A refractometer measures total dissolved solids as a percentage of the beverage — TDS. Combined with dose and yield on a 0.1 g scale, it yields extraction yield: the percentage of the coffee bed that actually ended up in the cup. Two instruments, one decision: grind finer, grind coarser, or change nothing.

What each number means

Terms, formulas and targets
Quantity Formula Worked example Target Instrument
Brew ratio n = yield ÷ dose 36 ÷ 18 = 1:2.00 1:1.5 – 1:2.5 Scale, 0.1 g
Strength (TDS) measured 10.0 % 8.5 – 11.5 % Refractometer
Extraction yield EY % = (yield × TDS) ÷ dose (36 × 10) ÷ 18 = 20.0 % 18 – 22 % Computed
Flow rate g/s = yield ÷ time 36 ÷ 27 = 1.33 g/s 1.0 – 1.6 g/s Scale + timer

Choosing the instrument

Refractometer options compared
Type Range Resolution Auto temp compensation Cost class Fit for espresso work
Digital coffee refractometer 0 – 30 % TDS 0.01 % Yes $$ – $$$ Standard choice — reads coffee units directly
Analog Brix handheld 10 – 30 °Bx ~0.2 °Bx No $ Workable — convert Brix × 0.85 to approximate TDS
Lab digital (low range) 0 – 10 % 0.01 % Yes $$$ Too narrow — espresso routinely exceeds 10 %
No instrument — — — Free Taste + time only — slower, more overshoot

The five-shot protocol

One bean, one recipe, one variable per shot. The protocol converges on a verified 18–22 % extraction in five pulls or fewer for most grinders with step sizes of one click or less.

Shot sequence — dose 18.0 g, yield 36.0 g locked throughout
Shot Change from previous Expected time Expected EY Decision rule
S1 Baseline — current grind record record Establishes the starting coordinate; no judgement yet
S2 Grind one step finer (if time < 25 s) or coarser (if > 32 s) ±2 – 4 s ±0.8 – 1.5 % If time enters 25–32 s, measure TDS before touching anything
S3 Nothing — repeat S2 settings ±1 s of S2 ±0.5 % of S2 Confirms the reading is real, not sampling noise
S4 Temperature ±1 °C only if EY stuck below 18 % or above 22 % ±0 – 2 s ±0.5 – 1.0 % Solubility lever — do not combine with a grind change
S5 Repeat the winning coordinate window 18 – 22 % Two consecutive in-window shots = dialled in

Reading the result — quadrant table

TDS × extraction yield → next move
TDS Extraction yield Diagnosis Next move
< 8.5 % < 18 % Weak and under-extracted Grind finer — it fixes both numbers at once
< 8.5 % 18 – 22 % Extracted correctly, cup too thin Shorten ratio toward 1:1.8 — strength only
8.5 – 11.5 % < 18 % Strong but under-extracted (sour, salty) Extend ratio toward 1:2.2 or raise temperature 1 °C
8.5 – 11.5 % 18 – 22 % Both windows satisfied Nothing — confirm with a repeat shot, then lock it
8.5 – 11.5 % > 23 % Over-extracted despite good strength Grind coarser, or drop temperature 1 °C
> 12 % 18 – 22 % Heavy, potentially salty finish Raise ratio toward 1:2.4 — dilute, do not grind coarser
> 12 % > 23 % Strong and bitter Grind coarser first, then reassess strength

Error sources, ranked by damage

Measurement errors and their fixes
Error source Typical impact Fix
Unstirred sample ±0.3 – 0.6 % TDS Stir the cup, pipette from the centre, read twice
Reading while hot (no ATC) ±0.2 – 0.4 % TDS Use auto temp compensation, or rest sample 30 s
Dirty prism +0.5 % or more Lint-free wipe with distilled water every session
Bubbles or incomplete cover erratic ±1 % Wait 5 s after closing the flap before reading
Forgot Brix → TDS conversion systematic ~15 % high Multiply Brix by 0.85 before any EY calculation
Scale drift or rounded dose ±0.3 % EY per 0.5 g Weigh to 0.1 g; tare the cup on the same scale
Shortcut: the formula and ratio tables turn dose, yield and TDS into extraction yield, flow rate and a quadrant verdict without guesswork.