Detecting Fake Coins: Magnet, Specific Gravity, Ping — and Why None Are Enough Alone

Every test you can run at your kitchen table can be defeated by a sufficiently determined counterfeiter. The defense is not a single test — it is a layered protocol, applied consistently.

In 2017, a London bullion dealer accepted a 1 kg gold bar bearing genuine PAMP Suisse markings, complete with assay card and serial number. The bar passed visual inspection. It passed the magnet test. The XRF gun read 99.99% gold on the surface. The bar was tungsten. Tungsten and gold share an almost identical density (19.25 vs 19.32 g/cm³), defeat surface-only XRF, and at the wholesale level can be machined to weight tolerances good enough to pass scales. The dealer caught it on ultrasonic testing, which detects the internal density discontinuity. He was lucky. The point of this article is that no single test is sufficient. The point is also that, properly stacked, simple tests will catch 99.5% of what retail buyers will encounter.

The basics: what counterfeiters actually fake

The counterfeit market is shaped by economics, not by what is most desirable to collectors. Three categories dominate:

• Low-cost tourist fakes from China. Brass or zinc-alloy "coins" with cosmetic gold or silver plating. Sold openly on certain wholesale marketplaces for under $5 each. Caught instantly by weight, magnet, or any other test. Aimed at flea-market and eBay traffic.
• Mid-tier silver fakes. Silver-plated copper or lead alloy in the form factor of American Silver Eagles, Morgans, or pre-1965 dimes. The plating thickness is often a fraction of a micron. These can fool a quick visual but not a scale or a calipers measurement.
• High-end gold bar fakes. Tungsten-cored bars with thick gold plating, machined to correct weight and dimensions. These targeted institutional and dealer purchases of 100g, kilo, and 400 oz bars. The retail risk on small denominations is genuinely low; the wholesale risk on large bars is real.

Notably, U.S. circulated junk silver is the hardest category to fake profitably. The metal value per coin is too low to support the cost of a convincing forgery. Most fake junk silver in circulation is poorly made and obvious to anyone who has handled the real thing.

The five field tests

1. The magnet test

Gold and silver are diamagnetic — very weakly repelled by magnetic fields, indistinguishable from non-magnetic at the household level. Iron, nickel, and steel are ferromagnetic. A neodymium rare-earth magnet (N52 grade, ~$10) held above or beside a real gold or silver coin will produce no detectable attraction.

What it catches: Crude steel-cored fakes. Some plated-base-metal fakes.

What it misses: Tungsten (also non-magnetic). Brass and zinc-alloy fakes (also non-magnetic). Lead and copper-cored fakes (also non-magnetic). The magnet is necessary but profoundly not sufficient. Roughly 70%+ of fakes pass it.

One useful refinement: the magnet slide test. Hold a coin at a 45-degree angle on a smooth, non-magnetic ramp. Slide a strong neodymium magnet down beside it. Genuine silver, due to eddy currents (Lenz's law), causes the magnet to slow visibly even though it is not attracted. This is more discriminating than a static magnet and catches some fakes that pass attraction tests.

2. Weight and dimensions

Every U.S. Mint coin has documented specifications. Tolerances are tight. A 0.05 g jeweler's scale and a digital caliper accurate to 0.01 mm cost less than $50 combined and catch the overwhelming majority of low-end and mid-tier fakes.

What it catches: Almost every fake outside the highest tier. A "silver" Morgan that weighs 22 g is not silver. A "Saint-Gaudens" that is 0.3 mm too thick has the wrong density.

What it misses: Tungsten gold fakes machined to spec. Plated fakes with denser cores chosen to match weight (lead in copper, etc).

3. Specific gravity

The most powerful kitchen-table test. Density of pure gold: 19.32 g/cm³. Density of fine silver: 10.49 g/cm³. Density of a 90% silver alloy: 10.34 g/cm³. The procedure: weigh the coin dry, then suspend it on a thread in a glass of water on the same scale, and record the difference. Specific gravity = dry weight ÷ (dry weight − submerged weight). Online calculators handle the math.

What it catches: All non-tungsten gold fakes. All non-exotic silver fakes. Lead-cored fakes (lead is 11.34 g/cm³, which on a small piece is detectably wrong).

What it misses: Tungsten gold (19.25 vs 19.32). Coins coated in tap water bubbles (re-dunk and tap to remove). Lacquered or plated coins where the coating shifts the result.

A purpose-built specific-gravity scale (Sigma Metalytics sells one, as does GemOro) automates the procedure for under $300.

4. The ping test

Strike the coin gently against another coin or balance it on a fingertip and tap it. Silver, gold, and copper have characteristic acoustic signatures. 90% silver U.S. coins ring with a clear high-pitched tone that sustains for 1–3 seconds. Cupronickel-clad clones thud and damp out almost immediately. Lead-cored fakes give a dead "tunk" sound.

Apps like CoinTrust (iOS) or Bullion Test capture the audio frequency profile and compare it against a database of genuine pings. They are surprisingly effective on common items.

What it catches: Most silver-plated base-metal fakes. Some gold fakes.

What it misses: High-quality silver-content fakes (sterling silver fakes can ping similar to 90%). Coins struck from real metal with wrong percentages. Gold bars (the form factor is too different from coins for ping comparisons to work cleanly).

5. Sigma Metalytics, XRF, and ultrasonic

The professional tools. Sigma Metalytics Precious Metal Verifier (~$700–$2,000 depending on probe set) measures electrical resistivity at multiple frequencies, which is genuinely metal-specific and difficult to fake. Used by major dealers as a first-line test. XRF (X-ray fluorescence) guns like the Bruker Tracer or Olympus Vanta read elemental composition with high accuracy — but only of the surface, typically the top 100 microns. Ultrasonic thickness gauges measure speed of sound through the metal and detect internal voids or density discontinuities, the only practical way to catch a tungsten core in a thick gold bar.

“A test that catches 99% of fakes is not a defense. It is an invitation to the 1%.”

Deeper context: why a layered protocol works

Each test catches a different subset of fakes. The intersection of "fakes that pass test A" and "fakes that pass test B" is much smaller than either set alone. A practical retail protocol:

1. Buy from someone whose reputation is the test. APMEX, JM Bullion, SD Bullion, Heritage, Stack's Bowers, GreatCollections, or a local shop you've vetted. The fake-coin failure rate from these sources is essentially zero.
2. Visual inspection. Compare to a known-genuine reference image. Look at edge reeding, lettering depth, and surface texture under magnification.
3. Magnet slide. 30 seconds. Catches the bottom tier.
4. Weight and calipers. Two minutes. Catches the middle tier.
5. Specific gravity (for anything over $500). Five minutes. Catches everything except tungsten cores in large gold.
6. Ping test. Confirms what the previous tests already established.
7. For large gold purchases: Sigma or third-party authentication. Or buy already-slabbed PCGS / NGC graded coins, which are vetted on submission and re-vetted on resale.

Common mistakes

• Trusting the seller's "test it yourself" demonstration. Some sellers carry a real coin to the meeting, demonstrate it passing all the tests, then hand you a different coin from a pocket. Conduct your own tests, on the coin you are actually buying, before money changes hands.
• Believing the "acid test" is sufficient. Nitric acid only tells you about the surface. Plated fakes pass acid tests until the plating wears through.
• Buying graded slabs without verifying the slab. PCGS and NGC slabs themselves are counterfeited. Verify the certification number on the grading service's website. Real slabs have specific security features (PCGS holographic logos, NGC barcode placement) that fake slabs typically miss.
• Skipping testing because "it came from a friend." Most precious-metals fraud at the retail level passes through trusted hands. The friend bought it from someone who lied to him.
• Assuming new buyer protections cover you. eBay's buyer-protection program excludes precious metals over a threshold value. PayPal's protection has similar carve-outs. Read the fine print before assuming the platform has your back.

What to do next

Order a 0.05 g jeweler's scale, digital calipers, and an N52 neodymium magnet today — total cost under $40 from any general-merchandise retailer. Buy a single, known-genuine 1923 Peace dollar from a reputable dealer. Run every test on it. Get a feel for what real silver weighs, how it pings, how the magnet slides past it. That single afternoon's calibration session is the foundation of every authentication you will do for the rest of your collecting life.