Before ordering the large and relatively expensive throttle quadrant PCB some real-world testing of the circuits was done, controlled by the Arduino.
The first test was to see that the clutches were correctly switched off by the limit switches.
They were, e.g. when a handle reaches full throttle position, but often (but not always) the clutches were also switched off immediately when the handle was moved back from an end position into the opposite direction, which should not happen.
As the clutch control circuit is edge triggered by a 1 -> 0 transition (falling edge) of the limit switch signal the wrong behaviour might be caused by bouncing of the limit switches.
A test with a debouncing circuit of two resistors, a capacitor and a Schmitt Trigger Inverter showed that this indeed fixed the problem.
As the inverter of the debouncing circuit flips the polarity of the limit switch signal the new 3-input NAND in the circuit had to be replaced by the original 3-input OR again and a debouncing circuit was added to the PCB design for each limit switch.
Next the limit switches, servos, clutches and position potentiometers were connected to the Arduino and some first lines of code were written to drive the servos and clutches.
Also the programmable step-down buck-converter was tested to see how much the voltage on the clutches could be lowered so that they still operate reliable and don’t get too hot during continuous operation.
A problem that can be seen in the video and that was also noticed already during earlier tests is that the left throttle handle has some backlash. Because of this it isn’t always exactly in sync with the right handle.
It seemed to be caused by the clutch but investigation showed that there was just too much space between the flat end of the D-axis that connects the servo with the clutch and the flat end of the clutch hole.
For the moment it has been replaced with one of the other D-axes which fits better but probably a new axis must be created with the CNC router to replace this one.
I’ve been thinking about a way to test the throttle quadrant handles standalone, i.e. not yet connected to the flight simulator PC as they are at a different location currently. Because the combination of breadboards, Arduino, throttle quadrant and power supplies needs more space than is available near the flightsim PC and it is also very fragile.
Once the PCB is available this will improve.
The idea that I have now for this test is to use the position of one throttle handle to drive the other one, i.e. when the left handle is moved, the right handle should follow it and vice versa.
The throttle quadrant PCB design meanwhile is almost finished and has gotten a couple of options:
– hardware direction detection for the trim wheel and flaps handle optical sensor wheels or not (can set by a jumper)
– servo or DC motor for the trim wheels; the motor can be supplied by the internal 5 Volt (jumper present) or from an external supply (connector from external supply used instead of jumper)
– clutch power switched by a relay or by a power transistor (depending on the mounted components and some jumper wires)