips mounting position on the PCB, with respect to a given reference point on the module. (the arrow or plug). Comparing to the front 0° of the aircraft.

I rang the guy I was speaking to about this. He doesn't have a Github account.
He gave me this example. Which did not make sense to him.

You can see that the front of the aircraft is the reference point. (green arrow and 0°.)
The arrow on the module is facing the rear as he requires.. Which is 180° from the front of the aircraft.
But the YAW slider has to be set to 90° instead of 180°, based on the modules rendered arrow facing rearward.. And to provide the correct settings.
Which all reference back to the Orientation Preset. Because its CW270° minus CW180° = CW90°.

I know this is simple if you understand, as you and I do.
But it makes the tool so confusing for new users from a visual perspective, especially at the times the tool is required the most for more diverse hardware orientations.

This is why I said the mag sliders should be set to zero when an orientation preset is defined and loaded, providing an active offset, referenced from the front of the aircraft.

[RomanLut]

If FC is mounted with arrow pointing forward:

If matek-m10q is mounted with arrow facing forward, then cw270flip orientation preset should be used.
The same orientation can be set using align_mag_roll =0, align_mag_pitch =180, align_mag_yaw=270, as shown by the tool:

If matek-m10q is mounted with arrow facing backward, then cw90flip orientation preset should be used.
The same orientation can be set using align_mag_roll =0, align_mag_pitch =180, align_mag_yaw=90, as shown by the tool:

Sliders do not show rotation which is "additionally applied on top of orientation preset". They show values of align_mag_xxx settings which corresponds to orientation of magnethometer as seen on screen.

Maybe this helps to understand: yaw slider has to be set to 90 because CW90flip means "rotation by 90 on yaw".

[rts18]

@RomanLut I've been watching this pull request along with #2025.
I was one of the people who pointed out the alignment tool had issues.

I would like to send a thank you for attempting to make this tool work correctly.
However, with all due respect. After reading the comments in #2025 and here. And being a long term iNAV user. I have to agree with @Jetrell .
I too understand your perspective as he does. But I don't see how that will make this tool useful for new users.

If matek-m10q is mounted with arrow facing backward, then cw90flip orientation preset should be used.

We may understand. But I guarantee you that know one else will. If I had to go to that much trouble to decipher what the tool is telling me, I wouldn't bother with it. As I don't. But this isn't about me.
Newbies need someone with your skills to make this tool easy to interpret. Most do not have the ability to work out the spatial axis offset.
They require a tool which allows them to replicate their models hardware layout by moving the sliders to a setting that makes sense. And provides the correct compass and flight controller settings, based on what the GUI is showing them.

[RomanLut]

User in NOT required to undestand how aling* settings work. User is NOT required to know preset name for default mounting of mag. User is NOT requred to know order or rotations and directions of euler angles. In extreme case, I could hide all text, numbers and dropdown, and display just three nameless sliders. All user is required to do - is to rotate mag model to match physical orientaton and click "save and reboot". Values of align_mag_* settings make no sense for the user? It is expected. In order for settings to make sense, user has to understand everything mentioned above.

[mmosca]

I was just looking at the alignment tool in 7.x and noticed that the default model arrow does not match the gps axix, which certainly does not help. Having these many models with correct alignment would be a great step forward, but I worry a bit about keeping it up to date in the long term.

[RomanLut]

@mmosca Y Axis shoud not point to the north, it should point in direction of magnethometer Y axis with current mounitng. In the long term, aligment is possible by alligning magnetomer chip orientation or magnetometer axes.
#2061

[Choolet]

I was going to report a problem with the compass side of the alignment tool in iNav 7.1. But its already being worked on. Thanks @RomanLut
I read through the comments. At the risk of sounding like another broken record, it leaves me wondering why you have not zeroed the magnetometer sliders once the orientation preset is applied.
This is what orientation preset is for. To reset them to the default position of CW0FLIP with the plug facing the opposite direction the model travels.
That's why its called a PRESET. It effectively clears the offset made by manufactures that use any other magnetometer chip position other than CW0FLIP.
And why does it have align_mag_pitch = 180 once the preset is applied. That's what flip does.

Sliders do not show rotation which is "additionally applied on top of orientation preset".

That is not correct. Additional align mag slider setting should be zero after the orientation preset 90, 180 or 270 are applied.
Another way of explaining this. Is if every magnetometer chip was mounted equivalent to CW0FLIP with its plug facing the rear as default, there would be no need of an orientation preset.

As for the XY coordinates. I fail to see the benefit of them being added to a module or to the tool. How are people to know if its X positive / negative, or Y positive / negative.

[RomanLut]

@Choolet >why you have not zeroed the magnetometer sliders once the orientation preset is applied

The are two methods to provide orientation of the mag in inav:
a) with align_mag preset
OR
b) with align_mag_pitch, align_mag_roll, align_mag_yaw angles.

If all (b) are zero, then align_mag is used. if any of (b) is non-zero, then (b) is used and align_mag is ignored.

This is stated in the title "2. Select a preset (align_mag) or create a custom configuration using the sliders
(align_mag_roll, align_mag_pitch, align_mag_yaw)"

So yes, in order for preset(align_mag) to take effect, align_mag_pitch, align_mag_roll, align_mag_yaw settings should be zeroed.

At the same time, it is usefull to show align_mag_pitch, align_mag_roll, align_mag_yaw settings which would provide the same orientation as selected preset with current FC rotation. If user desides to define orientation using (b), he can touch sliders.

For example: I am experienced user, I know that my FC is not rotated, my matek magnehometer is mounted with arrow forward, so I select CW270FLIP. Then I notice that mag is actually slightly off by 20 degrees, so I use sliders to adjust orientation slightly. Tool switches from preset to align_mag_* settings once sliders are touched.

In this PP, I hope I made enough changes in interface to indicate which settings take effect.

If preset takes effect, then align_mag_pitch, align_mag_roll, align_mag_yaw are striked out, and CLI string below shows "align_mag_pitch=0, align_mag_roll=0, align_mag_yaw=0" for thouse who understand how settings work:

If align_mag_pitch, align_mag_roll, align_mag_yaw takes effect, then preset dropdown is striked out:

Update: Also a label has been added which rexplicitly tell: "Magnetometer orientation is defined by: PRESET" or "Magnetometer orientation is defined by: ANGLES".

As for the XY coordinates. I fail to see the benefit of them being added to a module or to the tool. How are people to know if its X positive / negative, or Y positive / negative.

Actually, XYZ model is the most important model in the tool. This single model is enough to setup any magnetometer.

inav does not care about magnetometer or magnetometer chip manufacturer, arrow marking, default preset etc. All inav needs to know is how magnethometer local frame axes are oriented on the physical setup.

If magnethometer has XY axis markings, settings can be easily found in tool by aligning XYZ model to physical setup. So in future, if some magnetometer 3D model is not included in tool, setings could be found by alligning axes.

[Choolet]

The are two methods to provide orientation of the mag in inav:
a) with align_mag preset
OR
b) with align_mag_pitch, align_mag_roll, align_mag_yaw angles.

I don't agree with that assessment.
Both (a)align_mag and (b)align_mag_roll, align_mag_pitch, align_mag_yaw may work on a similar premise.
But they are purposely different settings.

(a) Should be applied first. align_mag is purposely meant to account for the manufactures mounting orientation of the magnetometer chip. With reference to a known marker on the module. Either the plug or arrow.
You practically have it working. But it baffles me why you can't seem to grasp the individual purpose of these two settings.

When any one of the orientation presets are applied. All the (b) sliders should be reset to ZERO. So they can be used to reference mounting position changes by the user.

If all (b) are zero, then align_mag is used. if any of (b) is non-zero, then (b) is used and align_mag is ignored.

(a) stands by itself. It should not be rendered void when the sliders are adjusted.
It can be likened to resetting the Trip meter in your automobile. It provides a new distance reference point. But it doesn't alter the Odometer distance, its the base reference.

[RomanLut]

align_mag_roll, align_mag_pitch, align_mag_yaw are not applied on top of preset (allign mag) in inav.
There are no settings in inav to apply additional rotation on top of preset.

If CW270FLIP preset takes effect, then settings are:
align_mag=CW270FLIP align_mag_rol=0 align_mag_pitch=0 align_mag_yaw=0

If magnethometer is off by 1 degree, then settings should be:
align_mag=[any value because it is ignored] align_mag_rol=0 align_mag_pitch=1800 align_mag_yaw=2710

While CW270FLIP preset takes effect, I prefer to set sliders in positions 0 180 270 and strke out "align_mag_xxx" title to indicate that sliders values do not have effect. This way, when user touches sliders, magnetometer keeps orientation and does not flip to 0,0,0 (CW0).

[DzikuVx]

@RomanLut shall we merge? It looks awesome enought :)

[RomanLut]

@RomanLut shall we merge? It looks awesome enought :)

PR is ready for merging.