Coffee Refractometer Guide: TDS, EY, Scientifically Precise Extraction

Updated: 7 September 2026

3 key takeaways

• A refractometer turns an optical property, the bending of light, into a concentration expressed as a percentage of dissolved solids.
• The coffee conversion is TDS = Brix × 0.85, so 1.50% Brix reads as roughly 1.28% TDS.
• Three coffee-calibrated models dominate European availability: the DiFluid R2 Extract, the Atago PAL-COFFEE (BX/TDS) and the VST LAB Coffee III.

You've dialed in your grind, nailed your dose, and the shot looks beautiful, but something's still off in the cup, and that is exactly where a coffee refractometer earns its place. Is it under-extracted? Too dilute? Is the recipe actually consistent from brew to brew? These are questions your palate can answer partly, but a refractometer answers them with numbers. This guide stays on the instrument and its protocol: what the prism actually reads, why Brix and TDS are different numbers, what the published precision figures are worth, and how to draw a sample that does not lie.

What exactly does the instrument read in your cup?

TDS stands for Total Dissolved Solids. It expresses, as a percentage, how much dissolved material is present in your brewed coffee. A filter coffee at 1.25% TDS contains 1.25 grams of dissolved solids in every 100 grams of liquid. Those solids include acids, sugars, soluble lipids, aromatic compounds, and caffeine.

Why does this matter? Because TDS is the measurable correlate of what you perceive as "strength" or "body." Too low a TDS and the coffee feels thin and watery, even if it's technically correctly extracted. Too high and it feels dense, almost heavy, nuances get drowned out.

One boundary matters here. TDS is a concentration, nothing else. Turning it into an extraction yield needs the dose and the beverage weight as well, and that calculation belongs to a different discussion, covered in the TDS and extraction yield guide. On this page the number stays what the prism gives you.

How does a bent light beam become a TDS percentage?

A refractometer measures the refractive index of a liquid, how sharply light bends as it passes through. The more dissolved solids in the liquid, the higher that index. The instrument never sees coffee as such: it sees an angle, and a calibration function turns that angle into a percentage.

Dedicated coffee units carry a coffee-specific calibration function and correct for sample temperature. That correction earns its keep, because refractive index shifts with heat and a reading taken on coffee still at 60 °C comes out wrong if nothing compensates. Analogue refractometers with an eyepiece read off an engraved scale by eye, with no electronic compensation, which makes them a poor fit for the small differences coffee work is chasing.

Do you multiply or divide a Brix reading to reach TDS?

You multiply. This is the single most common mistake made with these instruments, and it moves the answer by about 15%. The Brix scale was built for sucrose solutions, and the dissolved solids in coffee do not refract light the same way. The empirical relationship the industry has used since the mid-1990s is:

TDS (%) = Brix (%) × 0.85

A Brix reading of 1.50% therefore corresponds to roughly 1.28% TDS. An independent 2019 study recovered a coefficient of 0.85 by linear regression, with real spread inside the range: the fit sits nearer 0.83 at higher concentrations and nearer 0.90 at lower ones.

The practical consequence is short. Anything sold as a coffee refractometer already applies the conversion internally and displays TDS directly. A generic Brix refractometer, the thirty-euro one from a winemaking or gardening supplier, displays raw Brix: you apply the 0.85 factor yourself, and you accept an engraved scale with neither the resolution nor the thermal stability that brewed coffee demands.

Resolution, repeatability, accuracy: which number should you trust?

Three words get used as if they meant the same thing. Resolution is the display step: it tells you how finely the screen rounds, and nothing more. Repeatability describes how tightly several readings of one sample cluster. Accuracy describes the distance from the true value. A device can show 0.01% and still be 0.15% off.

Manufacturer datasheets are unambiguous on this. Here is what the three coffee-calibrated models available through European distributors publish, with prices checked in September 2026 and given as orders of magnitude.

Two things follow. The most expensive unit does not have the finest display step, since all three show 0.01%. And the accuracy gap between the Atago and the other two, ±0.15% against ±0.03%, is wider than the differences you are usually trying to detect between two filter recipes. That row, not the price tag, is the one that should drive a purchase.

Which steps make a reading repeatable?

1. Zero-calibrate with distilled water: Before each session, verify the reading with distilled water gives 0.00%. If not, recalibrate per the manufacturer's instructions.
2. Cool the sample: Even with temperature compensation, readings are most stable between 20-25°C. Draw 1-2 ml of coffee into a small dish and allow it to cool for 2-3 minutes.
3. Mix the sample: Stir before drawing: surface deposits and sediment can skew readings. For espresso, always stir (the crema has a different refractive index than the liquid below).
4. Place 2-3 drops on the prism: Use optical paper or a soft cloth to clean the prism between measurements. Never use abrasive materials.
5. Read and record: Wait for stabilization (2-5 seconds on digital models). Take 2-3 readings and average them if they diverge by more than 0.02%.
6. Log the reading with its conditions: Sample temperature, reading number, dose and beverage weight. A TDS figure without its sampling conditions cannot be compared with tomorrow's.

Why does an espresso sample need different handling?

Coffee-calibrated units cover both, but sampling does not look the same. Espresso is an emulsion loaded with fines and lipids, and those suspended particles scatter light instead of bending it cleanly, which inflates and destabilises the reading. Two standard workarounds exist: push the sample through a syringe filter before it touches the prism, or let it settle, stir, and draw from under the crema. Coffee equipment distributors sell those syringe filters in boxes of fifty for exactly this reason.

For filter methods (V60, Chemex, batch brew), direct measurement is simpler. The main consideration is sampling a representative amount: TDS varies slightly between the first and last drops of an extraction (the start is more concentrated). Sample after the brew has been mixed in a carafe or vessel.

What will the device never tell you?

Worth understanding before you invest. A refractometer measures the quantity of dissolved solids, not their character or their pleasantness. Two coffees from different origins reading identical TDS will taste nothing alike. The same figure can sit under a gorgeous cup or a muddy one, depending on the coffee, the roast and the process.

The refractometer also responds differently to different coffee matrices. A natural-process coffee (high in fruit sugars) may read slightly higher TDS than a washed coffee extracted to the same degree. These differences are small but real, and not corrected for in most consumer devices.

A refractometer doesn't tell you whether your coffee is delicious. It tells you precisely where you are on the extraction map. The destination is still yours to choose.

Which model fits your use and your budget?

For regular home use the DiFluid R2 Extract covers the need: dual temperature sensors, a 0 to 30% TDS range, a body you can rinse under the tap, and roughly €270 including VAT from European retailers in September 2026. The Atago PAL-COFFEE (BX/TDS), around €400 excluding VAT, brings workshop-grade IP65 sealing and a dual Brix and TDS scale, at the cost of a wider published accuracy figure.

The VST LAB Coffee III, around €880 excluding VAT, remains the roastery and laboratory reference: it averages several readings before displaying a result and publishes ±0.03% in the filter range. At that price the argument stops being the printed figure and becomes consistency in a device used dozens of times a day.

One caution applies to all three. A multi-purpose refractometer sold for wine, oils or aquariums is not calibrated for coffee. It reports Brix on a scale built for a different matrix, so you apply the 0.85 factor by hand and live with the imprecision that comes with it.

← Back to guides

Where do most measurement errors actually come from?

Almost never from the instrument. A refractometer reads the refractive index of whatever sits on its prism with the precision its datasheet claims; what varies is the sample and the conditions under which it is read. Operator technique, not brand choice, sets the ceiling on a reading.

Temperature is the most critical variable. Coffee refractometers are calibrated to produce accurate readings at a specific temperature, typically 20 °C (68 °F). Coffee fresh from the brewer is at 70-85 °C; measuring it immediately will produce a significantly erroneous TDS reading because the refractive index of a solution changes with temperature. Most quality coffee refractometers include automatic temperature compensation, an electronic correction applied to the measurement temperature, and manufacturers publish the range over which it operates: the Atago PAL-COFFEE, for instance, states compensation from 10 to 100 °C. The most reliable approach is to cool a small sample (5-10 ml) to room temperature in a sealed container before measuring, or to use a refractometer with a wide ATC range combined with ambient temperature close to 20 °C.

Calibration with distilled water is essential before each measurement session. Distilled water at 20 °C should read exactly 0.000 on the TDS scale; if it doesn't, the calibration offset should be adjusted before coffee measurements are taken. This calibration step takes 30 seconds and prevents systematic errors that would make all subsequent measurements consistently high or low. Tap water should never be used for calibration, its dissolved mineral content will register as a non-zero TDS and create a permanent offset in all readings.

The sample volume placed on the prism must cover the entire prism surface without forming bubbles or creating a film too thin to measure accurately. For most coffee refractometers, 2-3 drops (approximately 0.15 ml) is sufficient. Allowing the sample to equilibrate for 30-60 seconds on the prism before taking the reading, rather than measuring immediately, reduces thermal gradient errors. Clean the prism with distilled water between measurements; coffee residue on the prism creates carry-over contamination that inflates readings of subsequent, weaker samples.

How do you build a measurement log worth keeping?

A single reading is close to worthless. The instrument earns its place when readings accumulate in a form that allows comparison, which means recording the conditions alongside the number rather than the number alone. That habit is what turns a curiosity into a diagnostic instrument.

The most productive application is parameter isolation: changing one brewing variable at a time and measuring the TDS and calculated EY (Extraction Yield) impact of each change. For example: grind the same coffee at three different settings (fine, medium, coarse) at fixed temperature, ratio, and time, and measure TDS for each. The results will show how much of the extraction variation is attributable to grind size alone, information that allows future grind adjustments to be made with confidence that the cause-and-effect relationship is understood rather than guessed.

Tracking TDS across the post-roast window of a specific coffee reveals the degassing impact on extraction. Coffee at day 3 post-roast will typically extract at a lower TDS than the same coffee at day 12, because residual CO₂ creates extraction resistance. Quantifying this difference, rather than relying on subjective taste impression alone, provides calibrated information for recipe adjustments as the coffee ages through its optimal window and beyond.

Comparing TDS across brewing methods with the same coffee and same water reveals the inherent extraction efficiency differences between methods. French press, Chemex, and espresso will all show different TDS and EY values from the same coffee, even with equivalent ratios, because pressure, filter type, and contact time create genuinely different extraction environments. Understanding these baseline differences between methods prevents the error of assuming that a recipe optimised for one method will transfer directly to another.