Refractometers and measuring extraction

A refractometer is the tool that turns brewing from a taste argument into a measurement. It does not tell you whether your coffee is good. It tells you, repeatably, how much dissolved coffee is in the cup and what fraction of the grounds gave it up. Once you can put numbers on those two things, you can change one variable, measure, and know exactly what moved. This lesson assumes you already understand the difference between strength and yield. If not, read extraction-theory first, because a refractometer is useless until you know what the readings mean.

What the device actually measures

A coffee refractometer measures total dissolved solids (TDS), the percentage by mass of dissolved coffee solids in the liquid. It does this optically, not chemically. Dissolved solids change how much light bends (refracts) when it passes through the liquid, so the meter reads the refractive index of a small sample and converts it to a TDS percentage on a coffee-calibrated scale. A digital model such as the common Atago-style unit reports to 0.01 percent.

A few practical truths fall straight out of that mechanism:

• Temperature matters. Refractive index shifts with temperature, so good meters compensate automatically, but they need a moment to equilibrate. Sample the same way every time. Many people pull a few milliliters, let it cool toward room temperature, then read.
• Particles are noise. The meter wants dissolved solids, not suspended fines. An unfiltered immersion sample full of fines and oils reads high and erratic. For filter coffee, pull the sample through a syringe filter (0.45 micron is common) or at least let it settle and draw clear liquid off the top.
• Espresso needs dilution. Raw espresso is too concentrated and too full of emulsified oil and crema to read cleanly. The convention is to weigh a small mass of well-stirred shot, add a known mass of room-temperature water (often 1:4 or 1:9 by weight), read that, then multiply the TDS back up by the dilution factor.
• Calibrate with distilled water. It should read 0.00. Re-zero whenever the room temperature changes much.

Typical numbers to anchor yourself: filter coffee lands around 1.15 to 1.45 percent TDS, cold brew concentrate runs far higher before dilution, and undiluted espresso sits roughly 8 to 12 percent.

From TDS to extraction yield

TDS alone is just strength. The number you actually steer with is extraction yield (EY), the fraction of your dry dose that ended up dissolved in the cup. You get it with one piece of arithmetic, and the only extra data you need is the weight of liquid you actually brewed (the beverage mass), not the water you poured in. Those differ because the spent grounds keep a chunk of water: a wet puck or a soaked filter bed retains roughly 2 grams of water per gram of coffee.

The formula:

EY % = (beverage mass in grams x TDS %) / dose in grams

Worked example. You brew with a 15 g dose, and after the drawdown your carafe holds 240 g of liquid. The refractometer reads 1.38 percent.

EY = (240 x 1.38) / 15 = 22.08 percent.

That sits right at the top of the conventional 18 to 22 percent filter window, so this cup is fully, maybe slightly over, extracted. If it also tasted a touch drying, the number agrees with your tongue, and now you know which way to move.

For espresso the same formula applies, but plug in the diluted TDS multiplied back by your dilution factor, the shot’s beverage mass (the yield in the cup), and the dry dose. Espresso EY commonly lands around 18 to 22 percent as well, though many modern recipes happily run higher.

Using the number to dial recipes

Here is the workflow that makes a refractometer worth the money. Pick a target EY (say 20 to 21 percent for filter), brew, measure, then adjust one of the four dials and re-measure:

• EY too low, cup tastes thin or sour: under-extracted. Grind finer, raise water temperature, extend contact time, or add agitation. Each pushes more solids out.
• EY too high, cup tastes bitter or astringent: over-extracted. Grind coarser, drop temperature, shorten the brew, or pour more gently.
• EY fine but cup too weak or too strong: that is a ratio problem, not an extraction problem. Change the coffee-to-water ratio and leave grind alone.

That last bullet is the real payoff. Without measurement, people chase a weak cup by grinding finer, accidentally overshoot extraction, and end up with a stronger but bitter brew. The refractometer separates the two axes so you fix the right one. For pulling shots specifically, see espresso-dialing, where EY plus shot time plus taste form a tight feedback loop.

What it cannot do

A refractometer measures concentration, not deliciousness. Two cups at an identical 20 percent EY can taste worlds apart depending on grind distribution, evenness, water chemistry, and bean quality. A reading cannot detect channeling or bypass: an uneven brew can post a “perfect” average EY while half the bed is over-extracted and half is under, and only your palate catches that. The 18 to 22 band is a convention drawn from cupping panels, not a verdict, and plenty of excellent light-roast filter sits at 21 to 23 percent. Sampling sloppiness (hot samples, unfiltered fines, a dirty prism) produces numbers that look authoritative and are simply wrong.

Treat the meter as a way to make your tasting reproducible, not as a substitute for it. Brew, taste, write down the EY, and over a few weeks you build a personal map of which numbers your palate actually likes for each coffee.

Takeaway

Measure TDS, divide by dose using beverage mass to get EY, target 18 to 22 percent for filter, then move one dial at a time and re-measure. The refractometer’s job is not to grade your coffee. It is to make every adjustment honest, so you stop guessing which way you moved. Next: take the workflow to the bar in espresso-dialing.