Not quite Galloping Ghost, but I've designed & built a pulser for the old Adams-style magnetic actuators, its driven by the rudder (or aileron) channel of a conventional modern radio set. It connects inline between the rx and the mag actuator - so the receiver thinks the "pulser-plus-magnetic-actuator" combination is a standard propo servo.
The pulser comprises just two 8-pin chips, mounted in a 16-pin socket and heatshrunk, so its quite small and weighs just 5 or 6 grams.
It has a full bridge output so drives the mag actuator in both directions alternately, at 7 cycles per second, exactly like the old pulse proportional sets did.
The proportional output from the receiver channel precisely controls the mark-space ratio with 'mark' driving the actuator one way and 'space' driving it the other way. The rudder clacks away just like the old pulse sets of the 60s.
Tech details, the processor is a pic 12F683, chosen because it has hardware pulse-width modulation, which makes it easier to simultaneously measure and generate accurate pulses. The driver is a MIC4428YN mosfet full H-bridge capable of 1.5A output. All the clever stuff is done in the PIC software.
Its very easy to make (bug-style) if anyone has a redundant Adams actuator or similar. Theres a video of the installation in action, and also a flight video on my Youtube channel Philg2864.
The prototype was fitted into Shaun’sex-DB model which is fitted with an Adams Baby magnetic actuator, andis controlled by the rudder channel of his conventional Spekky DX7 2.4g radio.
I've made a few pulsers now, bug style, and although they're not as neat as I'd like, there's a worthwhile weight saving doing it this way and now I've settled on a method, the result is neat enough. Construction is a bit fiddly, but we need it to be kept small & light and I though I'd show how its made.
The circuitry is just 2 active components, a pic and a bridge, both 8-pin chips, so I start with a 16 pin low-profile DIL socket which will hold both chips. Here's the layout looking at the underside of the DIL socket:
Unused pins are snipped off the socket, and an SMT 100uF is soldered onto the socket between pins 1 and 16 which will be the power rails on Pic pins 1 & 8.
The pic output and the two bridge inputs are folded down flat and connected with soldered kynar wire:
A strip of 1/4" by 3/4" card lays over these pins for insulation, then the input lead is added next, above the card. Pos, neg and the servo signal:
On top of that the output connections are added, laying all the wires neatly along the module:
Thats the wiring finished, so time to plug in the chips. Note that the marker pips are opposite each other - the H-bridge chip is upside down wrt the pic:
Heatshrink the lot to finish it off, the result is a 5 or 6 gramme module capable of driving 1.5 amps at a varying duty cycle proportional to the rudder channel input:
Heres a brief clip of it working, unfortunately the youtube frame rate cant keep up with the movement but in practise its very smooth and precise. In fact you're better off just listening to the actuator & ignoring the video!:
There are two ways to connect the actuator and its probably best to experiment to see which gives the most torque. One way is with both actuator coils in series (ignoring the centre-tap) and driven in opposition by the bridge:
The other is to take the actuator centre-tap to battery pos and have the bridge alternately ground each coil:
In either case you will most probably need some resistance in series with the actuator connections, sat 10 to 20 ohms, as the original actuators were intended for 2.4v from either half of a centre-tapped 4.8v battery. Without the resistor, the actuator and pulser may get hot.
So there you go. 1960s pulse proportional in 2012, and on 2.4g!
Cheers
Phil