Loading…
Set aside the chrome and the ritual and an espresso machine is a small, exact instrument: it holds water a few degrees below boiling, leans on it with the weight of nine atmospheres, and pushes it through a thimble of ground coffee before the pressure has time to settle. Everything interesting happens in about twenty-five seconds.
An espresso machine drives hot water through a packed bed of coffee under pressure — about nine bars of it, water a few degrees off the boil, seven grams of grounds, done before you’ve found a saucer.
Roughly 7–9 g of coffee, water near 92–95 °C, 20–30 s, yielding 25–35 mL. In a real café, heavier — 18 g in, 36 g out.
Follow a single shot from tank to cup. Three things have to line up along the way — pressure has to be made, heat has to be held, and the coffee itself has to push back just the right amount. Step through the parts to see where each one lives.
Every shot begins here, at room temperature. From this tank the pump draws the few grams of water a single espresso needs.
“Nine bars for twenty-five seconds” sounds like a setting you dial in and hold. It isn’t. A shot is a short journey through three curves at once — pressure, flow, and temperature. Scrub the cursor and watch how the famous window is really a plateau the machine passes through, with a wetting phase before it and a decline after.
The nine-bar window. Full pressure. This plateau is the part the textbooks mean by “9 bar for 25 seconds.”
Illustrative reference profile — shaped to documented behaviour, not a log from one machine.
Here is the awkward fact a machine has to design around: brewing wants water around ninety-three degrees, and steaming milk wants something well past boiling. One vessel cannot be both at once. The three classic layouts are three different ways out of that corner.
One boiler heats brew water, then is cranked up for steam and cooled down again. Cheapest and most frugal, but you cannot brew and steam at once, and without a PID you time the shot against the thermostat's cycle.
The industry measures a shot by its extraction yield: the share of the dry coffee’s mass that actually ends up dissolved in the cup. Most of a roasted bean is insoluble structure, so there is a hard ceiling — only about a third of it will ever leave.
Grind finer and you expose more surface, which ought to raise the yield. A mathematical model of the puck predicts exactly that — so long as water flows evenly. But when Cameron, Hendon and their colleagues actually measured it, yield rose, peaked, and then fell as the grind got finer still.
The culprit is channeling. Pack the bed too fine and water stops soaking through evenly; instead it carves a few preferential paths, over-extracting along them and barely touching the rest. The cup comes out both bitter and sour at once, and no two shots agree.
Their fix runs against café instinct: find a shot you like, then grind coarser and use less coffee, steering the yield with the ratio of water to grounds rather than the clock. In a year-long café trial it cut the coffee in each drink by as much as a quarter, with no loss in the cup.
A smaller heresy from the same work: tamping force, within the range anyone actually uses, barely changes the shot. The tamp is there to level the bed, not to add resistance. And the lower the pressure, the wider the range of grinds that still flow evenly — their trials dropped to six bars to keep fine beds from clogging. Nine bars endures as the sensory standard, not a yield optimum.
of the bean is all that will ever dissolve — the rest is structure.
less coffee per shot, at the same yield, by grinding coarser.
That reddish-brown foam is real, and it is more interesting than the folklore around it. Roasting traps carbon dioxide inside the bean; the shot’s pressure forces that gas into solution, and when the coffee hits the open cup and the pressure drops, it fizzes back out as a haze of micro-bubbles suspended in an oil-in-water emulsion — 90% of the oil droplets under 10 µm, threaded with fragments of coffee too small to see.
What keeps it standing is chemistry borrowed from roasting: proteins and their browned descendants, the melanoidins, build the foam; dissolved sugars called polysaccharides make it last; oils firm up the walls. Drip coffee has the same ingredients and never foams, because it never sees the pressure.
The connoisseur’s rule of thumb — at least 10% of the cup, holding for two minutes or more — is a fine freshness cue, because stale, off-gassed beans foam less. But it is a poor flavour cue: low-regarded robusta foams more than prized arabica. Crema tells you there was pressure and the beans were fresh. It does not tell you the coffee is good.
None of which is an argument against the ritual. The tamp, the timing, the watching — they are part of the pleasure of the thing. The point is only to know which parts are the machine doing physics and which parts are us, enjoying ourselves.
Crema tracks freshness and trapped CO₂, not flavour. Stale beans foam less; robusta foams more than prized arabica.
Within the range tested, tamp force barely moved yield or shot time. Its real job is to level the bed.
Past a point, fine beds channel: yield peaks, then falls, as water carves preferential paths.
Time is an outcome, not a dial. Steer by coffee mass and water mass; some good shots finish in under fifteen seconds.