Over the past few years, we have tested over 90 espresso grinders – from entry-level machines at €150 to professional single-dosing grinders beyond €1,000. What emerged is not a gut feeling, but a structured measurement and test protocol that we have continuously refined with each test.
This article explains how we proceed, why we measure exactly these values, and what lies behind our rating thresholds. It also serves as a reference for all our individual grinder reports – we link to the corresponding sections here.
The test protocol at a glance
Our test is divided into four blocks, which are carried out at different times. This ensures that we capture both the cold start of a grinder and its behavior after break-in and under load.
All grinders are tested with the same coffee: the Apas Espresso, a naturally processed Brazilian from our partner cooperative APAS. Same coffee, same bean batch – this creates comparability across all grinders.
Block 1 – Break-in and initial testing: Accessory check, 1 kg coffee ground, first recipe setting, first particle distribution samples.
Block 2 – Main measurements: Temperature, noise level, grinding speed, consistency check, further particle samples with and without RDT (Ross Droplet Technique).
Block 3 – Adjustment testing: Ristretto, lungo, return to espresso – how does the grinder behave with major grind size changes?
Block 4 – Retention: Absolute, temporary, and permanent retention are measured precisely.
Tastings take place in parallel with the respective blocks.
Starting material and break-in
Before we take a single measurement, we grind 1 kg of coffee through the grinder. This simulates realistic use and ensures that flour and oils have settled in the grinder. A freshly unboxed grinder behaves differently than one that has 100 shots behind it – and we test the latter.
Then we check the included accessories: dosing cup, funnel, brush, key. We note what is included. We also note what is missing.
For the In-Out test, we grind three times 18 g of coffee for 30 seconds each and measure the output amount. This gives us a first assessment of retention and consistency – before any other measurements.
Noise level
We measure noise level with a calibrated sound level meter, positioned 20 cm from the grinder. Measurement is taken during continuous grinding of 250 g of coffee – not at cold start, but during steady-state operation.
| Rating | Decibels |
|---|---|
| Top / quiet | under 80 dB |
| Medium / rather loud | 80–90 dB |
| Negative / loud | over 90 dB |
Why these thresholds? 80 dB corresponds to roughly the noise of a vacuum cleaner at normal distance – this is the point at which a grinder becomes perceived as disruptive in a quiet kitchen. Over 90 dB, i.e., lawn mower level, is simply too loud for home use. Under 80 dB, however – such as the Mazzer Philos with 78.1 dB – a grinder is acceptable even early in the morning or in an open kitchen.
Important:
The grinding sound has two dimensions. The mere decibel number says little about the quality of the sound. A penetrating, stuttering whine at 73 dB (as with the Varia VS3) can subjectively be more bothersome than even humming at 78 dB. We document such qualitative impressions in the notes.
Grinding temperature
Heat is the invisible enemy of good extraction. Coffee ground too hot loses volatile aromatic compounds before extraction – and fats begin to oxidize. We measure the temperature of the ground coffee immediately after grinding, five times in a row with 18 g each and 20 seconds pause in between to make the warming effect visible.
| Rating | Temperature of ground coffee |
|---|---|
| Top | under 30 °C |
| Good | 30.1–35 °C |
| Medium / warm | 35.1–39 °C |
| Negative / very warm | over 39 °C |
Why these thresholds? Under 30 °C, the ground coffee is still near room temperature – no measurable thermal stress. From 35 °C onwards, we occasionally taste differences in tastings, especially with light, aromatically delicate roasts. Over 39 °C, as we measured with the Zuriga G2 at 40.2 °C, is a genuine limitation: It disqualifies the grinder for higher throughputs because the temperature continues to rise with multiple consecutive pulls.
We are aware that this measurement is only an approximation of the temperature in the grinding chamber. The temperature during grinding is transferred to the surrounding material, and the ejected coffee is always cooler than the coffee in the grinding chamber. But this type of measurement is the most precise available to us with our means.
Grinding speed
Speed is measured in two different ways, which reveal different things and are not directly proportionally comparable.
Measurement A – Grams in 10 seconds (full hopper): The grinder runs with a full bean container. Measured is how many grams are output in 10 seconds. This measurement is particularly relevant for grinders designed for hopper operation – it shows the maximum performance potential and helps assess whether a grinder could also be suitable for commercial use.
| Rating | Grams in 10 seconds |
|---|---|
| Top | over 40 g |
| Good | 30.1–40 g |
| Medium | 20.1–30 g |
| Slow | 10–20 g |
| Very slow | under 10 g |
Measurement B – Time for 18 g: For hopper grinders, this measurement is also performed with a full hopper – the value is therefore in direct proportion to the 10-second measurement. For single-dosing grinders, it is measured with an empty hopper. This also includes the time for additional puffs with the bellows needed to completely expel the coffee. The total duration is thus typically longer and not directly comparable to hopper values.
| Rating | Time for 18 g |
|---|---|
| Top | 1–6 seconds |
| Good | 6.1–10 seconds |
| Medium | 10.1–15 seconds |
| Slow | 15.1–20 seconds |
| Very slow | over 20 seconds |
Why is speed relevant? Very fast grinders grind with high power, which increases grinding temperature. Very slow grinders extend the contact time of the coffee with the burr – this doesn't have to be a weakness, but it changes the workflow noticeably. Our investigations of rotational speed also show that grinders with lower RPM tend to have a narrower main peak; this means they grind more uniformly.
Grinding speed is an important criterion for use in gastronomy or catering operations. A speed of less than 10 seconds for 18 grams is desirable with frequency.
Retention
Retention is the area in a grinder where already ground coffee remains. This coffee is not immediately ejected, but rather during the next grinding – it is always a portion older than the freshly ground coffee. The larger the retention, the stronger the flavor consequences: rancid, sharp, unbalanced. We describe this in detail in our article on retention.
We distinguish three types: Temporary retention empties and fills with each grinding – it is the dominant factor for everyday quality. Permanent retention sticks permanently in cracks, around the grinding discs or behind the flapper and affects the aromatic climate in the grinder. Absolute retention is the sum of both.
For measurement, we remove all removable parts, brush out the grinder completely, and weigh the residue.
| Rating | Absolute retention |
|---|---|
| Excellent | 0–2 g |
| Good / low | 2.1–3.5 g |
| Medium | 3.6–5 g |
| High | 5.1–9 g |
| Very high | over 9 g |
Why these thresholds? At 18 g portion size, 9 g retention corresponds to almost half the espresso – half fresh, half old. You can taste that. Below 2 g, the influence on the result is minimal as long as you grind regularly. Grinders in this range – such as the Niche Zero (1.1 g) or the Varia VS3 (1.1 g) – are particularly well suited for single dosing and switching between different coffee varieties.
Retention is also relevant when setting the grind size! The larger the temporary retention, the more influence the "old" coffee from the last grinding has on the brewing behavior of the next espresso. In other words: if a grinder has 9 grams of retention, 9 grams of the old grind size remain in the grinder after a grind size adjustment. It is essential to grind these out in order to see the effect of the grind size setting in the extraction. If this is not done, finding the right grind size becomes a cycle of incorrect adjustment.
Additionally, we measure single-dosing performance: How much coffee is left over when exactly 18 g is input and the grinder runs for 30 seconds?
| Rating | Deviation |
|---|---|
| Top | 0–0.15 g |
| Good | 0.16–0.3 g |
| Medium | 0.31–0.6 g |
| Poor | 0.61–1 g |
| Very poor | over 1 g |
Adjustment testing: How reliable is the grinder?
A grinder that cannot find its exact starting point after a major grind size adjustment is a real problem for anyone using multiple recipes or coffee varieties. We test this with a standardized sequence:
- Espresso (starting recipe): 18 g in, 45 g out, ~25 seconds
- Ristretto: 18 g in, 18 g out, 25 seconds
- Lungo / Café Crème: 18 g in, 120 g out, 25 seconds
- Return to espresso: 18 g in, 45 g out, 25 seconds
From each of these points, we take a 25 g sample for particle distribution measurement. This shows us not only whether the grinder returns to its setting, but also how the ground coffee physically changes at very fine and very coarse settings. The deviation of the x50 value between T4 (espresso) and T7 (espresso after return) is a precise indicator of grind size reproducibility.
Particle distribution
This is the core of our scientific analysis, and this is where our protocol differs from most other grinder comparisons. We analyze all samples in collaboration with the Coffee Excellence Center of the ZHAW (Zurich University of Teacher Education) using a Retsch Camsizer X2 – a high-precision device for dynamic image analysis.
All samples are divided using a sample splitter to prevent separation between fine and coarse particles. Up to seven samples per grinder are measured: factory condition, espresso, ristretto, lungo, return espresso – and sometimes further investigations, such as the effects of RPM adjustments or slow feeding experiments.
x50 (median particle size): The point at which 50% of the ground coffee is smaller and 50% larger. For espresso it typically lies between 150 and 350 µm, depending on grinder and setting. It describes the "average grind size" and shifts finer or coarser with adjustment. Alone, it is only of limited significance – it describes the middle, not the edges.
Fines content (Qf <100 µm): The percentage of particles smaller than 100 µm – the actual coffee dust content. A high fines content increases body and density of an espresso, but carries the risk of over-extraction and bitterness. It also forces grinding coarser to get a flowing shot at all. Grinders with few fines can be set much finer and often deliver clearer, more nuanced flavor profiles.
60% Coarse Peak Width (main peak width): The size range in which 60% of coarse particles (over 100 µm) are contained. A narrow distribution means most particles are similar in size – even extraction, balanced flavor. A wide distribution means heterogeneity: While some particles are already over-extracting, others lag far behind. In our tests we have observed: Once peak width exceeds ~300 µm, espressos often taste "scattered" – acidity and bitterness stand disconnected next to each other, sweetness is missing. We describe this in detail in the article on particle distribution.
Consistency check
A good grinder does not grind precisely only on the first shot – it does so on the fifteenth as well. For consistency measurement, we grind 15 × 18 g with 30-second intervals and measure each output amount to 0.1 g precision. From this we calculate mean and standard deviation.
Standard deviations around 0.1 g are solid. Below 0.05 g is excellent. Over 0.2 g begins to disrupt the workflow – if you dose to 0.1 g precision, you need a grinder that can maintain that.
For grinders with grind-by-weight function (such as the Eureka Libra 65 AP or Zuriga G2 GbW), we additionally test how precisely the scale works and how it handles grind size changes.
Espresso tasting protocol
All measurement values in the world cannot replace the decisive test: How does the espresso taste? We taste in parallel with each test phase and conduct tastings where at the end of a larger test series all tested grinders are evaluated side by side.
Our tasting protocol follows the scoring sheet of the World Barista Championship, but we have extended the scale to 10 points and divided the attributes more differentiated. We evaluate two dimensions:
- Taste balance encompasses sweetness (balancing, showing extraction evenness), acidity (clarity, transparency, liveliness – not sharpness) and bitterness (integrated and positive vs. harsh and disruptive).
- Tactile experience encompasses weight (heaviness in the mouth, independent of quality), texture (creamy, oily, silky, dusty – the quality of the weight) and aftertaste (length and quality after swallowing).
Each rating is supplemented by a free description. More on this in the espresso tasting protocol.
RPM setting (rotational speed)
Not all grinders offer variable speed control – when available, it is treated as a separate data point in the test. The rotational speed of the grinding discs affects the ratio of fine dust to coarse particles without the grind size being adjusted.
At higher RPM, beans are shattered with more kinetic energy, producing more splinter formation and thus more fines. At lower RPM, the grinding is more controlled. The consequences in the cup are noticeable: Low RPM delivers a narrower main peak in the particle curve, clearer aromas and more transparency – but less body. High RPM generates more fine dust, more body and denser texture, but also increases the risk of uneven extraction.
The RPM setting is thus a tool for flavor control that works independently of grind size – particularly valuable for single-dosing grinders like the DF64V, Lagom P64, or Varia VS6. It can help calm a frayed extraction profile by lowering RPM, or create a body-emphasized espresso by raising RPM.
What the protocol does not rate
A good test protocol also defines its limits.
Design and haptics do not factor into the measurement values. A beautifully crafted grinder gets no bonus points from us for that – but if you spend €1,000, you can expect no plastic parts to rattle, and we note that in the comments.
Long-term behavior cannot be captured in our test framework. Our old Mazzer grinders have been running for 15+ years – that is experiential knowledge, not a test result.
Filter coffee we only test where a grinder is explicitly marketed as all-purpose. Some grinders, like the Eureka Libra 65 AP, pleasantly surprise us here.
All espresso grinders in the test
You can find all results at a glance in our large espresso grinder comparison.
