Coffee Data Science
Understanding the stack
I have done a lot of tests on the grouphead for the Decent Espresso machine, and I have been taking a closer look at water distribution in general. So I wanted to understand how the shower screen, water diffuser, and water dispenser operated in how water went into the puck. Of course some of these parts are to facilitate a clean grouphead, but I went after a few questions:
- Does the shower screen help extraction?
- Does the diffuser help extraction?
I performed this test with 6 month old coffee, which means the coffee didn’t have any CO2 left or very little. I then did salami shots and measured TDS.
Total Dissolved Solids (TDS) is measured using a refractometer, and this number combined with the output weight of the shot and the input weight of the coffee is used to determine the percentage of coffee extracted into the cup, called Extraction Yield (EY).
For a shot profile, I did a flat profile at 90C at a flow of 2 ml/s. The end of this profile had a long slow ramp down to try to avoid sucking coffee grounds up into the grouphead.
For dose, I had to increase the dose to reduce the headspace. I also didn’t know how far to go, so I modified tamping. I dosed, distributed, and tamped half the grounds first, then I dosed and distributed the second half. I didn’t tamp the second half.
The lack of a tamp allowed there to be little headspace without knowing the exact minimum. So one could study this a bit more and vary some parameters, but I did not. I wanted to gain some data guided intuition.
I was a little surprised the shot looked so normal without a shower screen.
This shot looks pretty typical.
So then I removed the diffuser, and we went for broke.
This shot also looked normal. I was expecting major channeling from the side, but once it pressurized, it was the same.
In terms of measurements, there weren’t major differences in extraction yield.
