I fear that what I will be writing here will make me appear either inept or crazy, but here goes. I finally was able to borrow the Scace thermofilter and its Fluke electronic thermometer readout from my local roaster. I tried to follow the WBC Measurement Procedure. I attempted readings at the 199.4° setting, the 203° setting, and at my current setting for the roast I use, which is 3 clicks above 203° (approximating, I hope about 204°, which is where I was at for this roast on my PID'd Silvia, and which seems just about right by taste). All the Scace readings were about 6°F or more higher than expected!

203° on the dial = 209.8 ±0.2°
199.4° = 206.8 ±0.2°
203 = 3 clicks = 215.8 ±0.2°

I repeated some of these, getting the same results. I then tested boiling water and the ambient temperature with the device, and found it accurate to within 1 degree. So the thermofilter was reading correctly. I finally put the thermofilter on my PID'd Silvia and found that it read out 204°, exactly where it had been set. So probably I am neither crazy nor inept. I then phoned La Marzocco and spoke with a technician. He went through the expected material about flow rates, pressures, and the like, but did say that there were heaters just above the group head for temperature control. His basic position, with which I certainly agree, was that the Linea Mini was not conceived of for exact readouts, and one should go with one's taste and experience and, essentially, "if you've got it dialed in and are satisfied, you shouldn't be complaining if I were you".

He is undoubtedly correct. But I can't figure out what was happening.

. . . and the Temperature Sensor is a type N Thermocouple (Standard: +/- 2.2C or +/- .75% accuracy).

From the Linea Mini parts section - "Linea Mini NTC Temperature Probe". NTC is an abbreviation for negative temperature coefficient.

The temperature probe is a thermistor LIKELY having a resistance of 50K ohms at 25 degrees C. As the temperature rises, the resistance of the thermistor goes down. This particular thermistor value is very common in espresso machines fitted with a Gicar PID.

lcats wrote:His basic position, with which I certainly agree, was that the Linea Mini was not conceived of for exact readouts, and one should go with one's taste and experience and, essentially, "if you've got it dialed in and are satisfied, you shouldn't be complaining if I were you".

That's kind of the position I have accepted before purchasing. This machine isn't for that person that wants to manage temp by the numbers. I manage my temp by taste. The Linea Mini is not at all interested in giving you data. It's designed to giving you rock solid consistency from shot to shot so you can get on with obsessing over your technique rather than the machine itself. So in a way, ironically, it's a better machine to train on.

I have a hard time believing that the actual temp at the group is that much higher than the temp on the dial. As far as I know, 215 degrees is way too high to produce acceptable espresso, regardless of roast, blend, or grind. I suspect that something is off in the way you did your testing, but it's hard to say without more data.

In any event, if you're able to adjust the machine to get tasty results with your beans and setup, that's all that really matters. The rest is just a distraction.

And indeed today I got a different, slightly darker, roast and was able to dial it in very rapidly at the 203° setting on the dial, or slightly lower. So no problem with wrapping my mind around the machine and its dial.

Still, I'd like an explanation for what happened. I did the same manoever on the Silvia side-by-side with Linea Mini and the thermofilter behaved normally.