This doesn't answer your question about flavor, but it does show the effect of preheating the piston and/or portafilter on temperature:

Cafelat Robot User Experience

Yes, preheating the piston affects the rate of temperature decline significantly. That much is clear. The commonly assumed air gap theory is wrong. The much more complicated part is how the profile influences flavor. That's not just a Robot question. But in my experience preheating (piston and/or PF & basket) improves the flavor for some coffees, worsens it for some, and doesn't seem to make any difference for others.

Also, I found that a non-metal piston is essentially equivalent to a preheated metal piston.

Glen: thanks for linking to that graph, I had read through this thread back in January and its results didn't stick with me until now.

John: I appreciate the explanation and that graph you prepared linked by Glen is pretty cool.

I don't have tools to measure the water myself (I always like to do my own tests for fun), and also I'm still trying to figure out whether espresso tastes good or bad to me so that's why I've been reading about preheating. Inspired by Ken's suggestion to touch the piston, I did do a quick IR gun test without preheat: piston at my room temp (68F) and immediately after the first shot bottom was 96F, and the silicone gasket was 70F, and slightly above the gasket on the piston side was low-mid 80s. I suppose to increase heat on a chunk of SS metal 20-30F will suck quite a bit of heat from a small amount of water.

I'm really curious why the air gap doesn't work as well as one might think.. maybe it's not large enough or something to do with steam and pressure?

jedovaty wrote:I'm really curious why the air gap doesn't work as well as one might think.. maybe it's not large enough or something to do with steam and pressure?

It will matter if the gap is big enough. How high do you fill the basket with water?

The thing is, when you apply pressure you squeeze the gap down. Air will compress proportionally to the pressure, so a 6 bar pull will make the gap 1/6th it's original width -- if it were all air. But the hot water from the basket results in the headspace being something like 50% water vapor at the time you lock it in. When water vapor is put under pressure it doesn't compress, it condenses into a liquid. So the gap you get is really more like 1/12 the original width. If you pull 9 bar it would be more like 1/18th.

I usually fill to within about 5mm of the rim, give or take. That means that the gap under pressure would be 0.3-0.5mm. And it will be saturated with water vapor making it about twice as thermally conductive as dry air. And as Ken noted, the Barista model makes the gap even smaller because of the air in the tube and pressure gauge that also compresses.

Or at least that's what I imagine is going on.

I underfilled and pulled gently one time and the temperature decline was a lot less. But it was *way* underfilled and a very low pressure shot. It was not practical as a means of mitigating the temperature decline. Preheating is a PITA in my opinion but it works. I think a non-metal piston is a better idea. I made a piston out of PVC and silicone and it worked great but I didn't like the food safety aspect of it for long term use. I wish somebody would sell a food safe plastic piston. I've been meaning to build another myself that is safer but I've been drinking dark coffee for the last year or so and I don't need to preheat anything. Eventually I'll make that piston. I can't drink dark Italian coffee forever.

I wonder how accurate the IR reading is for the hot piston as I read that IR thermometers have issues reading off of different materials and that shiny metal can throw off readings, though I have no idea just how off the reading could be.

https://blog.thermoworks.com/thermomete ... ermometer/
Look under Emissivity

jpender wrote: Or at least that's what I imagine is going on.

Not sure what is going on myself, but I have been told from someone that does not have the metal sleeve on the plastic tube that water does go up into the tube, and I would assume that before any water gets up into the tube that the water would have complete contact with the piston.

Ken5 wrote:I wonder how accurate the IR reading is for the hot piston as I read that IR thermometers have issues reading off of different materials and that shiny metal can throw off readings, though I have no idea just how off the reading could be.

Some IR thermometers have an emissivity adjustment feature. But in this case the actual temperature isn't that important. It's just a metric that you could potentially use for comparison. I had a sensor against the inside face of the piston. The temperature there is always lower than the face that's immersed in water.

Ken5 wrote:Not sure what is going on myself, but I have been told from someone that does not have the metal sleeve on the plastic tube that water does go up into the tube, and I would assume that before any water gets up into the tube that the water would have complete contact with the piston.

It seems like it would have to be that way. When I took off the metal sheath on mine I was surprised how far up the tube the water went. I had convinced myself it wouldn't travel far at all but I was wrong. Maybe my pressure tube had a leak? Even now that I have removed the gauge and plugged the piston I still find that the piston has moisture on it. I dab it with a towel after a shot.

And that transparent portafilter video? People say they see an air gap. When I look at it I don't see a gap. What gap?

Thanks John, and pointing out the minimal air gap, keeping it real

Actually that line that looks like a minimal gap sort of looks like the metal piston that extends past the lower edge of the gasket, giving an illusion of an air gap.

When being nitpicky, how much difference do small variations of temperature really have on the flavor of extractions?

The Robot's been yielding very good results for me on medium-dark roasts (e.g. Redbird) for years now. To adjust to different beans or room temperature, I preheat the portafilter to varying degrees over a gas flame.

Still, I wonder what difference in the consistency of shot quality the finer temperature control of a PID-controlled lever might offer.

Any opinions based on experience? Thanks in advance!