beans equipment order espresso links email newsletter articles about

The pre-Christmas rush provided me with a lot of hard work and a good deal of food for thought. The background to this article is simple; in early November I received a long awaited shipment of Rancilio Silvia espresso machines at the same time that I ran out of stock of Quaha machines. In the following 6 weeks I set up and tested over 80 Silvias, in batches of 6 at a time.

When you need a precise grind, only the best will do.

This sort of thing takes a lot of standardization if you’re going to do it in reasonable time, and after about the 20th machine I began to ask myself where the variations in my shots were coming from. The grind was consistent, done on a high quality commercial grinder. The blend of coffee was consistent. The room temperature and humidity were consistent, and I was pulling each shot at the same point in the boiler heating cycle. But I was getting up to 5 seconds variation in 60ml shot pull times.

Worse, the variations weren't consistent, a second shot on the same machine might not be the same as the first. So I started to pay attention to precisely what I was doing, and tracked down some of the little things that make big differences. In the photos below I’ve let all the shots run for exactly 25 seconds to illustrate what happens.


All the components used in brewing, machine, portafilter and filter baskets, need to be at a stable temperature. The best way to achieve this is to let everything heat up together for at least an hour. Leaving the machine on overnight or longer is not a problem as long as the boiler is full and there are no leaks from the steam wand, solenoid etc.

Machine warmed up 10 minutes, portafilter & basket cold.

Machine, portafilter & basket warmed overnight.

Shot with 13.5g of coffee, double basketNote the differences in coffee volume and crema colour. The "warm" shot is on the right.


A little bit more or less coffee makes a huge difference. Once I’d worked out the temperature thing, this turned out to be the biggest single source of variation. As little as 0.5 gram of coffee can make the difference between an ordinary shot and espresso nirvana.

Shot with 13.5g of coffee, double basket

15 grams, double basket

13.5, 14.0 and 15.0 gram shots side by side, left to right.

Too much coffee?

A very well defined screw impression in the puck.

The absolute best shots were those I pulled when the double filter basket had as much coffee in it as possible, 15.0g. The coffee itself was in direct contact with the showerscreen, and another 0.5 grams caused leaking around the edges of  the portafilter. As it was, the screw impression left in the puck was clearly defined.

Tamp not a factor

One thing that putting as much coffee in as possible did was to remove the effects of tamping as a factor in the shot quality. A light “smoothing” tamp to remove voids and ensure good edge contact with the filter basket was just as effective (or not) as the heaviest “grunt” tamp I could manage. To prove to myself that this situation wasn’t exclusive to the Silvia, I tried the same thing out on a 1-group La Marzocco and got the same results.

Possible downsides

Leaving your machine on for extended periods more or less duplicates commercial practice, but certainly uses more electricity and will lead to faster deterioration of the group gasket. Using as much coffee as possible in the filter basket may lead to wear or damage of the showerscreen assembly (although I haven’t seen any) and certainly requires a strict showerscreen cleaning regime.