#2223 an24b: ARK (ADF) does not receive signals and remain 'operable=false'

[Ludovic Brenta]

This may be a bug between chair and yoke (i.e. a misunderstanding on my part) but here goes.

1. Launch the an24b at LOWI.
2. Autostart (I normally start manually but for the sake of reproducing this bug, this is fine).
3. Go to the radio navigator station (press 4).
4. On the left ARK-11 panel click the lower-right button "Switch Control over here".
5. Referring to the QuickStartupGuide.pdf page 9, set the switch 2 to "KOMPASS I", enable switch 3 "PROGRAM", rotate knobs 5, 6, 7 to select 420 kHz, press button labelled 1, disable "PROGRAM".
6. Observe that the "paper" sheet now displays INNSB 420.0. Note that the Innsbruck beacon is only a couple nautical miles away, so we should be receiving its signal.
7. Press button 1 again to recall the programmed frequency.

Symptoms:

1. The reception indicator remains at zero.
2. The radiocompass needles do not move.
3. /instrumentation/adf/operable alternates between "true" and "false" very quickly.
4. /instrumentation/adf/in-range stays 'false'.
5. The RMI on the pilot's panel is unusable.

The same symptoms also hold for the second ARK-11.

[James Turner]

The oscillation of the 'operable' property sounds like the ARK-11 is attempting to do some manual computation on the operable flag, which is not a great idea. Will need to see if it's just this aircraft or others.

[Ludovic Brenta]

Oh it's definitely just this aircraft. It has a lot of custom nasal for all instruments including the ARK radios, which even changes the value of /instrumentation/adf[0..3]/operable in a listener. I have not figured out what calls this listener repeatedly.

[James Turner]

Probably what's needed, is for the aircraft Nasal (or XML rule?) which drives 'operable' to instead drive power-button (which is an 'input' property, whereas 'operable' was supposed to be an output one).

That would cause the aircraft logic to turn the C++ radio on / off, which I guess is what was desired here.

[Ludovic Brenta]

thanks for the hint, this allowed me to fix the bug. There were two, actually. One was that the an24b set /systems/electrical/outputs/adf to the string '1' instead of a double number; the other was that it was driving /instrumentation/adf[]/operable instead of .../power-btn.