In short
No consumer machine reads through a gold layer, so the most accurate tester for jewelry is the one that sees deepest: XRF at a jeweler or refiner, which reads 20 to 50 microns into the metal and is usually free with a sale. Electronic testers with a probe and gel read the surface to about one karat and are sorting tools for stamped scrap, not lie detectors for unknown pieces; heavy plating and gold filled read as gold on them. Acid kits are surface tests too but cost $30 and, with a filed notch, catch most plating. Bullion verifiers such as the Sigma Metalytics PMV read the whole coin or bar, which is exactly why their maker says they are not for jewelry. For a one-time sale of family gold you need no machine at all: a density check at home and the buyer's XRF do the job.
Contents
What each machine actually measures
| Tester | What it measures | How deep | Reads karat? | Fooled by | Cost |
|---|---|---|---|---|---|
| Acid kit and touchstone | how the streak reacts to graded nitric acid | the surface, about a micron | yes, in steps: 10, 14, 18, 22 | gold filled and heavy plating unless a notch is filed | $20–40 |
| Electronic karat tester, probe and gel | electrochemical response of the surface | the surface, a few microns | yes, roughly to one karat | plating, gold filled, coatings, dirty or lacquered surfaces | $200–600 |
| Density kit: 0.01 g scale and water | weight over volume of the whole piece | the whole piece | no; a density band per karat | hollow pieces, stones, tungsten cores | $20–30 |
| Bullion verifier, electromagnetic | bulk resistivity against a known alloy | the whole coin or bar | no; pass or fail against a chosen alloy | not made for jewelry, by the maker's own statement | $999–3,500 |
| Handheld XRF | elemental composition from X-ray fluorescence | 20–50 microns | yes, to a fraction of a percent, with every alloy metal listed | a thick shell over a core; refiners file a spot | tens of thousands; free at buyers and refiners |
| Fire assay | gold recovered by melting a sample | the sample | yes, the reference method | nothing; it destroys the sample | refiners only |
Read the "how deep" column first. Every test that reads the surface passes gold filled and heavy plating, because the surface of those pieces is real karat gold. The home tests page shows what each free test proves; this page is about the machines people are tempted to buy.
Electronic testers: what the karat readout means
The pen-probe testers sold for jewelry, the GemOro AuRACLE family being the best known, touch a probe to the metal and read the electrochemical response of the surface through a conductive gel or a dry contact. The AGT3 is specified for 6 to 24 karat and platinum, calibrates against a 14k or 18k reference piece, and shows a karat and a gold percentage on a display; it launched at $549 and current models sit in the same range. Used on a clean spot of solid gold, it lands within about a karat of the truth in seconds, without acid, which is why scrap buyers use one to sort a tray of stamped pieces quickly.
What it cannot do is see below the surface. A gold filled watch case reads 12k because its surface is 12k gold; a heavy 18k plate on brass reads 18k until the layer is scratched through; a lacquered or dirty surface reads nonsense. Manufacturers say to file a spot before testing anything suspect, at which point the machine is doing what a $30 acid kit does. Buy one if you sort volumes of stamped gold and want speed without acid. Do not buy one to authenticate a single unknown chain from a stranger; for that, density and a notch beat it.
XRF: the answer jewelers actually use
An X-ray fluorescence gun excites the atoms in the metal and reads the light they emit, giving the percentage of every element present, gold, silver, copper, zinc, nickel, palladium, in a few seconds and without a mark. It reads 20 to 50 microns deep, enough to see through electroplate and to catch the brass under a thin gold filled layer, and it distinguishes 14k from 15k and 18k from 17k, which no acid or probe does. The printout is also the honest way to know what a white gold alloy contains before a nickel-sensitive person wears it.
Handheld XRF units cost tens of thousands of dollars, which is why the practical route is to use someone else's: refiners, pawn shops, coin dealers and many jewelers run one on anything they might buy, for free, and will usually show the screen. Ask for the reading before the offer. Its limit is a thick shell over a core, a bar problem more than a jewelry one; refiners file a spot or drill a suspicious bar. For a jewelry buyer, XRF plus a scale is the end of the argument.
Bullion verifiers are not jewelry testers
The Sigma Metalytics Precious Metal Verifier and its relatives measure how a small electromagnetic field passes through the whole sample and compare the result to the resistivity of a chosen alloy, so a tungsten core inside a gold bar, which passes weight, stamp and surface tests, fails in a second. They read through plastic holders and take no sample. Prices run from about $999 to $3,500 by model.
The maker is explicit that none of its devices are designed for jewelry: a ring or chain has no fixed geometry and no single reference alloy to compare against, so the reading means nothing. Buy a verifier if you buy coins and bars; for jewelry, spend the money on a scale, an acid kit and the bus fare to a refiner.
Acid kits still earn their place
A kit is graded nitric acid solutions for 10k, 14k, 18k and 22k, a black touchstone and a file. Rub the piece on the stone, apply the acid for the karat you expect: a streak that survives is at least that karat, one that dissolves is below it or not gold. It costs the price of lunch, keeps for years, and with the file used on a hidden spot it catches gold filled and heavy plating that fool the electronic probe. The costs are a tiny mark on the piece, gloves and ventilation, and no gemstones near the acid.
Combined with a pocket scale and a glass of water for a density check, an acid kit covers what a private seller or a flea-market buyer needs for under $60. It does not give a percentage, so for pricing scrap the buyer's XRF still has the last word.
Which to buy, by situation
| Situation | Buy | Skip |
|---|---|---|
| Selling inherited gold once | nothing; a pocket scale for weights, then the buyer's XRF | any machine; it will not change the offer |
| Buying at estate sales and flea markets | acid kit, file, 0.01 g scale, a magnet | electronic tester as the only tool |
| Buying and selling gold regularly | electronic tester for sorting, acid kit for doubts, a refiner's XRF for settlement | trusting the probe on unstamped pieces |
| Coins and bars | a bullion verifier, or the dealer's | using it on jewelry |
| A jewelry store or pawn counter | handheld XRF | acid as the only test on gold filled |
The most accurate tester is not a product, it is a sequence: stamp, magnet and weight at home, acid on a notch if anything is doubtful, XRF before money changes hands. Fire assay sits behind all of it as the reference refiners settle on, and it is the one test none of these machines replace.
Fakes that beat each machine
- Heavy plate and gold filled beat acid on the surface and every electronic probe. Only a notch, density or XRF catches them.
- Tungsten-cored bars pass stamp, magnet, density and a shallow XRF; only a verifier, ultrasound or a drill finds them. They are a bullion problem, not a jewelry one.
- Fake stamps cost nothing: 14K on brass, 750 with an Italian-looking star, 18K on plated steel. A stamp is a claim, and the hallmark guide covers which marks are guarantees.
- Hollow chains pass everything and simply weigh less; the chain weight calculator gives the solid figure to compare.
- Plated over silver, vermeil, passes acid and probe, and rings true in density only because silver is heavier than brass; XRF reads the silver at once.
Questions people ask
Are electronic gold testers accurate?
On clean solid gold, roughly to within a karat, which is enough for sorting stamped scrap. They read only the surface, so gold filled, heavy plating and lacquered pieces read as gold. Makers advise filing a spot first, which is what an acid kit does for a tenth of the price. Use one for speed on known material, not to authenticate an unknown piece.
What is the most accurate gold tester?
For jewelry, XRF: it reads every element in the alloy to a fraction of a percent, 20 to 50 microns deep, without a mark. Fire assay is the reference behind it and destroys the sample. Handheld XRF costs tens of thousands of dollars, so use a refiner's or a pawn shop's, usually free, and ask to see the reading before the offer.
How much does an XRF gold tester cost?
Handheld units run to tens of thousands of dollars and are built for refiners, dealers and jewelers who test all day. Nobody selling a few pieces should buy one; refiners, coin dealers and pawn shops run XRF on anything they might buy and will show you the screen. The practical cost of an XRF test is a visit.
Can a gold tester detect gold plated jewelry?
Surface testers cannot: acid on the surface and electronic probes read the plating as gold, and gold filled reads as its karat layer. Filing a notch and testing the base metal catches plating, a density check catches brass under gold, and XRF reads through electroplate. Anything sold as a machine that spots plating from the surface alone is overpromising.
Is the Sigma Metalytics verifier good for jewelry?
No, and the maker says so: its devices compare a sample's bulk resistivity to a chosen coin or bar alloy and are not designed for jewelry of any kind. For coins and bars it is the fast way to catch tungsten cores. For a ring or chain, use density, an acid notch and a buyer's XRF.
Is a gold acid test kit worth buying?
Yes, for anyone who buys gold from strangers: for about $30 it grades 10k, 14k, 18k and 22k, and with the file on a hidden spot it catches gold filled and heavy plating that beat electronic testers. It marks the piece slightly, needs gloves and ventilation, and cannot give a percentage, so a refiner's XRF still settles the price.