I've been studying HX performance to better understand the dynamics of the cooling flush amount, pressurestat setting, and design differences (e.g., length of the HX, how much is immersed in water, path length, diameter of the thermosyphon, etc.). Independent of the espresso machine, they all exhibit what I call the "HX hump" at the beginning of the shot, shown in the chart below. At first I tried to hone my technique to produce as small a hump as possible, since the conventional wisdom of recognized experts in the field said "flat is king."

I've noticed two things the operator controls that affect the HX hump, especially on prosumer E61-type espresso machines—how much is flushed, and the length of the interval between flushing the water out of the HX and starting the extraction. The flush amount tends to affect the mid-to-tail end of the curve. The rebound time ("pause before the pull") affects the very early part of the shot, either producing a prominent hump, a flat one, or if there's little or no delay, it disappears completely and the curve becomes more like a rising straight line that barely reaches the desired peak temperature (or worse yet, forms an inverted U as the temperature plummets soon after the extraction begins).

The ideal rebound time for the E61s I've measured is somewhere around 15-35 seconds, depending on the boiler temperature. It is certainly less than one minute since the heat exchanger water is above 212°F at that point for all reasonable pressurestat settings.

The chart to the left shows an example of the prominent HX hump of the Cimbali Junior. Commercial espresso machines like it and the double-boiler La Marzocco produce a profile that is uniquely theirs and there's little you can do as an operator to change its shape—which is a good thing in a commercial environment. I've come to appreciate how each of these temperature profiles favors certain extraction flavors.

Interestingly enough, I also discovered that unlike their bigger commercial brothers, you can "work" the temperature of a prosumer machine to produce different curves that enhance different extraction characteristics. So if a particular blend favors a "La Marzocco" like temperature profile, it's not difficult to emulate it on your home machine. Don't laugh, but I've found that for our house favorite, Black Cat, I prefer the slant-L profile for straight shots and a very "high hump" profile for cappuccinos. The initially higher temperature boosts flavors that are otherwise lost in the background, producing a more flavorful cappuccino with prominent chocolate and caramel overtones. On the other hand, the Toscano espresso of our local roaster, Counter Culture Coffee, is a brighter blend and it benefits noticeably from an HX hump whether it's served straight or in milk.

At first I thought this was only my imagination until Jim Schulman tried the same experiment on his equipment. Jim is a home coffee roaster extraordinaire and master of the experimental method, so you could imagine my delight when he reported these results (quoted with permission):

"I've done about five to six shots with the 45 second pause, and presumably the humped profile, using the Kenya/Ghimbi blend I've been working on for the last six months. The results are definitely more complex than the flat profile shots. I did the blend for a 'fruit bomb' taste— massive blackberries and raspberries, a little orange and caramel —which the flat profile delivers. The humped profile lowered the volume on the fruit, but brings out the clove, balsam and minty notes of the Kenya. As you said before, the hump favors complexity by lowering the volume on the 'loud' bitters and sours, and allowing the softer flavors to be perceptible.

It's not a question of which is better; different blends, tastes and moods will need different treatments. But it's nice to have this as a variable—an aspect of the shot that can be 'worked.'"

I really appreciate his confirmation and also his descriptive explanation of the effects. As for me, I'm not sure what balsam is (isn't that used on baseball bats for a good grip?) and yet this guy can taste it in an espresso. Cool!

Conclusion

The purpose of this article is to share our experimentation and encourage speculation. I don't claim to have all the answers. I continue to learn by experimenting, sharing, and listening to the ideas of other enthusiasts. It's an intriguing aspect of my espresso hobby that keeps me interested, and I hope you will find it interesting too. Finally, keep in mind that to a large extent, what I've outlined in this last section borders on temperature accuracy for sport. Getting within a degree or two is more than accurate enough for my taste abilities, which isn't difficult to do consistently just by observing the water dance during the cooling flush.

If you have questions or wish to join the discussion of this article, see the forums. The next how-to in this series, Perfecting the Naked Extraction, continues by explaining the equally important dosing, distribution, and tamping skills the barista must master.