Most people use an electric solenoid valve from the automotive N2O industry. But there are issues with those valves, they quickly overheat under prolonged "open" condition, they also tend to be rather low flow rate valves and fairly expensive. I looked at commercial solenoid valves, at $300... um, no thanks. A very good alternative is a standard ball valve, simply 1/4 turn from off to full flow. Now the problem is, how to actuate the ball valve. I suppose the simplest method would be a set of two ropes and pulleys, of course that becomes impractical over long distances, and offers no assurance the valve truly is open or closed.
I looked at geared stepper motors, probably an ideal solution. But these require a microcontroller driving the cost and complexity up. While they do offer variable opening ability, that wasn't really needed for the ground based fill system. I looked at just solenoids as well, but I didn't find any that had both the throw distance and the power required.
I decided to take a walk around my shop and see what I had available, about the only thing I could think of that would have the torque needed would be a windshield wiper motor. It seemed a likely candidate, 12 volts, lots of torque and pretty robust construction. So I proceeded to to tear the wiper motor out of my old 1987 Chevy van. Of course the motor is mounted under the dash, if you've ever worked under the dash on a van, you'll sympathize with me. After 2 hours, a few cuts and scrapes, half the dash taken out and many curse words, the motor was mine!
The motor had two sets of wires coming out of it, one with three wires and the other a pair. After some testing, I found the three wire set was negative, fast and slow motor speed. The two wire set was a "park" indicator for where the wiper should position themselves at turn off. The park wires would be of no use, as there was just one spot where the wires went to short circuit. The fast speed seemed a little too fast, so I decided on the slow speed, and did a quick torque test by trying to hold the arm back by hand. I'm not going to say it was impossible to hold it back, but it wouldn't be easy. It seems like there should be more than enough power to actuate my valve. The way it is set up, there is an arm bolted to the motor shaft with a pin coming up at a 90 degree angle to the arm. So the pin moves in about a 2.5" circle, which looked just about right to me for mounting directly to the ball valve lever.
Here's the bottom side of the motor mounted to a plywood box.
Here you can see the motor shaft, arm and pin. I simply drilled the valve handle to accommodate the pin, then retained it with a small clip.
The valve needed a strong mount, so I made two supports from 1"x1/8" steel and welded two high pressure steel couplers to the mounts. A pair of high pressure nipples completed the mounting hardware.
Here is the finished controller.
Next I needed a way to control the valve. Since it's a DC motor, rotation can be changed by simply reversing the polarity to the motor. I accomplished that with a double pole double throw switch. The battery power connects to the center two positions, the output to one side of the switch. When the switch is thrown to that position, the motor is powered in one direction. To get it to switch directions, the second switch position contacts are reverse wired to the output set. Now when the switch is thrown the other direction the motor turns in the opposite direction. The toggle switch was a bit clumsy to flip on and off to move the motor in small increments, so I added a momentary "push on" switch to make control easier. I ran the output into an extension cord end, since that's what I use for my launch controller as well. Extension cord wire is robust, inexpensive and nice to have around anyway...
I could see there was one more thing I'd really like to have with this controller. That's a way to know for sure, if the valve is open or closed. I really debated this one for a while, but in the end I decided to keep it simple and use another wire for indicator lights on the controller. I simply mounted two momentary on switches at either end of the valve lever travel, so when the lever depressed a switch it would light up an indicator on my control panel.
Here's the backside of the controller. The two small jacks on the back are for each indicator light. I used different types of jacks so they couldn't be accidentally wired backwards. On the side you see a 4 pin XLR power jack, I use these on most of my 12 volt systems as a standard power plug.
Here's the controller opened up.
Here's the set up in action. In this picture you can see the valve is in the open position and the "open" indicator light is on.
Here the valve is in the closed position and the appropriate light is on.
I spooled up 270' of phone wire as the indicator light signal wire. A much neater package would be possible if I bought a 6 conductor cable and had everything in one wire. But I had the phone wire on hand and decided to make use of it. I used another one of my XLR ports on the valve end of the wiring to make it simple to hook up at that end.
While this certainly isn't a nice, compact, high tech system. It works very well and cost very little. I had almost everything on hand. I had to order the stainless steel ball valve, and I spent about $6 on fittings. One more piece of the puzzle come together...