Sitemap

The LC Oscillator Inside a Proximity Sensor Barrel

5 min readSep 10, 2026

--

An M18 inductive proximity switch draws no more than 13 mA with nothing in front of it. That current is not idling. It pays for an oscillation that has to keep running at a constant amplitude for years, in a panel that gets to 55 degrees in May, on a supply that sags when the drive next to it accelerates. Everything the sensor does, and every tolerance on its datasheet, comes out of what happens to that oscillation when a piece of steel approaches the face.

The tank circuit behind the sensing face

Behind the plastic face of a cylindrical inductive switch sits a coil wound on a ferrite core. The core is open towards the front, which is why the field leaves the barrel in one direction instead of closing on itself. The coil is one half of a resonant circuit; a capacitor is the other half, and an active stage feeds back just enough energy to keep the pair oscillating. That is the LC tank.

The field in front of the face is therefore alternating, not static. On an Accent D30C the face material is PBT and the barrel is nickel-plated brass, with stainless steel available on request. Brass is non-magnetic, so it does not short out the field it surrounds. Where the metal housing extends forward to the sensing face, the sides of the coil are screened and the sensor can be buried flush in a steel bracket. Where it does not, the field spreads sideways as well as forward, the sensor reaches further, and it needs a metal-free zone around the face.

What a target does to the oscillation

Put a metal plate, painted or bare, in an alternating magnetic field and the field induces circulating currents in its surface. Those eddy currents dissipate real power as heat in the target’s own resistance, and that power comes out of the tank circuit. As the target closes in, the coupling improves, the loss rises, and the amplitude of the oscillation falls. Push the target close enough and the oscillation collapses altogether.

A demodulator stage rectifies the oscillation and produces a DC level proportional to its amplitude. A trigger stage compares that level with a threshold and switches the output. A target sitting on the threshold with a little machine vibration would make the output chatter, so the trigger has deliberate hysteresis, specified on the D30C range as not more than 15 per cent of Sn. On an M18 flush sensor with a 5 mm rated distance that is up to 0.75 mm of difference between the operate and release points.

Get Aniket Patil’s stories in your inbox

Join Medium for free to get updates from this writer.

How much energy a target drains depends on what it is made of. Mild steel is the 100 per cent reference, chrome-nickel about 90 per cent, brass 50 per cent, and aluminium and copper 40 per cent. Copper reading shorter than steel surprises people who expect the better conductor to be detected better. Steel adds magnetisation losses on top of the eddy-current losses, while the very low resistivity of copper means the induced currents oppose the incoming field more than they dissipate energy from it. Less dissipation means less amplitude drop, so the target has to come closer before the trigger sees it.

Why the barrel diameter sets the range

Reach is set mainly by how large a coil will fit and how far its field projects. The coil cannot be wider than the bore of the thread, so the thread size on the drawing is a ceiling on the sensing physics. Below are the published rated distance Sn and assured distance Sa for the D30C series, all in millimetres. The coil itself is not a published dimension; the bore is simply the ceiling it cannot exceed. The range fills in the gaps at M6.5 and M36, and the rated and assured distances for each barrel size are worth reading against your own bracket depth.

Thread  Sn flush  Sa flush  Sn non-fl  Sa non-fl
------ -------- -------- --------- ---------
M8 1 0.8 2 1.6
M12 2 1.6 4 3.2
M18 5 4.0 8 6.4
M30 10 8.1 15 12.1
M50 n/a n/a 30 24.3

Range does not scale linearly with diameter. An M8 flush sensor reaches 1 mm, or 12.5 per cent of its thread diameter. An M30 flush reaches 10 mm, or 33 per cent. An M50 non-flush reaches 30 mm, 60 per cent of its diameter. Coil area grows with the square of the diameter, and the unscreened front face lets more of that field escape forward.

The same physics pushes the other way on speed. A larger coil stores more energy, so it takes longer to collapse and longer to rebuild. Switching frequency on the D30C range therefore falls from 5000 Hz at M6.5, M8 and M12, to 2000 Hz at M18, 1000 Hz at M30, 500 Hz at M36 and 100 Hz at M50. Reach and speed are traded against each other by the choice of barrel, before anyone writes a line of PLC code.

What the electronics behind the coil have to hold steady

The oscillator, demodulator, trigger and output stage occupy a few millimetres of board behind the coil, and their job is to keep the trip point still while the world moves around them. On the D30C series the supply may be anywhere from 10 to 30 V DC and the ambient from minus 10 to plus 70 degrees Celsius, with temperature drift across that whole span held within 10 per cent of Sn and repeat accuracy over eight hours within 10 per cent. The unit is expected to survive 1 mm of vibration from 10 to 55 Hz for 30 minutes in each of three axes, 30 g of shock for 1 ms in each axis, and 1000 V AC at 50 Hz for one minute across its insulation.

Two protections are not uniform across the range. Reverse polarity protection is standard. Short-circuit protection is not fitted on M6.5, M8 or flush M12 sensors, and is fitted from non-flush M12 upwards. Where it is fitted, the behaviour is a gradual shutdown rather than a latch: the switch becomes inoperative while the short is present and recovers on its own once it is removed, with no need to cycle the supply. On the smaller barrels the wiring discipline has to supply that protection instead.

Fitting all of that behind a coil is a packaging problem as much as a circuit problem. Accent Controls took thick-film hybrid microelectronic circuitry into Indian proximity switch manufacture in 1987, through its collaboration with Kastl Electronic of Germany, and moved to surface mount technology in 1990. Both steps had the same point: fewer hand-made joints behind the coil, and less of the drift that comes with them. The 13 mA the sensor draws while doing nothing is the price of holding that trip point still.

Application engineering team, Accent Controls, Mumbai.

--

--

Aniket Patil
Aniket Patil

Written by Aniket Patil

0 followers

Writer at Peepals, a seven-section news network. https://peepalsglobal.com/