Lesson Sources and induction · Induction and spikes
Induced voltage spikes
Open an inductive DC circuit and the coil’s collapsing flux induces a voltage spike — often hundreds or thousands of volts — as the inductor tries to keep current flowing (Lenz). Arcs pit relay contacts and destroy transistors. Fix: give current a controlled path — freewheel diode on DC, MOV or RC snubber on AC/DC, or a parallel resistor so energy dissipates safely instead of across the opening gap.
When spikes happen
While coil current is steady, flux is steady — little induced EMF.
Interrupt current (open switch, transistor off, contact bounce) → flux collapses fast → large dΦ/dt → large induced voltage per e = −L (ΔI/Δt).
The minus sign is Lenz: induced EMF polarity opposes the decrease of current — it adds in series to push current through whatever path exists.
If the only path is air between opening contacts, voltage rises until arc ionizes the gap — spike energy goes into heat and pitting.
How high can it go?
Spike magnitude depends on:
- Inductance L
- How fast current changes (Δt very small → huge EMF)
- Total resistance of the loop during collapse
Example logic: 1.5 H coil, current drops 2 A in 5 ms:
|EMF| ≈ L × ΔI/Δt = 1.5 × (2 / 0.005) = 600 V
Open switch with infinite gap resistance → inductor attempts any voltage needed to maintain current — arcs measured in kV on small control coils are common without suppression.
Safe decay with a parallel resistor
Resistor in parallel with coil (or coil+driver): when supply opens, induced current circulates through coil + resistor — series path.
Higher resistance → higher peak voltage, slower decay.
Lower resistance → lower peak, faster energy burn.
Trade heat in resistor vs contact damage. Often used on large DC contactors where diode drop is unacceptable for fast release timing.
Freewheel (flyback) diode — DC workhorse
Silicon diode across coil, cathode to positive side of coil (reverse-biased while coil energized).
Switch on: diode blocks; coil charges normally.
Switch opens: induced polarity forward-biases diode → current circulates through coil + diode → voltage clamped ~0.7 V (silicon) across the loop.
Energy dissipates as heat in coil wire and diode. DC only — AC reversals would short half-cycles through the diode.
Perfect for PLC-driven relays, solenoids, small DC motors field windings where release delay of one τ path is acceptable.
MOV — bidirectional clamp for AC or DC
Metal oxide varistor (MOV) connects across coil (or supply lines). Normally megohms of resistance.
When voltage exceeds rating (e.g., 140 V on 120 V control), MOV resistance drops sharply in nanoseconds — clamps spike.
Used on AC relays, motor starters, line-side protection — conducts either polarity.
Choose rating above normal operating V but below insulation breakdown of sensitive parts.
Common on surge strips and across starter coils in MCC buckets.
RC snubber and contact protection
Resistor + capacitor in series across contacts absorbs switching energy and limits dV/dt on solid-state outputs.
Tuned for coil L and supply — too small C → still arcs; too large → leakage or slow release.
Manufacturer kits often specify snubber module per contactor family.
Combine strategies on harsh loads: diode internal to DC coil driver + MOV at board input.
Design habits for the field
- Never drive bare coil from transistor/PLC output without suppression spec.
- Replace pitted contacts with root-cause fix — snubber missing beats filing contacts every month.
- DC diode polarity wrong = short on energize — verify cathode to +V side of coil.
- AC coil → no single diode; use MOV or RC rated for AC.
- Scope or transient recorder on first article if spikes exceed device absolute max.
Field case
Situation. New building: hundreds of 24 VDC damper actuators on one PLC output card. Within six months, 30% of outputs fail shorted. Coil drivers have no diodes; actuators are inductive 0.5 H class.
Root cause: Each off command generates spike across open collector; integrated driver dies before mechanical limit is reached.
Fix: External diode at each actuator (or relay with suppressed coil), verify 5τ release still meets control sequence, replace failed modules once — not monthly.
In the field
Symptom
Pitted contacts, failed PLC outputs, nuisance GFCI, RF noise on open
Where to look
Coil suppression missing/wrong polarity, cable length, switching device rating
Likely causes
- No diode/MOV, AC snub on DC only, open circuit with no decay path
What to measure
- Peak V on scope at open
- verify diode orientation cold
What not to do
- “Just file contacts”
- reverse-bias diode across DC coil
Checklist
- I explain spike as Lenz response to fast ΔI
- I use e ≈ L(ΔI/Δt) for order-of-magnitude estimates
- I wire flyback diode for DC (cathode to + coil)
- I choose MOV for AC or bidirectional clamp
- I know resistor parallel lowers peak but burns energy
- I treat spikes as normal unless suppressed