A typical metal detector listing reads like someone dumped a datasheet into a word salad generator. 12 kHz operating frequency. 11-inch DD coil. Target ID 00-99. Notch discrimination. Multifrequency. Four figures, zero field context, and if you’re shopping for your first machine, none of it tells you whether this thing will find a quarter or chase ghosts in a park.
Here’s the decoder premise: every number on that listing controls exactly one physical thing. Frequency controls the signal cycles per second. Coil size and shape control the geometry of the field. Target ID measures how fast a returned signal decays.
Three spec families actually change what happens in the dirt: operating frequency, discrimination/target ID, and coil geometry. The rest are ergonomics, presets, or marketing rounding. And none of them work alone, which is the part the spec sheet never tells you.
Two findings along the way that make this worth reading to the end. First, the frequency printed on the box is often not the frequency the machine emits: a White’s Classic II was labeled 6.6 kHz when it actually ran at 6.592 kHz. Second, a platinum wedding ring once read on screen exactly like a pull tab, at target ID 76, and was only found because the detectorist ignored the number.
Key Takeaways
Operating frequency is a conductivity trade: low frequencies (3-7 kHz) reach deeper for silver, copper, and brass, while high frequencies (14+ kHz) are what actually detect small gold; the box label is often a rounded number, not a measured constant (White’s Classic II: printed 6.6, measured 6.592).
Target ID is a measured phase shift, not a label, but it’s a probability table, not an answer key: one beach detectorist found 50+ pull tabs all reading ID 76-77, then a heavy platinum wedding ring the next morning at ID 76 too.
Coil geometry is a terrain decision, not a depth knob: bigger means deeper but heavier and blinder to small targets in trash, and DD coils stay stable in mineralized or salty soil specifically because their overlapping windings cancel their own magnetic field.
Table of Contents
Operating frequency: what the kHz number really means
Operating frequency is the number of transmit/receive cycles per second, measured in kilohertz, so 12 kHz means the coil sends and listens 12,000 times every second. That’s the whole definition. Everything else is what that number buys you.
The decoder beat here is that the number on the box is a rounded label, not a measured constant. The White’s Classic II was originally labeled 6.592 kHz, then printed as 6.59, then rounded to 6.6 in later listings. That’s not a manufacturer lying, it’s standard rounding practice. But when the Huygens Optics teardown put an oscilloscope on a Garrett AT Pro, it measured 15.3 kHz coming off the coil. Close enough to the spec to be honest, but you get the point: spec sheets round.
General-purpose machines cluster in the 5-15 kHz range because engineer George Payne pegged roughly 12.5 kHz as the sweet spot for all-around detecting, which is why the Tesoro Compadre ships at 12 kHz and most coin/jewelry/relic machines live in that band. Single-frequency units tend to sit at or under 10 kHz or above 30 kHz, since the extremes serve specific niches rather than general use.
That’s the whole kHz story for general detecting. Gold prospecting moves to 17-70+ kHz, and that’s the next section’s problem.
Frequency by target: low vs. high and the conductivity trade
Low frequency penetrates deeper and favors conductive targets like silver, copper, and brass; high frequency excels on small, low-conductivity targets like gold nuggets and thin jewelry. That’s the trade in one sentence, and it’s conductivity physics, not brand preference.
Low frequency: the depth play
At 3-7 kHz you’re in silver, copper, and brass territory. Low frequencies also handle iron, pipes, cables, and relics reasonably well. The win is penetration: a low-frequency signal gets deeper into the ground before fading. The loss is sensitivity to small stuff. A thin gold chain is basically invisible to a machine tuned for silver dollars.
High frequency: the gold play
At 14 kHz and up, and especially in the 17-70+ kHz gold prospecting range, the machine gets genuinely good at small gold nuggets and fine jewelry. The costs are two-fold: less depth, and higher sensitivity to mineralized soil, which is where ground balancing (letting the detector calibrate out ground minerals) earns its spot on the feature sheet.
The frequency-by-target chart
Approximate, conductivity-based, and not gospel:
- Coins: 10 kHz and lower
- Large objects: 7 kHz and above
- Small objects: 7 kHz and below
- Ferrous/iron: 10 kHz and above
- Gold: 14 kHz and above
- Silver, copper, brass: 3-7 kHz
- Nickel, aluminum: 4-8 kHz
A forum thread from long-time engineers adds nuance with appropriate hedging: dropping to roughly ±5 kHz may respond slightly better to high-conductivity coins with a possible depth gain, and 20-25 kHz may favor nickels and thin gold at a slight depth cost. These are tendencies, not guarantees.
The honest caveat that kills the “right frequency = treasure” forum folklore: frequency alone doesn’t determine performance. Coil size and shape, circuitry design, signal processing, recovery speed, and discrimination quality all matter. Picking the perfect kHz and ignoring everything else is like optimizing a CPU clock speed and shipping it in a machine with no RAM.
Single-frequency vs. multifrequency: what the premium actually buys
Multifrequency is a real capability, but in Wirecutter’s 2024 testing it wasn’t demonstrably better for beginners. That’s the verdict, and it cuts against most of the marketing.
Single-frequency continuous-wave VLF detectors, like the Garrett Ace 400, are cheaper and simpler. The weakness is mineralized soil and wet salt beaches, where ground minerals and salt conductivity drown the single signal you’re listening for.
Multifrequency evolved in two generations. The first used FBS (Minelab’s original approach). Newer models run FMF or SMF, which is what you get on the Nokta Makro The Legend and XP Deus II. The acronyms matter less than the capability: multiple frequencies at once, so low-conductivity gold and high-conductivity silver both get picked up in one sweep.
Where multifrequency genuinely earns its premium: wet saltwater sand, mineralized ground, and situations where you’re chasing low- and high-conductivity targets simultaneously. The worked example from Wirecutter’s testing is clean: the single-frequency Simplex Lite underperformed the multifrequency Minelab Vanquish 440 on mineralized saltwater sand. On a dry park lawn, that gap mostly closes.
But here’s the contrarian core: the advantage only appears once you’ve learned to adjust settings to exploit it. Experienced detectorists prize multifrequency for exactly that reason. A beginner with default settings can’t access most of what the premium buys. If you’re shopping a first machine, you’re paying for capability you won’t use for months. That’s not “multifrequency bad,” it’s “know your stage.”
There’s a third technology worth knowing: pulse induction (PI), which sends a powerful current pulse and measures how long the signal takes to decay to zero volts. PI is largely immune to ground mineralization, which makes it the saltwater and deep-ground specialist, and it’s strictly high-end territory.
Discrimination and target ID explained: the physics behind the numbers
Discrimination mode on a metal detector identifies metal type by measuring the phase shift of the returned signal, which is the speed at which the metal’s induced field decays. When the coil’s field hits a metal object, it induces eddy currents in that object. Better conductors produce stronger eddy currents and a higher phase shift. The detector reads that shift and translates it into the number on screen.
This is where it gets genuinely cool. The Huygens Optics oscilloscope teardown measured phase shifts per material and found that target ID is literally a time delay you can watch on a scope:
| Material | Phase shift | Target ID |
|---|---|---|
| Silver | ~180° | ~90+ |
| Aluminum block | 176° | Well over 90 |
| Lead | 160° | 55-80 |
| Aluminum foil | 82° | As low as 40 |
| Iron | ~20° | 10-25 |
| Ferrite | 0° | ~00 |
The Garrett AT Pro’s 00-99 target ID scale is derived directly from this measured phase shift. That number on the display isn’t magic, it’s an oscilloscope reading wearing a costume. I find this honestly delightful.
Where the ID lies
Now the failure case, fully told. In one documented beach outing, the detectorist dug more than 50 pull tabs, all reading ID 76-77. The next morning, the first find was a heavy platinum wedding ring, also reading ID 76. It got found only because the detectorist ignored the number and dug anyway.
That’s a common beach-hunting pattern: one repeating ID all session, and then the “bad” ID produces the good find. I’m not going to tell you this happens every time, but I will tell you it’s documented.
The reason: target ID depends on material plus shape and size. A crumpled foil square and a ring can land in the same range because shape and size contribute as much as material. So the ID chart is a probability table, not an answer key. Digging is generally the safer choice when the signal sounds interesting.
Field note: When one ID range repeats all session, dig the next one anyway. Shape and size can make trash and treasure read identically.
That said, discrimination has genuine upsides. It filters trash in junky ground, and in areas with old military ordnance, where grenades can still be in the soil, discrimination can literally be a lifesaver. Not all features have to be fun to be important.
Notch discrimination and adjustable iron audio are the two settings most listings tout. Notching lets you blank out specific ID ranges you’ve decided are trash, and iron audio adjusts how loudly iron signals report. Both are common trash-filtering tools you’ll see on nearly every spec sheet, and both are exactly as good as your assumption about which IDs are trash.
Coil size: depth vs. sensitivity, with real numbers
Coil size on a metal detector trades detection depth against sensitivity to small targets. Bigger reaches deeper but is worse at finding the fine stuff; smaller finds the fine stuff but quits on deep targets. That’s the whole trade.
Large coils (13-15 inches)
Best for caches and deeply buried relics, where depth is the point. The costs: they’re heavier, more susceptible to electromagnetic interference, and blind to tiny targets in trashy ground. A 15-inch coil over a pull-tab mine just averages the noise.
Small coils (5-8 inches)
Exceptional sensitivity to fine jewelry and small nuggets, easier to swing for hours at a stretch, and at their best in trash-filled sites and dense woods, where a small field means better target separation. This is the coil you want in a junky park, not the 15-inch one.
Medium coils (9-12 inches)
The general-purpose balance for coins and relics. Most first-time buyers should probably be here and don’t know it.
The buyer trap
Here’s the failure pattern: first-timers buy the biggest coil listed because depth is the spec that sounds impressive, then discover it’s unusable in a trashy park, where weight over long swings plus poor separation turns a fun afternoon into arm-day punishment. So is bigger better? Yes for depth, no as a default.
Buyer rule: Match coil size to the ground you’ll actually hunt — medium for parks and relics, small for trash, large only when depth is the mission.
Real numbers as anchors. The Nokta Simplex Ultra’s 11-inch DD coil covers 95 square inches. The Minelab X-Terra Pro’s 12×9 covers 83. The Vanquish 440’s 10×7 elliptical covers 55.
Those aren’t cosmetic differences. And the fatigue math: roughly 1,000 swings per hour-long hunt, so ounces are a real spec, not trivia. Two pounds at swing 50 is a different hobby than two pounds at swing 900.
Jeff Lubbert, who has been detecting since 1980 and has recovered roughly $980,000 worth of finds, puts a number on it: at least 11 inches under most conditions. His second rule matters more: own both a compact and a large coil rather than hunting for one perfect size. Aftermarket options exist from Minelab, Garrett, Nokta, XP, Coiltek, NEL, and Detech if you want to build that kit.
Coil shapes and configurations
The right configuration depends entirely on terrain: mineralized soil and saltwater favor Double-D coils, low-mineralization ground and easy pinpointing favor concentric, and serious PI-driven depth favors monoloop. Shape, meanwhile, is mostly ergonomics.
Round, elliptical, open-web
Round coils offer stability and even depth and sensitivity across the sweep, which is why gold prospectors tend to favor them. Elliptical coils navigate rocks, bushes, and tight terrain better. Open-web coils have cutouts that shed weight and drag, useful for water hunting and long sessions.
Concentric
Inner transmit winding and outer receive winding, producing a cone-shaped field. Easiest pinpointing since the target is dead center, and great on low-mineralization ground. Weaknesses: poor mineral filtering, needs tighter sweep overlap, and can get noisy in mineralized soil.
Double-D
Two overlapping D-shaped windings. The physics is genuinely elegant: the overlap makes the net magnetic flux through the receive coil zero, meaning the coil is blind to its own transmitted field but highly sensitive to disturbances caused by metal targets. That’s the mechanism behind why DDs dominate mineralized and salty ground. They pinpoint near the heel of the coil, and on Minelab GPX detectors, DD coils enable Iron Reject. High-end detectors can run two DD configurations in a single coil.
Monoloop
One winding doing both transmit and receive duty, used on PI detectors like the Minelab SD, GP, and GPX series. Produces the deepest cone field, often more depth than DD in comparable conditions, but demands precise ground balancing in mineralized soil or it’ll lie to you all day.
Scenario pairings, quickly: medium DD for coins and relics, round monoloop for gold in mineralized soils, open-web DD for beach and saltwater work.
Quick VLF context if you’re new: the coil’s transmitter and receiver generate the field, the metal response is what the control box interprets, and proper swing technique is wide arcs parallel to and close to the ground, driven by torso rotation rather than wrist flicks. We cover the full onboarding in how to use a metal detector for the first time.
How the settings interact, and where the ground wins
Ground conditions can override every other spec. The same detector changes personality between a quiet park and a wet saltwater beach, and the spec sheet won’t warn you because the spec sheet assumes ideal ground.
The hidden coupling first: larger coils produce lower effective frequencies, and smaller coils produce higher ones. So kHz and coil inches aren’t two independent line items like a listing implies; in design, they’re partly one decision. That’s general design physics rather than a hard constant, but the coupling is real.
Mineralization punishes high frequencies, and ground balancing is the standard mitigation, with pulse induction as the design that sidesteps the problem entirely. On heavily mineralized ground, lower frequency is usually the right pick. The concrete proof is the same one from earlier: the single-frequency Simplex Lite losing to the multifrequency Vanquish 440 on wet saltwater sand. Different design, same ground, different result.
Sensitivity is a reach-versus-stability trade. In mineralized or trashy ground, turning sensitivity down buys you a stable signal at the cost of some depth. If the machine is chattering constantly, that’s not a defective unit, that’s a setting.
Preset modes are a readable design philosophy if you know to look. The Simplex Ultra’s presets are environment-based (beach, park, field). The Vanquish 440’s are object-type (coin, relic, jewelry). Two detectors can both say “preset modes” and mean opposite things. Read the philosophy, not the mode count.
If you already own a machine, your manual’s frequency chart is the logical follow-up now that the vocabulary decodes.
Reading a spec sheet like a detectorist
Reading a metal detector spec sheet means translating each line item into a field consequence before you care about the price. Once you can do that, the sheet stops being marketing and starts being an engineering document.
The line-by-line decoder
- Frequency ? the conductivity trade (low for silver/depth, high for gold/sensitivity)
- Coil size and type ? depth vs. separation, terrain fit
- Target ID range ? a probability table, not a label
- Discrimination/notch ? trash filtering with known failure modes
- Sensitivity ? reach vs. stability
- Waterproof rating ? where you can hunt at all; the XP Deus II at 20 meters vs. 5 meters for other picks is a real difference, not a rounding
Tested picks as case studies
Each of these proves a tradeoff already covered above.
The Nokta Simplex Ultra is Wirecutter’s beginner top pick: 11-inch DD coil (95 sq in), 2.6 pounds, waterproof to 16 feet, environment-based presets, roughly 12 hours of battery, Bluetooth headphone support. In testing it pulled a gold ring, a steel bolt, and a copper piece from 8 inches deep. That’s the beginner benchmark.
The Minelab Vanquish 440 is the multifrequency-at-a-reasonable-price runner-up: 10×7 elliptical DD (55 sq in), AA batteries, object-type presets, and only the coil and pole waterproof. In testing it found a child’s toy six inches deep in the surf, which is the multifrequency-saltwater advantage in one anecdote. All planted targets were found by every tested detector, but the Equinox 900 pinpointed all three most precisely.
The Minelab X-Terra Pro runs an 83 sq in coil, weighs 5 ounces more than the Simplex Ultra, and has notably better magnetic charging.
The Minelab Equinox 900 offers six customizable frequencies and haptic grip feedback, which is genuinely useful for hearing-impaired detectorists. It also costs more than triple the top pick.
The XP Deus II is the underwater specialist, added to the guide in an August 27, 2026 editor’s update: 12 preset modes, less beginner-friendly, and “optional” accessories that aren’t really optional, like a $749 underwater remote and $280 WS6 headphones. Real total cost of ownership, not just sticker price.
The Minelab Manticore sits in the expert tier with a 2D ground map showing depths and IDs, plus 10 search modes, and it’s Jeff Lubbert’s expert recommendation.
What the sheet leaves out
Anti-recommendations are curriculum too. Lacey Metal Detectors, a specialist retailer that opened in 2001, supplied one such verdict: the Bounty Hunter Tracker IV, around $100 with no display, is a machine people use once and put away in a closet forever. The Garrett Ace 300 is decent under $300 but outclassed on features and ergonomics. The Nokta FindX Pro kit, under $200, is a capable VLF machine that performs poorly on wet sand and underwater. Two of the four brands on Amazon’s 2022 most-popular-detector list were gone by 2024, which tells you something about market churn in this space.
The spec sheet also won’t mention what ownership feels like. A pinpointer with ferrous discrimination (Minelab’s is waterproof with five sensitivity levels) matters in trashy parks. Headphones, preferably Bluetooth, are effectively required for beach hunting, where soft signals get drowned out. Warranty and service: Nokta advertises a three-year warranty on the Simplex Ultra, but customer service is reportedly poor or unreachable, while Minelab brought repair in-house and rates highly for speed. That difference matters more at year two than any spec at day one.
Try before you buy if you can. Local clubs and store trial hunts exist, and specialist tour operators in the US and UK arrange permission to search farms and private land. Meanwhile, the ethical baseline: fill your holes, don’t trespass, ask permission, pack out your trash. Detecting is prohibited on federal property and often restricted or licensed on public land, so check local rules before you swing. Expectations, honestly set: you’ll mostly dig worthless scraps, and practitioners treat the search itself as the point.
Specs inform, ground decides
No single setting guarantees finds. Frequency is a conductivity trade. Target ID is a probability table. Coil geometry is a terrain decision. The spec sheet tells you what a machine can do; the ground tells you what it will do, and the gap between those two is where experienced detectorists live.
If you’re starting from zero and want help narrowing models, our how to choose a metal detector guide walks the full selection logic. And if you’re weighing cost against capability, how to choose a metal detector for the money breaks down which features justify their price. Otherwise: find a local club, get your hands on machines before spending, and enjoy the search. That’s the actual hobby.
Frequently Asked Questions
Is a bigger coil better on a metal detector?
Bigger reaches deeper, but it’s heavier, more prone to electromagnetic interference, and blind to small targets in trashy ground. A 15-inch coil over a pull-tab-strewn park just averages the noise, while a small 5-8 inch coil excels at separation in junk. Most first-time buyers should start with a medium 9-12 inch coil and treat large coils as a depth-specific tool, not a default.
What are the different settings on a metal detector?
The three spec families that actually change what happens in the dirt are operating frequency (the conductivity trade), discrimination and target ID (trash filtering with known failure modes), and coil geometry (depth vs. separation and terrain fit). Sensitivity is a reach-versus-stability trade, ground balancing calibrates out soil minerals, and preset modes vary in philosophy — some machines organize presets by environment (beach, park, field) while others use object type (coin, relic, jewelry). Everything else on a listing is ergonomics or marketing rounding.
What size coil is best for metal detecting?
It depends entirely on the ground you’ll hunt. Medium coils (9-12 inches) are the general-purpose balance for coins and relics and the right starting point for most buyers. Small coils (5-8 inches) win in trash-filled parks and dense woods thanks to better target separation; large coils (13-15 inches) are only for caches and deeply buried relics where depth is the mission. Many experienced detectorists own both a compact and a large coil rather than hunting for one perfect size.
Are multi-frequency metal detectors worth it for beginners?
Usually not as a first machine. In Wirecutter’s 2024 testing, multifrequency wasn’t demonstrably better for beginners: the advantage shows up on wet saltwater sand and mineralized ground, but only once you’ve learned to adjust settings to exploit it. A beginner on a dry park lawn with default settings can’t access most of what the premium buys — you’d be paying for capability you won’t use for months.
