What Is Pulse Induction Metal Detection? (PI Circuits)
Pulse-induction metal detection sends brief current pulses through a search coil, then measures how the coil’s voltage or current decays after each pulse. Nearby metal changes that decay through electromagnetic induction. A PI circuit repeats this cycle, rejects the switching transient, and compares later samples. Coil design, current limits, clamp voltage, timing, and noise control shape its performance.
A PI detector’s hardest circuit problem is not making a coil pulse. It is safely controlling the energy left in that coil when the pulse ends, then measuring a much smaller signal without confusing it with electrical noise. A sound design treats the transmit and receive paths as related but separate jobs.
I would begin with a measured coil and a current-limited supply, not an assumed coil specification. The reference values below are starting points for a bench prototype, not guaranteed detector settings or code requirements. Adapt them to the parts, layout, and enclosure you actually use.
Planning the PI pulse and detection principle
Pulse induction, or PI, is a way to detect metal by watching a search coil’s response after its drive current is switched off. A nearby target changes the coil’s transient decay. The circuit times this change and compares it with a target-free reading; it does not directly identify every metal by a unique signal.
When current flows through the coil, it creates a magnetic field. Switching the current off makes the field collapse. By Faraday’s law, that changing field induces voltage; by Lenz’s law, the induced effect opposes the change that created it.
A conductive target, such as copper or aluminum, can carry eddy currents. Those currents create their own field and alter the coil’s decay. Ferromagnetic materials, such as iron, can also affect the response through their magnetic properties. The result depends on target size, shape, orientation, distance, and the detector’s coil and timing.
A PI circuit repeats four actions: charge the coil, switch it off, wait through a blanking period, then sample the response. Blanking means ignoring the early interval when the transmit switch and amplifier are still recovering. Later samples may reveal target effects, but their size and timing are design-specific.
A useful first test uses 12 V DC, a nominal 300 µH coil, and a 100 µs pulse at 100 pulses per second. These are example values, not universal specifications. With a 1.5 Ω coil, the idealized RL current after 100 µs is about 3.15 A:
[ I(t)=\frac{V}{R}\left(1-e^{-tR/L}\right) ]
Here, (V) is supply voltage, (R) is coil resistance, and (L) is inductance. Real wiring, switch behavior, and supply limits affect the result, so start with a 3 A current limit and confirm the waveform. The pulse repetition rate is 100 Hz, and a 100 µs pulse gives a 1% on-time.
Coil, switching, and signal-path clearances
The coil is both the magnetic transmitter and the receive sensor, so its resistance, inductance, wiring, and placement all affect readings. High-current switching also creates sharp voltage changes that can upset nearby signal circuitry. There is no single PI clearance rule; use component ratings and a layout that separates noisy and sensitive paths.
Measure coil inductance and DC resistance before choosing pulse timing. A nominal 300 µH, 1.5 Ω coil stores about 1.35 mJ at 3 A, using (E=\frac12LI^2). That energy must go somewhere when the switch opens. It cannot simply vanish, and the clamp network must manage it on every pulse.
Use a low-side MOSFET with a defined clamp path. For a nominal 60 V drain-clamp design, choose a MOSFET rated at least 100 V, then verify actual drain voltage with an oscilloscope. The rating leaves some margin, but it does not prove the circuit is safe: layout inductance can cause overshoot above the expected clamp level.
| Design item | Example starting point | What to verify |
|---|---|---|
| Supply | 12 V DC | Polarity and current capability |
| Coil | 300 µH, 1.5 Ω | Measured inductance and resistance |
| Pulse | 100 µs, 100 pulses/s | Actual gate and coil waveforms |
| Current limit | 3 A | Peak coil current |
| Clamp and switch | Nominal 60 V clamp; MOSFET rated at least 100 V | Drain peak and ringing |
| First receive sample | About 30 µs after switch-off | Repeatable decay, not switching recovery |
Keep coil leads short and paired to reduce loop area. Keep the pulse-current loop compact and apart from the receive amplifier and ADC wiring. Do not assume a fixed spacing is safe for every board. Follow the insulation, creepage, and layout rules in the actual component data sheets for the voltage and environment.
Building the transmit and receive circuits
A practical PI circuit has a transmit stage, a protected receive stage, and timing logic. The transmit stage creates controlled coil current. The receive stage measures the later decay without taking the full switching hit. Timing logic sets the pulse, blanking delay, and sample window.
Survey the parts. Measure the coil’s inductance and resistance, check supply polarity, and choose wiring and a fuse for the source and measured operating current. Put the fuse close to the supply. Set a current limit before applying repeated pulses.
Build the pulse stage. Connect the coil, MOSFET, gate driver, and clamp in a compact loop. Add a current-sense resistor or another suitable current-limiting method. Do not drive the coil from a microcontroller pin. A logic pin cannot provide the required coil current or safely absorb the turn-off energy.
Add timing and receive protection. Use a gate driver suited to the MOSFET and switching rate. Keep the receive input protected from the transmit transient. Add blanking so the amplifier or ADC ignores the initial recovery period. Do not connect an unprotected receive input directly to the transmit node, and do not sample during an uncontrolled turn-off transient.
Set and test sample timing. Begin testing around 30 µs after switch-off, then move the delay and sample window while observing the signal. That 30 µs value is a starting point, not a universal optimum. Compare target-present and target-absent readings using the same coil position and timing.
Enclose and retest. Secure the coil and board, insulate exposed conductors, and provide strain relief. Retest current, temperature, drain voltage, noise, and repeatability in the finished assembly. Enclosures and cable routing can change interference, so a successful open-bench test is not final validation.
Verifying waveforms, limits, and detection repeatability
Verification means measuring the circuit under controlled conditions, not relying on a single beep or reading. Check coil current, MOSFET drain voltage, receive recovery, and repeated target-free samples. Use a properly rated oscilloscope probe and a safe measurement setup; probe grounding can itself create a short or distort a fast waveform.
The key failure mode is an inadequate clamp. When the MOSFET turns off, coil energy can drive its drain voltage far above the supply. A drain spike can exceed the MOSFET rating and damage the switch. Do not rely on avalanche breakdown as the normal energy-management method. Confirm the clamp waveform directly.
A clamp that makes the field decay too slowly can also hide the target response. Conversely, sampling too soon may capture switching recovery rather than the target signal. Adjust one timing value at a time and keep a record of pulse width, current, clamp level, blanking delay, and sample window.
For a repeatability check, collect multiple target-free samples with the coil held still, then repeat with a target at a fixed distance and orientation. Compare the spread and average of the readings. Averaging can reduce random variation, but it cannot correct a changing coil position, unstable current, or a poorly controlled switching transient.
Two illustrative bench cases
These are worked test scenarios, not reports of field installations. Their purpose is to show how the same PI principles guide diagnosis.
Current lower than expected: With a 12 V supply and a 300 µH, 1.5 Ω coil, a 100 µs pulse gives an idealized current near 3.15 A. If the measured peak is much lower, check the current limit, supply sag, pulse width, coil resistance, and MOSFET drive. Do not increase the limit until the wiring, switch, and clamp are rated for the change.
Unstable target-free readings: If the reading changes with no target movement, first compare coil current and drain waveform from pulse to pulse. Then inspect lead routing, grounding, amplifier recovery, and sample timing. A repeatable baseline is needed before a target comparison is useful.
These checks separate a real target response from changes caused by the circuit itself. The next step is to make controlled adjustments, then repeat the same measurement.
Comparison: PI timing and circuit choices
PI circuit choices trade signal quality, switching stress, and measurement speed. No single pulse width or sample delay suits every coil and target. The table below compares practical decisions for an initial prototype; its values are test starting points, not performance promises.
| Choice | Starting approach | Main trade-off |
|---|---|---|
| Pulse current | Limit near 3 A for the example coil | More current raises stored energy and switch stress |
| Blanking delay | Test near 30 µs after switch-off | Too short may capture recovery; too long may miss useful decay |
| Clamp level | Design around a nominal 60 V drain clamp | Higher voltage may speed decay but increases switch stress |
| Coil leads | Short and paired | Smaller loop helps limit interference pickup |
| Sampling | Compare repeated averaged readings | Averaging helps random noise, not systematic drift |
There is no PI-specific building-code standard that sets these example values. For a product, identify applicable electrical-safety and electromagnetic-compatibility requirements. In the United States, FCC Part 15 limits may apply to unintentional radiators. Applicability depends on the finished device and how it is marketed or used; consult the relevant rules and qualified compliance support.
Conclusion and FAQ
A reliable PI metal detector depends on controlled coil current, a deliberate clamp, protected receive electronics, and repeatable timing. Measure the coil, start with a current limit, inspect switch-off voltage, and adjust sampling only after the waveform is understood. Treat example values as bench starting points, not guarantees or code mandates.
What does PI mean in metal detection?
PI means pulse induction. The detector pulses current through a coil, then measures how the coil’s response decays after switch-off.
How does a PI metal detector sense metal?
A metal target changes the coil’s transient response through electromagnetic induction. Conductive targets can carry eddy currents; magnetic materials can also change the response.
What is the purpose of the coil pulse?
The pulse creates a magnetic field around the coil. Switching it off creates the changing field whose decay the circuit measures.
Why does a PI circuit need a clamp?
The coil stores energy while current flows. A clamp provides a controlled path for that energy when the MOSFET switches off and helps limit drain voltage.
Can I connect a coil directly to a microcontroller pin?
No. A microcontroller pin is not a coil driver and cannot safely manage the coil’s current or turn-off energy. Use a suitable switch, driver, current limit, and clamp.
Why does the circuit ignore the first part of the decay?
The switching transient can overwhelm or disturb the receive electronics. Blanking gives the transmit switch and amplifier time to recover before sampling.
Is a 30 µs sampling delay always correct?
No. It is an example starting point. The useful delay depends on the coil, clamp, switching behavior, amplifier recovery, and target response.
Does a 100 V MOSFET make the circuit safe?
Not by itself. Verify the actual drain peak and overshoot with a properly rated probe, and check current, temperature, layout, and clamp behavior.
Do these example values define a metal detector standard?
No. The 12 V supply, 300 µH coil, 100 µs pulse, 100 pulses per second, 3 A limit, and 30 µs initial delay are prototype values, not universal specifications.
Does a PI circuit need regulatory review?
A prototype has no single PI-specific code standard, but a finished product may fall under electrical-safety and EMC rules. In the U.S., FCC Part 15 limits may apply to unintentional radiators.
(This article was written by one of our staff writers, Garrett Vance. Visit our Meet the Team page.)
