Coins deeper than 20 cm

Rob Henke — A field guide to metal detecting: coil | robhenke.com

What this guide is

An independent field reference built from stated, testable behaviour: measured coil depths, published target ID bands and the reporting duties named by the authorities themselves.

What it is not

Not a shop, not a manufacturer's brochure and not legal advice. Depth figures are typical rather than guaranteed, and the rules summarised here should be confirmed with the named authorities before you rely on them.

25-35 cmdepth a 25-30 cm coil reaches on good ground

5-15 kHzsingle-frequency band with the deepest reach

14 dayswindow to report an old find in England and Wales

The short answer: a bigger coil on a lower frequency

For coins deeper than 20 cm, fit a 25-30 cm coil, run the lowest frequency the ground allows, and judge signals by repeatability rather than by the number on the screen.

A 15 cm coil — the size most machines ship with — typically reaches 15-20 cm on good ground, enough for fresh drops but not for a coin that has been sinking for a century. The 25-30 cm coil reaches 25-35 cm, putting a coin at 20-30 cm inside working range instead of at the ragged edge of it. On mineralised soil a 40 cm coil reaches 35-50 cm, at the price of weight and speed.

Frequency decides how much of that depth you keep. Lower frequencies go deeper: a single-frequency 5-15 kHz coil out-reaches an 18-38 kHz coil, though it separates small gold from trash less well. A wide-band multi-frequency coil sweeping roughly 4-30 kHz is the flexible option, finding silver coins at 20-30 cm while still catching gold at 10-15 cm.

The signature mark of this site, drawn as a plate

Match the frequency to the ground

Mineralisation above about 10-15% by volume makes a single-frequency machine chatter, and that is when 15-18 kHz earns its keep.

Ground mineralisation is the silent variable in every depth claim. Above about 10-15% by volume, a single-frequency machine produces chatter — false signals that sound like targets and bury real ones — and the usual fix is to run 15-18 kHz, which calms the ground response at some cost to small-target finesse.

The baseline coin-hunting setup assumes ordinary soil: 12-15 kHz, a 15 cm elliptical coil, sensitivity at 70-80% and a discrimination notch around iron at 15-20. Change one variable at a time — coil first for depth, frequency second for ground — and re-test after each change.

The signature mark of this site, drawn as a plate

The setup table

Coil and frequency setups against the depth they actually reach
Setup	Realistic depth	Ground it suits	What it does best
15 cm coil at 12-15 kHz	15-20 cm	Good ground	Everyday coin hunting
25-30 cm coil at 5-15 kHz	25-35 cm	Good ground	Coins deeper than 20 cm
40 cm coil	35-50 cm	Mineralised soil	Maximum depth on difficult fields
Single frequency, 5-15 kHz	Deepest for the coil fitted	Clean to moderate ground	Depth over small-target separation
Single frequency, 18-38 kHz	Shallower for the coil fitted	Clean ground	Separating small gold from trash
Wide-band, 4-30 kHz	Silver at 20-30 cm; gold at 10-15 cm	Variable ground	Mixed coin and jewellery sites
15-18 kHz on any coil	Set by the coil fitted	Mineralisation above 10-15%	Calming chatter and falsing

Coil and frequency setups against the depth they actually reach

Terms in plain language

Target ID: The number a detector assigns to a target, running 0-95 on most machines. It estimates conductivity, so different metals — and different depths — can share one number.
Discrimination notch: A filter that rejects a chosen band of target IDs, commonly set around iron at 15-20. Whatever genuine target reads inside the notch is rejected with the junk.
Mineralisation: The share of magnetic and conductive minerals in the soil. Above about 10-15% by volume it makes a single-frequency machine chatter and usually calls for 15-18 kHz.
Air test: A bench check with targets held in the air — typically a US quarter at 15 cm and a euro at 20 cm. A healthy machine returns the same ID across 3-4 sweeps.

A twenty-minute routine before the first hole

Five checks in order, each with a named failure that tells you to stop and fix settings instead of digging blind.

The whole routine takes about twenty minutes at the start of a session and then seconds per signal. Each step has a failure mode with a name, and each failure has a fix that does not involve a test hole.

Step 1, air test, five minutes: sweep a US quarter at 15 cm and a euro at 20 cm; both should return the same ID across 3-4 sweeps. Failure: ID drift — the number wanders between passes, so settings or ground are unstable.
Step 2, ground check, two minutes: sweep clean ground at 70-80% sensitivity. Failure: falsing — chatter with no target present, the mark of mineralisation above 10-15%; move to 15-18 kHz.
Step 3, notch and gain, one minute: set the discrimination notch around iron at 15-20 and keep sensitivity at 70-80%. Failure: depth loss — maximum gain on bad ground costs more depth through noise than it buys.
Step 4, sweep pattern, all session: overlap passes so the coil works within its 8-12 inch discrimination spacing. Failure: masking — a coin beside iron reads as low iron because the machine never got a clean look.
Step 5, two-angle check, per signal: re-sweep from a second direction before digging. Failure: one-way signal — a tone that sounds in one direction only is usually junk or ground effect, not a coin.