Espresso Extraction Troubleshooting Chart: Symptom, Cause, Fix

Every espresso fault is a deviation from one reference recipe: 18 g in, 36 g out, 25 to 30 seconds, at roughly 9 bar and 90 to 96 C. Fix that recipe in place, then read the symptom off the chart below and make the single change it names. The column that matters most is the last one, what not to change, because the usual reason a shot cannot be recovered is that three variables moved at once and there is no longer a way back to a known state.

What is the espresso troubleshooting chart?

Read down the first column until you find what your shot is doing, then apply the fix in the fourth column and pull one more shot. One change, one shot, then taste again. This takes four or five shots and roughly 90 g of coffee, which is the real cost of dialling in a new bag.

What are the target numbers a shot is measured against?

Extraction yield is the percentage of the dry coffee that dissolved into the drink, and the conventional target window is 18 to 22%. Below that the shot reads sour and thin, above it bitter and drying. Almost nobody at home owns a refractometer, so taste is the proxy, and the numbers below are the inputs you can actually control and measure.

Note the tolerance on dose. A 1 g error on an 18 g dose is a 5.5% change to the recipe before the grinder has done anything, and it is invisible unless you weigh. This is why a scale such as the Maestri House S2 Espresso Coffee Scale with Timer belongs in the first purchase rather than the fifth, and why volumetric buttons alone are not enough: they meter water, not coffee.

In what order should you change variables?

Diagnosis fails when people start at the expensive end. The order below runs from the variable with the largest effect per unit of effort to the one with the smallest, and it is also cheapest to most expensive, which is not a coincidence.

The order matters because the faults interact. A channelled puck produces a fast shot, so it looks exactly like a coarse grind. If you respond by grinding finer, the next shot channels harder, because a finer bed has more resistance and the water pushes even more forcefully through the existing weak point. Two or three iterations later the grinder is far off where it should be and the shot is worse. Fix distribution first and the grind question becomes answerable.

How do you tell channelling apart from a grind problem?

Channelling is uneven flow through the puck, not too much or too little flow. The tell is unevenness in space rather than in time. A grind problem produces a shot that is uniformly too fast or too slow. A channel produces a shot that starts normally then suddenly blondes, or that pours heavily from one spout and drips from the other.

The definitive diagnostic is a bottomless portafilter, because it removes the spouts that hide what the basket is doing. With a bottomless portafilter you watch the underside of the basket directly: an even shot forms droplets across the whole surface that converge into a single stream, while a channelled shot fires visible jets sideways within the first ten seconds. It is a measuring instrument rather than a luxury, and it is usually the cheapest way to stop guessing.

The fix chain for channelling is short and ordered. First, break up clumps with a needle distribution tool: the WDT tool uses fine needles to stir the grounds before tamping, which is the single highest impact accessory in espresso. Second, keep grounds inside the basket with a magnetic dosing funnel so the rim stays clean and the puck seals against the basket wall. Third, tamp level once, without twisting or tapping. The puck prep guide covers the whole sequence in order.

When is the problem the water or the machine rather than the coffee?

Three faults are not extraction faults at all and will not respond to grind changes.

Scale. A machine that has lost flow rate over months, takes longer to reach temperature, or makes new gurgling noises is scaling internally rather than developing a grind problem. Espresso machines scale from calcium and magnesium, and the target water is roughly 50 to 100 ppm total hardness as calcium carbonate with alkalinity near 40 ppm. The water hardness chart gives the descale intervals by band and the conversion between ppm and grains per gallon. Descaling chemicals are not food, so follow the manufacturer's rinse instructions exactly.

Old coffee oils. A group head that has never been backflushed will make every shot taste bitter regardless of the recipe, because rancid oils behind the shower screen dissolve into the water on its way through. A weekly backflush with Cafiza detergent on any machine with a three-way solenoid valve, plus a wipe of the screen with a group brush , removes a fault that otherwise gets blamed on the beans for years. Machines without a three-way valve should be cleaned per the manufacturer's instructions instead, since blind backflushing is not appropriate for all of them. The maintenance guide covers which routine belongs to which machine.

Temperature instability. On a machine without PID control, brew temperature can swing several degrees depending on where in the heating cycle you happen to press the button. That turns identical recipes into different shots, and it is the main reason single boiler machines reward a fixed routine: flush, wait, pull, always in the same rhythm. If your shots are reproducible in time but not in taste, temperature is the likely variable.

What does a bad grinder look like in the cup?

A grinder that cannot hold a setting produces the eighth row of the chart: the same recipe tasting different every time. The mechanisms are specific rather than mysterious. Retention means several grams of yesterday's grind sit in the chute and mix into today's dose, so an adjustment does not fully take effect until two or three doses later. Wide particle distribution means an excess of fines that extract instantly and boulders that barely extract at all, producing a shot that is simultaneously bitter and sour. Stepped adjustment sized for filter brewing means the gap between one setting and the next is larger than the entire espresso window.

The Baratza Encore ESP Coffee Grinder is the usual entry point out of that situation, and the MiiCoffee DF64 II adds stepless adjustment and near-zero retention, which is what makes single dosing different beans practical. Before spending, though, measure: a week of recorded shots on the grinder you own will tell you whether the variance is coming from the grinder or from the dose. The grind size chart shows why the espresso window is so narrow, and the grinder roundup covers what each price band actually buys.

What should a dialling-in session look like?

Set the dose to your basket and leave it. Pull the first shot at whatever grind the previous coffee used, and record the time to reach 36 g. If it is under 20 seconds, move two or three steps finer. If it is over 35 seconds, move two or three steps coarser. Pull again. Once the time lands between 25 and 30 seconds, taste. Sour means one more step finer, bitter means one step coarser. Most coffees land within four shots.

Then stop. The most common failure at this point is continuing to tune past the point of improvement, which mostly generates evidence that you cannot taste the difference between adjacent steps. Write down the dose, yield, time and grind setting, and drink the coffee. When the bag is two weeks older and the shots start running fast, you will have a reference point to return to rather than a memory. The dial-in guide covers the same sequence at length, including what to do when the bag ages and the shots start running fast at a setting that worked a fortnight ago.