Power supply for a Solenoid ?

[petemoore]

  I wrapped about 300 turns around a 1/2'' bobbin, and this is the solenoid coil.
  I spits out a ferrous slug with a neodynim magnet on it pretty good, even with just 9v battery.
  I want to try 12v, 24v on it, and have FWR'd [and filtercap'd] a transformer [big] with secondary taps for 9/12/24vac.
  Alternatively I have a SMPS which says 16vdc and 24vdc.
  Something about an inductor coil of the solenoid makes me thing that even with short duty cycles the PS will see this as a near short and...
  I still want to mess around with these solenoids but would like to be able to overkill the power supply by at least ''a bit'' in order to avoid blue smoke releases.
even though it's 'yer basic solenoid coil', I have no numbers for equating other than seems to warm up a bit if I unload a battery into it with longer/repeated duty cycles.
 

[R.G.]

The average power of the power supply is almost immaterial. What counts is its ability to spit out peak currents, yes?

So put a resistor from the power supply to a BFC* sufficient to not overload the power supply if the capccitor is shorted. When the cap fills up from the resistor, then short the capacitor into the coil. The capacitor powers the coil, and the power supply itself is not shorted.

This is a page directly from the book of the coil-gun people, who use capacitor banks to fling steel projectiles.

* Big Freaking Capacitor 

[petemoore]

  I've got one in there, a BFC IEC and all of course.
  I was wonderin' if may be 2 BFC'd be better with a resistor inbetween.
  oh boy I was googlin' thanks for the tips.
  I saw many for sale [SMPS mostly] and not a whole lot on the power supply for solenoids, managed anyway to find one as you've described, showed a 2k resistor before each BFC per power bank [see: "pinball machine backbox".

[R.G.]

The thing about solenoids is that they're current devices, not voltage devices. It trips up us voltage-centric types.

What you want a solenoid for is the magnetic field. The M-field is determined by the number of turns (N), the current through the turns, and the magnetic permeability of the space inside and wrapping around the outside of the coil.

Putting a voltage on the solenoid induces a current change in the winding determined by the inductor equation: V = L*di/dt . Put another way, the current ramps up and down at a rate determined by the voltage across it divided by the inductance.

There are a couple of things to notice in that equation that are not obvious until you've messed with solenoids a bit. First, there is no built-in limit for the current. As long as you provide a constant voltage across it and can supply the current from an external supply, the current can, in theory, ramp up forever. The other is that "L" is usually thought of as constant. In a solenoid, L is not constant.

L is the inductance, and it depends on the permeability of the space inside the coil and wrapping around the coil. The inside of the coil is the magnetic slug you're putting in there. It moves. As it moves into and out of the coil, that changes the magnetic path length and the gap in the magnetic path, with the non-ferromagnetic parts becoming a permeability of air (that is, unity) and the iron parts being several thousand.

The bottom line for voltage-centric types is that the inductance of a solenoid changes radically, and the current has to be limited by something else. The 'something else' is usually the resistance of the coil wire. That's how solenoids get a "voltage" rating. The resistance of the wire is put to good use to limit the coil current when the metal slug is missing or when the voltage is left on. The current ramps up to I = V/R, where V is the supplied voltage and R is the total resistance of the wire in the coil. What this does is to keep the coil wires from burning out whether or not the slug is in the center of the coil. It also slows down di/dt and makes it the result of a differential equation instead of algebra.

So solenoids get separated into continuous and intermittent duty. For continuous duty, like holding a relay "on" or opening a sprinkler valve, or any place where it must be held with the plunger inside the coil, the voltage rating and coil resistance are set up so if you put the recommended voltage on the coil, the resistance stops the current at a safe level greater than the hold-in value. For intermittent duty, you can run the coil current as high as you like, assuming things will cool down before the next use. That intermittency may be a few seconds or several hours.

I had a section on coilguns at GEO from back when I gadflied over to looking at them for a while. For coil guns, the on time is so small that every use is considered the only use, with long times between uses, so the current can be so high that things heat up to just below destruction. These coils are pumped from capacitors pumped to high voltages to store enough energy (E = 1/2 * CV²) to change the current in the coil FAST (V = L*di/dt) and get large energies into the coil ( E = 1/2 * L*I²) to move the slug fast. This sounds something like what you're doing.

The capacitors only need to charge up over time, like a camera flash capacitor. In fact, lots of coilgun enthusiasts use camera-flash capacitors. The caps can be fed slowly from a low *power* supply to high energies, as power is energy divided by time. Energy stored for a long time, released very quickly  all at once.

In your case, the cap can be "loaded" to some voltage and then fired. The power supply can be so trivially smaller than the coil current that it takes seconds or minutes to load up the cap. When the cap is dumped into the coil, the power supply essentially does not matter. It's only the cap's voltage (equals stored energy by E = 1/2 * CV² ) and the power supply can't possibly keep up.  The power supply's power and current rating only matter when you are thinking about how soon you will take the next shot. Holding in a solenoid for a gate, door, valve, etc. needs a power supply that does the full current all the time. Running it intermittent means the power supply only needs to run up the voltage over time. If the solenoid is powered for one second every minute, the power supply needs to provide energy for one minute to the cap, and then can even be disconnected for the shot, then reconnected. Additional R-C filtering is counterproductive. Use all your capacitance for the shot.

This duty is hard on the *caps* too. The cap current ramps up with the inductor current obviously, but goes to large values. So the power wasted in the cap's ESR gets significant in heating the cap, and the mechanical forces of electrostatic repulsion and attraction inside the cap's plates get significant. That's why flash capacitors are special things.