Chert wrote:I just recall that the probe on top of a puck of coffee made readings that I could not fathom. I seem to remember thinking that the thermometer could not adjust quickly enough as the water moved. If I were willing to ruin a few more thermocouples, I could try again and grind finely enough so that the water would stay above the puck for at least ten seconds.
A thermocouple's speed largely depends on the thermal mass of the junction point; a tiny beat thermocouple will react much faster than one encased in a large probe.
Chert wrote: Both lever groups I have worked with have superheated water behind them (Pavoni and CMA) so that if the water can pass rapidly through the puck, I expect temperatures over 100c to hit the coffee. An open boiler design, as some folks have and describe here, would seem to allow gravity extraction followed by pressurized extraction from the lever without much change in temperature.
Now this is interesting, and it is interesting to think of what happens to the superheated water in a lever group. Is there any reason to think that superheated water is actually making large scale contact with the coffee?
The lever is brought down, the 125c water starts to flow, hits the group and starts cooling (while heating the group). The shower screen keeps the coffee from mixing in any large quantity in the chamber. The water from the boiler fills the chamber until it is full, pressurizing it with boiler pressure. There's nowhere for the air that previously occupied the chamber to go, so it is there too, compressed.
The water will cool as it goes through the group screen under the boiler pressure, starting the preinfusion. It further drops in temperature as it hits the coffee. Then the lever is raised and the rest of the water is forced under spring pressure through the puck.
It might be interesting to bury a thermocouple in the coffee (tamped into the coffee) to see what temperatures it sees during preinfusion. Of course the thermocouple's very presence will alter the water flow at this stage.
I guess a Scace tool might be the best we can model, but it measures the water as it would sit on top of the puck, not in it. I can't imagine a large difference though there might be convection effects in such a configuration which would throw off the reading.
One thing that the lever group has over a pump machine in this regard is that it allows a hot preinfusion followed by a cooler extraction, because the water for the extraction is pulled from the boiler and cools during preinfusion. Trying to do a hot preinfusion with an unflushed pump HX machine would do a hot preinfusion followed by a hot extraction. The temperature of a lever group isn't constant, that is clear - it would be interesting to log the temperature during a shot as well, and graphing the changes during the whole process.