I am using the AX 1201728SG 4 output driver for the x27 gauges with 3 outputs used (RPM ASI and VSI) and the board is fed from a 5V stabilised Switch Mode PSU. The volt meter shows it at 5V or within a tenth of a volt. The Continuous rotation gauges (ADF and DG) are driven with the A3967 boards wired for full step mode with the MS1 and 2 terminals grounded and use switches for the zero as your diagrams. I tried optical but it was not as reliable as the hard switching Zero Sensors. These are fed from the same 5V PSU as I wired a 5V and a 12V bus through the cockpit that feeds all the loads. All the PSUs use a common ground and are bonded to the metal of the fuselage on the low voltage side only (for obvious safety reasons). Mains wiring is earthed as UK regulations.
The Stepper board with your firmware on it was fed from the master board 5V output but that was causing issues so I fed it and the Servo board from the 12V bus in accordance with the Arduino specs. 7-12V is OK from what they say. I have gone into the Power inlet port so the power on the board was checked and was happily at 5V when I tested it after the regulator. I did note your comments about power dissipation but as I only have small loads and limited current, and I checked the regulator was not running warm all seemed well. I could feed it from the 5V PSU but I wondered if any slight voltage drop on that would then reduce the board voltage too as it will be below the regulator output. I will however try that tonight anyway as the power busbars are close together and I could go into the Vin on the board at 5V rather than into the cannon connector or USB port. The 5V PSU is rated at 10A as is the 12V PSU so there should be ample current available before the voltage fluctuates.
Xplane FPS on my laptop is in the high 30s but I purposely kept the processor and GPU loads low for testing only . When I put the main Sim PC into the system I will be higher than that again running high graphics as its an overclocked I9 pushing a 2080TI GPU Card.
I am going to recheck all the grounds and make sure I have no earth loops and also all the voltages on the main boards, slave boards and the Stepper Driver boards so I can be sure they are all in your tolerances. I did this before but worth checking again.
One other thing, when I do a reconnect, the DG and ADF do not always do the full CCW rotation and then they go to some random point. Other times they do the full rotation, activate the Zero switch and then step to the right place but lose it later. Its mysterious and has got me really puzzled.
Usually if I do a reload it fails to do the full rotation but if I unplug and then force a hardware reconnect it works as it should.
Not sure pictures would help you as its a multiple wire loom that runs from the instrument panel to the processor boards on the front of the fuselage so nothing can really be seen together for connections. I will send you some pics of the sim for the gallery once I get it working.
I am going to re-check everything tonight in the hope I find the problem and that it is precisely as your diagrams (I think it is but again I will re-check) but in the meantime, if you have any other thoughts, please let me know.
Sorry a long message but I wanted to give you as much information as possible.
Once again many thanks for your help.