Ok, the Big Picture goal is thermal consistency at the grouphead. Well, the Big Picture is perfect espresso, but this step is just thermal consistency. For the non-coffee-geeks here, the grouphead is the big piece of metal (usually chromed brass) that you mount the portafilter into. The portafilter is the heavy doodad with the handle and the little coffee-puck-basket thingie. I'm posting this here, rather than at a coffee forum, because there's a frightfully large number of geniuses who hang out here, and who have huge amounts of knowledge on thermodynamics and heat transfer.
There are two flows of water (and thus heat) to the grouphead:
1 - There is a "natural circulation" of flow from the boiler, maintained at about 120C, to the grouphead. This flow maintains the GH at ~100-105C. Unfortunately, optimal temperature is ~93-95C, so it's a bit too hot.
2 - When a shot is being pulled, water is pumped from a room temperature reservoir through heat exchange coils in the 120C boiler out through a short piece of pipe into the GH. Assuming it's been in there for a moment, the water ought to come out at about 93-95C, ideally.
The conventional policy is to "flush" the system before pulling shots. Run 8-10oz of water through the system to cool the GH and fill the heat exchange tubing inside the boiler with fresh room temperature water. Then wait 30 seconds (experimentally determined) and let the water in the HX coils come up to temp. Then, pull the shot. This is non-optimal, for a variety of obvious reasons.
So here's my thoughts:
1 - Both flows obviously have to be temperature controlled. I don't ever want the grouphead to be more than about 1C off of the target temperature.
a - The shot flow is easy (sort of). Mount a thermocouple near the outlet of the flow path, somewhere close to the grouphead. Use a digital controller to control a pair of fans (like the 12V CPU fans would be about right, I think?) to cool the pipes. If necessary, the piping length could be "extended" by wrapping it into a coil-shape or brazing on cooling fins (ala Elephant Racing oil lines).
b - The continuous heating flow is harder. If I use a fan/heatsink combo to cool the grouphead, I'll just boost the NC flow. Temp shouldn't change (much). So instead, I'd need to reduce the actual heating flow -- say, with a servo-controlled valve. So I'd need a TC on the GH feeding a digital controller inputs on whether or not that valve should be open.
Thoughts? Ideas on where to get parts for this kind of insanity without breaking the bank? Problems with my temperature controls? Simpler ways to accomplish the same thing? Problems with my mental state or addiction?
Thanks in advance, all,
Dan