The answer is more easily found by modeling the coil and surrounding elements.
First of all, we're generally not dealing with coils in the abstract, but flyback / inductive EMF is only a concern where it's an unintended/inevitable side-effect of a component, like a relay or solenoid.
Coils, inductors, in general use, we aren't concerned with the EMF because
that's entirely the reason we've chosen them. Whether as a resonant coil in a tuning network, or the active element in a flyback converter, the EMF is fully constrained by the circuit it is embedded within.
So, assuming you're working with relays or solenoids here.
Typically such a coil has an overall L || R characteristic, due to eddy currents in the core. This is exaggerated for solenoids (usually a solid piece of iron), significant for small solenoids and relays (often using pieces of stamped metal, less thickness = less time for field to penetrate the material), and optimized out of real inductors and transformers (using laminated sheets, or materials like ferrite that carry significantly less eddy current). For turn-off flyback purposes, we can assume the core linear, and therefore characterize it at signal amplitudes.
The easiest way to characterize a coil is probably, using a sine generator and scope, connect the generator through a resistor to the solenoid, and measure the voltages at both ends of the resistor, with respect to ground. (So, scope probes, generator and solenoid are all common ground, and resistor connects to generator and solenoid, and both scope probes.)
Use a calculation like this,
https://www.seventransistorlabs.com/Calc/RLC.html#vecand tabulate values over a range of frequencies. Say 10Hz to 100kHz or more, taking a few points per decade, more if you find an interesting "kink" in the response.
Likely, you'll need to switch resistors from time to time, or voltage scales; just make note when/where you're doing that.
Then plot it: take abs(Z); compare Re(Z) to Im(Z); etc. Most likely you'll see Re(Z) ~ Im(Z) over much of the range (the signature of eddy currents in bulk metal), maybe you'll see a long asymptote where it behaves this way, or in fact behaves inductively (|Z| ~ F).
The maximum impedance |Z|, at whatever frequency corresponds to turn-off rate, gives the peak flyback voltage for that device, when turned off at that rate.
You might never be able to achieve such rate; for example a 1A solenoid driven by IRF520 will still be loaded by the few 100 pF of the transistor, plus some 10pF and kohms equivalent of whatever probe you connect to it. You would be able to infer, from these data, that something very fast and high performance, like an air spark (switching can occur within some ~ns; fractional ~pF loading is possible), could generate such-and-such voltages.
For coils other than electromechanical ones, the answer is much simpler: don't use it that way! The EMF is not something god-given, it's an expression of the coil's response into the surrounding circuit. Simply don't let that voltage appear, and you won't see it! Hence clamping diodes or zeners being a common sight in coil driver circuits. An SMPS fully constrains the inductor voltage at all times (between switching transistors or clamping/rectifying diodes); even the boost/flyback converter (suitable for generating high voltages) can be constrained by, not just device characteristics (you can't go higher than the avalanche rating of transistors/diodes used..), but the control circuit as well.
Tim
Home
Help
Search
Login
Register
