In early May, Robert Pavlis asked if I would try a device for measuring temperature inside the piston of a La Pavoni Europiccola. I thought that would be fun and soon received this custom piston rod fabricated on his lathe. He included a lever pin drilled to allow passage of a temperature probe and a tiny set screw to maintain the alignment of the lever pin.



He only had a long set screw, and I couldn't easily find a short replacement. Then I realized the set screw wouldn't be necessary if I could maintain lever pin alignment with something rigid, so I obtained a brass tube in a hardware store narrow enough to pass through that opening and inserted it through the piston rod and lever pin.



To measure temperature differences between the outside of the group and the piston, I used thin wire thermocouples connected to an Amprobe TMD-56 USB thermometer. This would allow me to record my readings in Artisan roasting software. On the first try water began leaking through the hollow piston rod. I realized I hadn't fully tightened the piston rod into the piston and added some Loxeal food-safe thread sealer for insurance. It worked.



I shortened the brass tube sufficiently to fit underneath a drilled acorn nut. Here's a close-up.



Robert's thinking was correct, of course. Brass is an excellent conductor and quickly responds to changes inside the group. While starting to pull shots I employed my usual thermometer with a probe attached to the outside of the group. I could see that the outside group temperature rose far higher than the piston temperature. It's for this reason I've been using outside group temperature only as a starting point. But now, using the piston temperature, my initial shots seem even better tuned. Here's a close-up of the Artisan readout. The shot looks rather long because I'm ramping up group temperature with half pumps before introducing water to the coffee cake. You can see that temperature outside the group reaches about 210F while remaining at about 205F at the piston, which is perfectly tuned to the Compass Coffee La Berlina Panama single origin I was brewing. The blue line shows temperature in the piston, red is the outside of the group, near the back.



My next temperature readouts will compare piston temperature with temperature in the coffee cake.

The piston rod thermowell will provide a much cleaner installation of a temperature probe than previous thermometer adaptations. I've drilled the acorn nut to accommodate the shaft of the Taylor model 608 thermometer I'm using on the group already. It can be fastened to the acorn nut with JBWeld, aligned and then held in place by the locking nut underneath. A piston rod thermowell is also a more universally application, since you don't need to be concerned with how to attach the thermometer to the machine or the group. For instance this will work with a first-generation La Pavoni Europiccola, an Olympia Express Cremina or any lever machine where the top of the piston rod is exposed throughout the pull.

Neat!

This is fantastic! This may be the only practical way to measure the water temperature inside a lever group during a shot.

You might consider shifting from a brass tube to a stainless steel tube or even a teflon tube to give thermal isolation between the piston who's temperature you are measuring and the piston rod/tube. With brass being such a good heat conductor the brass tube will pull down the reading of the temperature a bit. This may not be significant but I've seen it make a substantial difference when making measurements on the outside of a group head or boiler. Also, you might want to put some heat-conductive grease where the tip of the thermocouple sits to get the best heat transfer to the thermocouple. I was surprised at how big a difference these two made.

If you're just using the temperature to maintain consistence between shots and aren't trying to get an accurate temperature reading of the piston/water then this wouldn't be worth addressing.

Since the thermocouple tip is all the way down and inside the piston, I'm not concerned about the brass tube being a heat sink. If anything that's another advantage of it and the brass piston rod, for that matter. Also using brass for the piston rod creates additional heat sink properties. It was fashioned that way because it's easier to machine. I believe Robert actually drilled the hole all the way through. One should be able to buy brass or stainless steel tubes of appropriate diameter and thickness, so the vertical hole doesn't need to be drilled. Perhaps Robert will join the thread and will speak to that.

I like the idea of using thermal paste to more quickly conduct an accurate reading. I'll have to try it, although I may not get to that in this initial testing.

Hi Gary,

Funny that we started a similar topic almost the same time . Have I understood correctly? Is the tip of your thermocouple inside the piston rod?

Cheers

Gabor

Yes Gabor, the thermocouple's inside the piston rod. Robert wrote me about his idea back on May 2, so I guess great minds were thinking alike. Actually they'd been thinking this way as far back as when Faema was measuring temperature inside the group in the 1950s. It's a shame they didn't continue that project. This is the thread by Paul Pratt.

Faema Mercurio with thermometer



Here's a link to Gabor's thread, where he's experimenting with temperature measurement on different generations of Olympia Creminas, including fashioning a piston rod for that purpose.

Olympia Cremina Temperature Study, Part 4

I used brass because it is easy to machine! It has a moderately good heat conductivity, about half that of aluminium. I had slightly larger brass bar stock and turned it down to the correct size and then drilled the hole all the way through. It would be very difficult to bore this using a drill press! I had thought I would drill to near the end, but decided that it would be best to drill all the way through so the probe could actually contact the base of the piston. I considered very carefully what diameter the hole needed to be in the cross pin that would leave it strong enough yet permit the passage of a probe! It would be easy to make these with a CNC machine instead of by hand. I seldom have need to make more than one of thing, so do not have such equipment. (And I usually only make small things like parts for microscopes and the like.)

I could not find tubing the right diameter that seemed thick enough to withstand the pressure on it during operation. There must be a place someplace that sells such tubing. It also needs to be thick enough to enable cutting the threads on both ends of it without weakening it substantially.

I thought this to be the ideal temperature probe for first and third generation La Pavonis and for Creminas. There are other machines where this could be done without too much trouble. I thought it did not make too much sense for second generation La Pavonis, because the constant high temperature above the piston would interfere with measurement. With the first and second generation machines pumping the handle up and down brings water in and out of the group, and one can use this method to know exactly when this is done correctly!!! The same seems true for Creminas and others. (Remember the Cremina has hot WATER above the piston, not steam as second generation La Pavonis, and pumping the handle also introduces new hot water above the piston.)

Second generation La Pavonis can be consistently temperature controlled, and being able to measure inside the piston would be more accurate.

Reaching the Zen Zone with a Two Switch La Pavoni

Hi Robert and Gary,

Great. It will be interesting to compare our results. I'm having sealing difficulties anyway and can't experiment with the Creminas during my holiday. It will be fun to follow your topic during that time.

Cheers

Gábor

I suspect a 2nd generation (1974-2000) La Pavoni probe piston rod would best be made of stainless steel, because it is an unusually poor heat conductor. The hole would certainly need to be drilled all the way through (like the one described here) so the probe touched the brass piston with stainless. Otherwise the conductivity of brass would likely result in a too high reading from heat transfer downward from the chamber above the piston that is normally at boiler temperature. Grade V Titanium would be even better for this, but making such parts requires special skills and equipment.