Wireless switch for power

If I was terse before, let me start again.  One complication in electronics is that there are two load parameters: Resistive(Re) and reactive(Im).  Resistance, except in the world of op-amp filters and gyrators is typically a positive number.  Positive (imaginaries) reactance implies 'inductor' or 'coil,' negative 'capacitor.'  This contrasts with mechanical reactance, which is a single-parameter joint typically called 'inertia.'  Resistors are (typically) IR lightbulbs, they act like a frictive element.  Reactors store energy, almost invariably to return it later.  The exception would be some EMP devices.  Inductors (what you got) would store energy in the magnetic field, capacitors in the charge field.  Capacitors are, typically, profoundly more volumetrically efficient than similarly-sized coils, making them more popular.  For the physics wonks in the audience, I feel I must note the symmetry breakdown:  Inductors exhibit induction, so we can make things like transformers and transductors with them.  Caps, not so much.  Anyway, your load is +Re,+ Im.  If you turn on a fan, get it up to speed, then unplug it, you will likely see a spark.  This is an inductive load.  If you plug in an old, large analog stereo amplifier, you will see a spark when you plug it in.  This is a capacitive load.

So your load has a resistive term, the resistance of the wire in the coil.  This wire is long and thin, so R is non-trivial.  In shunt with this load, is an inductance.  Without gettin' too wonkish, there is the natural free-range inductance of your coil, + a larger, less-well behaved inductance caused by the insertion of the (probably ferrite) slug that makes the solenoid.  Electromagnets typically don't have great reach, as the force diminishes as the square of distance, so without a linear motor, rotary motor + mechanics, or fluidics, you cannot move stuff very far this way.

Anyway, this shunting inductance diminishes in influence the longer we energize, its a logarithm thing, so it never ends except that our universe ain't made so much out of continuous stuff.  We have a duration that is when 0.68 of the final value (here current, capacitors potential) is reached.  We call this value a time-constant which we typically measure in seconds.  Now, there is a subsequent moment when we attempt to shut off this load.  Your powerQ is probably a (greater than an ampere-and-half static rating) MOSFET.  These moieties, especially, like all transistors, would probably prefer to operate in a class-D or saturation mode.  If they are off, our Qdiss is low because we have little current, if they are on, we have little potential difference across our conduction channel.  Remember Power (Watts) = Potential difference (Volts) * Current (Amperes).  So ya' turn off the power device quickly.  This leaves our magnetic field in place.  It collapses.  It induces an instantaneous potential of a magnitude great enough (for practical purposes, physics wonks) to maintain the current in our conduction channel.  Its total energy is the 'area under the curve' we provided to get things going.

Fortunately, your load does not require a 'soft landing.'  Certain stepper motor drives (and stepper motor situs) and large solenoid-operated valves require a complex clamping scheme.  You almost certainly don't.  What you need is to shunt the load with a diode that seems 'backwards.'  This sort of clamp is of no avail across the switch, one is there anyway.  Bipolar stepper drives that are really tiny use this diode, but the surge in QA is quenched by by QA' and vice-versa, through auto-transformer action. I digress.   Alternately, we could quench with a Zobel network, passive and fast, but we might be ringing into November.  Hybrid solns are possible and work great, model them in Spice. How do we rate our Q and D?  Max current is set by the purely resistive model of our load and is thus inferable from Ohm's law.  Remember to derate!  Breakdown potential is supply potential, remember to derate, plus spike voltage, which is largely dependent on the capacitive reactance of our diode.  Derate, derate, derate!  Don't go crazy-high though, as the way the manufactures achieve (presuming like technology) is to reduce the conductivity of the bulk conductivity by doping less.  Practically speaking, if your potentials are low enough, you can use Shottky diodes.  At higher potentials, I would employ UF (ultra-fast) diodes.

So what happens if I don't clamp?  Maybe nothing.  Typically, though, hot electrons will poke little holes in my power switch and it will degrade over time.  I knew some guys in a machine shop in the eighties who thought their stepper-motors were wearing out (not too likely) it was probably their drivers power Qs.

==========

Wanna play a philosophical game?  Propose imaginary terms for real values.  For example one possible imaginary term for the for a commodities' value is: fungibility.  I digress.

If I was terse before, let me start again.  One complication in electronics is that there are two load parameters: Resistive(Re) and reactive(Im).  Resistance, except in the world of op-amp filters and gyrators is typically a positive number.  Positive (imaginaries) reactance implies 'inductor' or 'coil,' negative 'capacitor.'  This contrasts with mechanical reactance, which is a single-parameter joint typically called 'inertia.'  Resistors are (typically) IR lightbulbs, they act like a frictive element.  Reactors store energy, almost invariably to return it later.  The exception would be some EMP devices.  Inductors (what you got) would store energy in the magnetic field, capacitors in the charge field.  Capacitors are, typically, profoundly more volumetrically efficient than similarly-sized coils, making them more popular.  For the physics wonks in the audience, I feel I must note the symmetry breakdown:  Inductors exhibit induction, so we can make things like transformers and transductors with them.  Caps, not so much.  Anyway, your load is +Re,+ Im.  If you turn on a fan, get it up to speed, then unplug it, you will likely see a spark.  This is an inductive load.  If you plug in an old, large analog stereo amplifier, you will see a spark when you plug it in.  This is a capacitive load.

So your load has a resistive term, the resistance of the wire in the coil.  This wire is long and thin, so R is non-trivial.  In shunt with this load, is an inductance.  Without gettin' too wonkish, there is the natural free-range inductance of your coil, + a larger, less-well behaved inductance caused by the insertion of the (probably ferrite) slug that makes the solenoid.  Electromagnets typically don't have great reach, as the force diminishes as the square of distance, so without a linear motor, rotary motor + mechanics, or fluidics, you cannot move stuff very far this way.

Anyway, this shunting inductance diminishes in influence the longer we energize, its a logarithm thing, so it never ends except that our universe ain't made so much out of continuous stuff.  We have a duration that is when 0.68 of the final value (here current, capacitors potential) is reached.  We call this value a time-constant which we typically measure in seconds.  Now, there is a subsequent moment when we attempt to shut off this load.  Your powerQ is probably a (greater than an ampere-and-half static rating) MOSFET.  These moieties, especially, like all transistors, would probably prefer to operate in a class-D or saturation mode.  If they are off, our Qdiss is low because we have little current, if they are on, we have little potential difference across our conduction channel.  Remember Power (Watts) = Potential difference (Volts) * Current (Amperes).  So ya' turn off the power device quickly.  This leaves our magnetic field in place.  It collapses.  It induces an instantaneous potential of a magnitude great enough (for practical purposes, physics wonks) to maintain the current in our conduction channel.  Its total energy is the 'area under the curve' we provided to get things going.

Fortunately, your load does not require a 'soft landing.'  Certain stepper motor drives (and stepper motor situs) and large solenoid-operated valves require a complex clamping scheme.  You almost certainly don't.  What you need is to shunt the load with a diode that seems 'backwards.'  This sort of clamp is of no avail across the switch, one is there anyway.  Bipolar stepper drives that are really tiny use this diode, but the surge in QA is quenched by by QA' and vice-versa, through auto-transformer action. I digress.   Alternately, we could quench with a Zobel network, passive and fast, but we might be ringing into November.  Hybrid solns are possible and work great, model them in Spice. How do we rate our Q and D?  Max current is set by the purely resistive model of our load and is thus inferable from Ohm's law.  Remember to derate!  Breakdown potential is supply potential, remember to derate, plus spike voltage, which is largely dependent on the capacitive reactance of our diode.  Derate, derate, derate!  Don't go crazy-high though, as the way the manufactures achieve (presuming like technology) is to reduce the conductivity of the bulk conductivity by doping less.  Practically speaking, if your potentials are low enough, you can use Shottky diodes.  At higher potentials, I would employ UF (ultra-fast) diodes.

So what happens if I don't clamp?  Maybe nothing.  Typically, though, hot electrons will poke little holes in my power switch and it will degrade over time.  I knew some guys in a machine shop in the eighties who thought their stepper-motors were wearing out (not too likely) it was probably their drivers power Qs.

==========

Wanna play a philosophical game?  Propose imaginary terms for real values.  For example one possible imaginary term for the for a commodities' value is: fungibility.  I digress.