DIY Solenoid

After trying out many different solenoids for the pinball machine, and being very disappointed with almost all of them, I decided to try making one by myself.

After reading a couple of text from the internet, it seemed that making one should be easy. Just roll an insulated copper coil around a nonconducting core. And add a metallic pole inside the core. And the power of the magnetic field should just be dependent on the number of loops and the current, or B = uNI or strength = air resistance * loops * current. At least according to this website. Of course, the length of the coil increases with the loops, which in turn increases resistance, which reduces the current as long as the voltage remains the same.

I found a piece of plastic pipe and almost perfect bolt that fit inside the pipe from our Fablab. I used them as the core and the pole for the solenoid.

The copper wire itself was rolled around the core with an electric hand drill, where the core was attached to the drill where drill bits are normally attached to, and the wire was initially held in place with a tape. Tape was also use after coiling to keep the wire from unspooling afterwards. I left about 10cm of slack on both sided of the looped wire to attach them to electical apparatuses.

I taped tips of my fingers and side of my palm, so that they would not burn as meters of wire would pass along them. Gloves would probably have been better, but I did not have them at my disposal then.

A better solution would probably have been to drill a hole to the core pipe a bit farther than the bolt could reach, and feed the coil through it before rolling the coil around the core.

After looping the wire, I sanded the wire ends with steel wool and scraped it with a box cutter, to remove the insulation from the wire. The insulation would hinder attaching electrical connections to the wires.

The first copper coil I tried was 38 gauge copper wire. First I tested if the wire had an insulating enamel (or similar) layer on top of it by measuring the resistance along the wire. As the multimeter showed 0 ohms, it was enough proof that no electricity went through it, which indicated some kind of insulation. I looped it around the core about 200 times.

The solenoid was very weak, barely noticeable when lifting it agains the bolt. But there clearly was some kind of magneticism happening. As most tutorial mentioned that a couple hundred loops should be enough, I asked around.

We decided to measure the resistance of the wire, and noticed that the electricity consumtion for the coil was 24 volts and ~800mA, which would mean that the wire had a reistance of around 30 ohms. Which was a lot, given that 200 loops around a 1 cm diameter core is just pi * 1cm * 200, i.e. around 7 meters. This would mean that the wire had resistance of about 3-4ohms per meter. The wire was a bit too thin to properly conduct electricity though it.

I also measured the resistance, proving that the calculations were correct enough.

After that I tested a couple of different thicker wires. One wire was taken from a dumpster, and another wire was found from a cabinet. Neither wire had any better markings to differentiate them from any other wire. I cut a one meter long strip from both wires, and measured the resistance trough it using a multimeter. The bumpster wire had a resistance of 0,7ohms and the cabinet wire’s resistance was around 0,4ohms. With those measurements done, I decided to do another solenoid with the wire from the cabinet. I made it the same way I had done the previous solenoid, with around 200 loops. The loop count was probably a bit less this time, as the drill ran out of battery.

The new solenoid was clearly powerful enough to pull the bolt towards it. It was not as strong as my previous store bought ones, but at least it worked.

Given that my strongest bought solenoid was pulling around 1,25A with 24V, which would mean that the resistance of that solenoid is 24V/1,25A = 19ohms. If we assume that the wire inside of it has similar resistance per meter than the wire from the cabinet (i.e. 0,4ohms per meter), it would mean that there is about 40 meters of wire inside that solenoid. My better solenoid only had at maximum 7 meters (pi * diameter * loops) of wire inside it, probably less. Increasing the loop count should increase the power.

Next day

I decided to redo the solenoid with more loops. I recreated my spooling setup and started spinning. This time I did drill a hole to keep the wire initially in place, but I couldn’t use it to fasten the last slack, as the spooled wire was too tight and covered the drilled hole. I needed to use tape to fasten it in place.

I wasn’t gonna count the number of spins in my head, as I targeted to a number closer to a thousand. From previous day, I expected one full length of the core to contain about 100 loops, so I was gonna count only the full up and down cycles. And even then, I ended after 8.

I ended up a lot less than a thousand loops this time. Multimeter showed a total resitance of the solenoid to be 5.2 ohms, which would indicate a total wire length of about 13-14 meters. 13-14 meters would mean 400-450 loops of wire.

But the loop count did not really matter. The solenoid is already a beast, at least compared to my bought ones. I did not measure the strenght of the solenoid with any pulley, scale or Newton measurement device, but the power consumption and ball speed of the new solenoid is quite something already. 107 watts of power consumption (with 24V and 5 amps) is quite much. And the ball moves at a reasonable speed for a small pinball machine.

Lift test.

Ball launch speed.