Following on from my PPM Encoder mods (PPM Encoder Updated), I needed to be able to test and tune the main rotor speed controller.  I figured that a small setup was needed to simulate the pulses generated by the RPM sensor, that was motor driven and speed controllable by the ACM board.

So on the right of the photo, you see a my RPM test set consisting of plastic bucket lid, with a hole to allow light onto a sensor below.  The lid is bolted to a small brushless motor, and the ESC (which I think I just burned out) is connected to the throttle servo output on the ACM.

When it's fired up, the PPM Encoder is measuring the RPM pulse timing and outputting the result to Ch5 into the APM chip.  However, due to the method I calculate the timing, the resolution on the RPM speed is useless.

So as far as the PPM encoder mods go - its back to the drawing board.

(The things we do in the name of progress)....

If you're interested, the new PPM encoder project with RPM measurement, including the .hex file, is attached to this post.

PPMEncoder.zip

And here's a picky of the Heli static display at our local club's public display held yesterday at Hastings foreshore.  I'll upload the other photo soon on the clubs web site (www.wpmac.com.au).

My heli is on the right with the ACM connected - prompting a few questions from onlookers and members alike.  The green coloured box sitting on the muffler is a fibreglass crashproof box for the Ardupilot.  I'll provide some details on how to make one in a later post.

Now - On to the Heli code.

So today, I have got the updated PPM encoder s/ware working to do RPM measurement on one input channel.

For an RPM range of about 300 - 2000, the output pulse to the AP CPU is ranging from 2mS to 0.5mS.  The ACM code might need modding to accept this, without thinking its a sync pulse, but thats the next task.  All this will need further calibration when I test it with the motor running.

The RPM sensor is an IR LED and IR photodiode, glued into holes drilled into the frame either side of the heli.  One of the holes was used to mark the rotor shaft, which I then drilled through (offset about 2mm from center).  This gives a very clean pulse, hopefully long enough to be detected at full RPM.

Above you can see the wires leading to the photodiode, just above the main gear.

Many years ago, I tried building a model heli autopilot out of some old mouse position sensors, an 8031 micro, and my Hirobo shuttle.  The whole thing seemed to work on the bench, but quickly broke a few blades when I slammed the throttle on for an air test.  Such is life.

Now - its happening all over again, but this time, due to the excellent work done by the Ardupilot teams, I expect much better results.

So I got my APM for xmas, and its taken a month of playing to decide to put into a heli (as opposed to one of my planes).

So far, I have the following setup.

• APM with the shield.
• GPS (mediatek)
• EZ1 sonar sensor for low altitude above gnd. - Landing / takeoff assistance.
• Airspeed sensor with pitot tube (designed for plane but I think I'll put it on the heli too)
• Magnetometer
• Custom rotor RPM sensor

Airframe is Gohbee Stinger 50 (AKA JS models TZ50) - Raptor copy.

Radio is Spektrum DX7. RX is AR7000 with header board removed.

The plan is to build an autopilot system based on Arducopter, that lets you fly the @&#^$ out your heli, and not crash - regardless of your level of experience. (mine is pretty average, but I'm determined to get better).  This will be the tool for the job I hope.

When I get the hang of this blog thing, I'll post some pickys.