Add extra magnetometer to improve the single-GNSS variant

[realtimepeople]

Meanwhile we have quite a lot of experience with Larus installations in many gliders.

It becomes more and more clear that mounting the sensor close to or inside of the instrument panel
is favorable because we only need short wires and tubes making the installation simple and fast.

Unfortunately, this place is a bad one for the magnetometer.
In many cases there are power lines with varying currents close to the induction sensor.
This does drastically impact the performance of the magnetic compass.

Example:
A wire carrying 1A at a distance of 10cm to the induction sensor creates a distortion
of the earth induction corresponding to 3 degrees error in magnetic heading.
So the feed lines of canopy flashers or of the radio greatly downgrades the precision of the AHRS.

We shall add an extra hardware component:
An extra magnetometer that can be mounted in a place with no power-lines in the vicinity.
This sensor shall be connected to the Larus system using the CAN bus.

We do already have the algorithms to calibrate the orientation and offset/gain of such an extra sensor
automatically. So the implementation of the extra algorithms shall be quite simple.

[Simsys]

We have encountered some problems with the Larus Essential versions. Unfortunately, the cause of the problems is not understood.

The problems no longer occur when playing back the recorded data on the PC. The device malfunctioned in the aircraft. However, the recorded data was usable.

Furthermore, the conclusion presented above is contradicted by the fact that the device ran for many hours without any problems in the same installation location under the same conditions.

These points clearly contradict the assertion that the installation situation is the cause of the problems.

We should not open a new work order to solve an unknown problem. It could be that the remote sensor does not bring any improvement.

[Simsys]

From an economic perspective, such a solution is not attractive, either for the manufacturer or for the user, even if it were to work.

The number of Larus systems sold is manageable. A fraction of these are sold as the Essential version. And of these, a discounted version could potentially help. This would amount to little more than a handful of systems sold per year.

The manufacturer would lose money on this because they cannot recoup their costs.

The user would then end up with costs that are close to those of the advanced version.

In addition, the installation situation can always be resolved by mounting the essential version in the supposedly ideal location. So if the installation situation really is the problem, it can already be solved today - without additional hardware.

[Simsys]

In my opinion, the approach should be to really understand the problem. To do this, you have to bring the problem into the lab, where you can reproduce it.

The first step is to collect all relevant data; a proposal for this is already on the table (see #157).

It is quite likely that the problem can be solved by better algorithms or by fixing bugs.

[realtimepeople]

An additional magnetometer plus micro-controller and CAN phy-converter costs some 10-20 EUR, only the electronics. In comparison, the difference single vs dual GNSS is several hundred EUR. So it may very well be attractive to look at such a solution.

[Ventusfahrer]

It becomes more and more clear that mounting the sensor close to or inside of the instrument panel
is favorable because we only need short wires and tubes making the installation simple and fast.

So, with another sensor installed in a supposedly better location, you lose exactly this very advantage.

Another fact that speaks against this argument: we see the problem both in aircraft where the Essential Box is located near the canopy flasher (7v) and in those where the box is located in the supposedly more favourable location (DU, HG). In this respect, I find your argument unconvincing.

[dragonfly771]

In general it sounds like a good idea but the precise installation of the (small) magnetic sensor will be also difficult and may induce some more errors. There are indeed some difficulties with the installation of the Essential Larus in some gliders and the satisfaction level of D-GNSS users is higher. I´m not sure if the external magnetic sensor solves a lot of problems with bad installations. On the other hand, if it is possible without too much work and the hardware costs are manageable, we should just try it.

[Simsys]

In the end, it takes a lot of work to get such a product ready for series production. The device has to be designed and the interfaces defined, hardware, housing and software have to be developed. Finally, documentation for marketing and installation, including error situations, must be available.

I also don't see the benefit. The sensor box is already a small remote sensor that can be placed anywhere. How should a second sensor box help us? It will only make everything more complicated. The quantities will be negligible.

We have so many unfinished work to do. We urgently need to improve the reliability of the system. We keep seeing glitches in the functionality, especially in the essential version, but also in the dual GNSS version. In the event of a GNSS failure, we need a good fallback to a decent standard vario.

We shouldn't waste our limited developer resources. I advise against starting this project.