jonr wrote:I start the sequence (ie, raising the lever, with descent about 15 seconds later) when the outside of the group head (side, narrow part) rises from cold to 76C. While I haven't checked this machine, I can tell with a high degree of certainty that I'm in the correct low 90s for coffee brew temp (similar to Matt's numbers). So I'm curious why the vastly different results.

With the boiler at ~117C and stable, pre-brew group head temps of over 100C, it's not clear to me where 93C water could possibly come from.

I lift the lever when the thermocouple (mounted just above the flare, midway between the boiler attachment plate and the front of the group) reads circa 74C. That is with a pre-millennium, pressurestat equipped LP. I presume the starting temperatures would be different with a sleeved model.

Matt

I had to add this in here. I traveled to my paretns house over Christmas and broke the thermocouple on my gauge. I was pretty upset, but then I found this for 8 bucks on amazon.

http://www.amazon.com/gp/product/B00L32 ... UTF8&psc=1

It is very consistent and constant error of .5C

It accepts Ktype so a variety of other apoplications, it is cheaper than a polder, and there is no risk of breaking anything. I just route the wire to the side and have the device on a shelf, leave it on nonstop and look at the temp whenever I need it. If I want it off, no problem, just take the rubberband/oring off and its gone. Perfect and simple.

So after playing around with the thermostrips I decided that I prefer the accuracy and speed of a digital thermometer. Mounting them in a non invasive way can be a challenge however. So far I like pizzigri's brass ovp valve fixture, as well as the liner between the group head and the boiler (however some of my older machines don't have the bolt detachable group head).

I did come up with my own, low effort solution however; not quite as elegant but it is simple so I figured I would post it here.

This is the thermometer I am using, a $15 jobber from Walmart. It has some other nifty features like a timer which I can use to time shots and an alarm when it hits a certain temperature. This is helpful to let me know when my group head hits 195F.



It's model number is 1487-9
It has a nice long lead for the thermistor.


Getting at the thermistor is rather easy, just undo the crimp on the metal sheath with some pliers.




I found these guys at Home Depot for super cheap. They have them in various sizes and are made of a heat resistant rubber


I just wrap it around the neck of the group to hold the thermistor against the group head, and the thermometer/timer itself has a magnet on the back that lets me just clink it right to the base when I'm not actively looking at it.

Well, I wanted to share some slight changing in adding thermometry that so far is proven very effective. It consist of two parts

1: improved accuracy of measuring pipe temperatures with faster update speeds

2: a different method of attaching probe, though not fully proven yet.

3: How to use Ramiken method and not get false reading.


----------------

I was looking into better K-type probes for reading pipe temperatures and I came across an article (will link if I find it again) by someone that had designed a form of pipe clamp probe. He talked about the process of developing the probe and one of his first and most successful methods involved using a bare wire probe and peeling back the insulation about an 3/4 an inch as I recall. It is of course critical to keep the wires separate and not in contact, but if they are laid out full length along the wall of the pipe, along with the bead, then he found it gave far more accurate and rapid readings of pipe temperature (presumably because of the increased heat transfer to the bead through the wires.

Thus, I peeled back the insulation on mine and found he was correct.

I then took a simple piece of heat shrink tubing and ran the thermocouple through the tubing, cut out a slit and let the bare wire just come out and lay out 3/4 of an inch with the bead and then superglued the inner insulated part so that it would hold in place and ensure the wires stayed out of contact with one another. I then wrapped the tubing around the group and on the second wrap used some brush on super glue to hold it together. I let that set and then turned on the machine and hit the exterior with a hair drier to shrink the tubing and pull everything close together. It worked well except at the back side got a little loose, I may want to redo it at some point to make the whole thing tighter, but once the heat tube shrank it now pulls very snug against the group.

It would probably have been even better to use some thermal base on the probe exposed wires too.

What I have found has been really interesting. In the same spot I used to find readings of 86 degree C to be the sweet spot, but it was somewhat inconsistent and I could lower the temperature very quickly with a ramiken, but the temperature would raise very quickly.

WIth the new set up the temperature readings are far more stable. THe ramiken method still works wonderfully, but it takes more pumps of cold water to lower the temperature-especially after a couple of shots. My temperature reading for the sweet spot is now around 91C on my current coffee- previously this was way too hot in the same location, and there is much more flexibility with usable range-suggesting that the temperature reading is more accurate (before the reading was less accurate and therefore had a larger margin of error, imo). Most importantly, the temperatuer increase once I have lowered to my target temp is now MUCH more stable and steady on the increase. THis means I can much more easily control what the actual shot temperature will be when I pull the shot, whereas before I did this if I had pulled a couple of shots the temperature would be jumping up really fast and I would have to race against it to get the shot in the range I wanted (of course the thermocouple is slower than the pipe, therefore that method was inherently flawed).

Finally, on point 3, I have learned to make sure that I use short pumps of ice water, not long raises. It is not obvious to me that the long raise with water is ineffective because most of the heat is in the actual piston. When raising the piston you move the primary heat source away from the location of the temperature measurement, and therefore get a false lower temperature reading. The spot on the pipe the measurement is taken may be lower, but the piston itself is the same temperature and as soon as you lower it the temperature will sky rocket (I noticed this a while back, but am just throwing it in with this post for others that might not realize that yet). Thus short pulls of cold water work better to actually lower the piston temperature, and therefore more effecitvely control shot temp.

Paul Pratt has shown us that Faema did this first on one of their very early commercial levers.

Faema Mercurio with thermometer

Notice the left side of the group.



Here is a close-up.



And here's what Paul writes about these machines.

Paul Pratt wrote:How cool is that? It is a group head thermometer that displays the group temperature and still works. The inside of the machine was completely original, no pressure switch! This machine had a high and lo element system like the 1st Pavonis.

My thermometry solution form back in 2012:



I used Gary's idea of an O-ring to hold the temperature bead to the bell, then a little piece of steel plumber's strap folded over into a U-shape to clamp around the remaining stub of the thermometer shaft, with the group-mounting bolt to hold it all on. Note, I have subsequently gone to stainless steel Allen headed screws, a little bit longer, to get more depth/bite into the group-mounting pad on the boiler. I could have taken more time to form the little piece of steel strap into a more conformal fit around the thermometer shaft, but it's hardly noticeable unless you are looking for it. Tooting my own horn here but I feel my thermometer mount is one of the cleaner and simpler that I've seen in this thread.

-Peter

mathof wrote:In addition to the pleasure of seeing the numbers go in the intended direction, I find the taste of espresso is softer and less harsh as compared to shots pulled without the dissipator.

Matt

Matt,

Where do you find the heat dissipator? I've been googling like crazy and can't find one.

Got to try out mine, maybe late of this evening and see how accurate. Ideal group temperature is 92C??

Ideal group temperature is what tastes right with the coffee you are drinking. Your Europiccola is one of the very old ones with a brass sleeve and before they had a drip tray. Once it is warmed up on the low heating element (minimo) the temperature should be pretty stable and good. Half pumps will bring the group temperature up for lighter roasts.

drgary wrote: Half pumps will bring the group temperature up for lighter roasts.

Hi Dr.Gary, All my coffee beans are in Full City Roast, so do I still need to half pumps?

Thank you