If you want the dynamics of actual shots, the only way I know is to rig up a Schomer basket. This is a regular basket with one or more TCs protruding to monitor the temperature at various depths of the puck (classically, use only one at the top of the puck, but Dan rigged one up with three TCs at various depths). Then fill the basket with coffee as normal, and make your shot. A TC at the top of the puck will react more or less the same way as a Scace; those deeper down will have long lags associated with them.

If you make shots at very different flow rates, things get complicated. Lets say it's an old Linea with a monster heater in the brew boiler. The group is at 92C and the brew boiler is at about 94C, and stays there regardless of flow rate. Then the shot will start at 92 and rise to 94C at a rate depending on how slow or fast the flow is. Now take a DB E61, with the group at 92C and the boiler at 104C. Again, the shot begins at 92C. If the designers are really incompetent and put in a monster heater, the shot temperature will steadily rise and start boiling. If the designers are competent, they will underpower the heater, tweak the PID constants, and put in delays, so the brew boiler temperature drops during the shot. The ideal tweak has the brew boiler temperature dropping at a rate that offsets the rise in group temperature regardless of flow rate or shot volume.

This kind of tweaking has dramatically improved the performance of every kind of DB (even the saturated group ones, where the problem is less dramatic, have been tweaked to slow down brew boiler heating). On HX machines, there are similar kinds of tuning tricks that distinguish the great machines from the also rans. The need to flush espresso machines for temperature management has gradually passed, as semi-commercial machines have benefited from these tricks.

The Scace was initially designed for testing and and calbrating Barista competion machines. The testing part was to see if the machines could maintain a flat line profile at various traffic rates; the calibrationg part was to make sure the offset produced an accurate display of the shot temperature. Greg's advice was to use the reading after 5 seconds of flow, since the first few seconds were affected by the thermal mass of the filter, and the later part by the interplay of group and brew boiler temperatures

For this calibration purpose, using flash boiling will get the same result (or given the ambiguities of simmering and boiling, within 1 degree). The method won't tell you anything about the temperature dynamics or stability of the machine. It won't give you any bragging rights. But given that the early part of the shot is most critical for the taste, It will give you a fairly accurate indicator of the effective shot temperature.

another_jim wrote:The goal of the calibration is to set the panel temperature to the group temperature, which is also the initial shot temperature. <snippage>

Jim,

I want to provide some data from my Eric S E61 GH thermometer in a Vetrano DB machine. I did your calibration last night to determine the offset and it indicated that I should reduce my offset from 15 C (27 F) to 12-13C (22-23). I actually only reduced it to 24 F. I then used my usual 202 F SV (panel display).

This morning after several hours idling with bottomless portafilter in the group, the PID read 202, the GH thermometer read 196, indicating that the GH is idling much cooler than the PID display with offset. When pulling the shot, the thermometer read ~206-207 while the GH pressurizes for a couple seconds, it then fell to ~205 and held fairly steady through the rest of the shot and then it dropped to ~198-199 when I stopped the shot.

In the past when I had my offset at the factory 27 degrees F, and the SV (panel display) at 202, the machine would idle at 199, pressurizing peak was ~209-210, reading during rest of the shot was 207, and temp was ~200 after shot. BTW, CC factory setting is offset of 27 and SV of 203

Ignoring the fact that different parts of the GH are at different temperatures, my interpretation of this morning's measurements is

• The GH temp is 196 before the shot where the thermometer is inserted
• The water initially reaching the GH from the boiler has a temperature of ~207 when it gets to where the thermometer is inserted
• As the GH pressurizes the water is heating the GH and the GH is cooling the water to maintain a relatively steady water temp
• The GH temp is ~198 after the shot where the thermometer is inserted

I am not sure what is happening with the water flow through the head during the calibration since I presume the head is not pressurizing as it would against a blind filter or a Scace device or when pulling a shot. I believe the thermometer is measuring water temp during the shot and GH temp before and after. But I don't think that the calibration is leading to a display of the GH temp. I do think that it would be nice to get the displayed SV close to what we expect for the initial shot temp or average shot temp or ... .

My own approach is that I'd like the display to be in the ballpark so I can use it as a guide if I have a new coffee and a brew temp recommendation, but I make my temp adjustments by taste. I'll see what I think of the shots over the next day or so since if I haven't confused myself I am likely now pulling my shots 3 degrees F cooler.

Alan

First, what was the GH thermometer reading when you did the flash boil test?

When idling, the thermometer monitors the cooler saucer part of the group, which does not reflect the initial shot temperature. The best approximation of initial shot temperature is the idle temperature of at the group's neck. Your thermometer stabilized at 205 when your PID readout read 202. If you had reduced the offset to the one indicated by the test, it would have been at 203F when the readout was at 202F.

Eric's GH thermometer has become a de facto standard for many E61 owners. If had used its initial shot value to calibrate the offset, you would have set the offset to 1C less than the one recommended by the flash boil procedure.

So I'm not sure what's up here. You used the flash boil procedure, and decide to go with an offset 3F too hot according to both your thermometer's readings and the flash boil test. Then you conclude it's best to just wing it. Why on earth don't set your panel display temperature to equal the stabilized GH thermometer readings at the start of the shot? Then you can at least wing it with equal length wings.

Hi Jim,

I'd like to begin by apologizing if anything in my content or tone offended. I have found your help here extremely useful over many years, and I was not attempting to provoke a debate, but rather just trying to better understand what is going on.

Yesterday I collected some flash boiling data leaving my PID in F mode since it seems to work fine that way. The offset I determined using your technique was 22 F. BTW at this offset and a display value of 212 F, the GH thermometer was cruising at 200 F and it read 211 and rising within a second or so of drawing water through the group. This morning at a 22 F offset and my SV displaying 202 F, the GH thermometer was cruising at 193. When I pulled the shot, the GH thermometer shot up to 203 then quickly dropped to 201.5 and slowly rose to 202.5 over the course of a 26 second shot. It dropped to 196 F when I ended the shot. So it appears that at least in my hands the calibration process gives me a display temperature that is close to the GH thermometer "plateau" reading (see below) of water temp during the shot.

If I may, I'd like to explain why I was skeptical about dropping my offset (and presumably my brew temp) by five degrees from the factory setting. Using a Scace device and GH thermometer on my old Vetrano HX machine after a long cooling flush, and reviewing Eric's published data using a GH thermometer and similar thermofilter devices on both HX and DB machines, I see the GH thermometer reading a water temperature significantly hotter than the thermofilter during the shot. Typically, after an initial higher temp hump, the GH thermometer settles to a plateau that is hotter than the the thermofilter during the shot and then the GH temp drops to a temperature very close to the thermofilter temperature immediately afterward the shot. See figure in post two of this thread showing temp profiles for GH and thermofilter for two walk up shots on a Vibiemme Domobar DB.

Perhaps I don't understand what the data in Eric's figure is telling me, but it looks like the GH thermometer is cruising at about 197 F before the shot. During the shot it rapidly increases to about 210 and then settles down to 206-207 before dropping to 200 immediately after the shot. OTH, the thermofilter heats over a few seconds to 200-201 F and pretty much stays there. I presume that the portafilter is supposed to approximate a brew profile and that the target brew temperature was 200 not 206. If that is the case, then I expect that my brew temperature this morning was 4-7 degrees F lower than the 202 F display.

Sincerely,
Alan

Hi Alan,

I'm glad you tried the test; but it looks like you are not getting what Dan and I got.

I don't have a GH thermometer, so I don't know its peculiarities. From the posts about it, it looks like it systematically reads higher than the Scace during a trial. I guess its because the the basket and group bell are shedding heat, while the GH thermometer location is more insulated. So the slower the flow, the bigger the difference will be.

In any case, I compared my flash boil calibration to the calibration I got using a home built device that is similar to the Scace (we were building TCs into padded baskets for several years before Greg started marketing his version, mine dates from 2003). The reading in the basket after 10 seconds at double espresso flow rates coincided with the flash boil test, as I said in the original post. Dan tried it his Bianca and Strada using the Scace II and also got agreement within 1C.

I'm wondering what others have found. Maybe the technique doesn't work on all DBs and needs to go back to the drawing board?

One possibility. If you look at the video, you'll see that you will get a boil after 5 or 6 seconds even if the pour starts at 98C, since the brew boiler water heats the group. Over 10 seconds of unrestricted flow, you can get as much as a 4C rise, as the hotter boiler water heats the cooler group. The flash boil point is that temperature at which the boil is almost immediate (after 2 to 3 seconds) and very vigorous. If you waited longer for a vigorous boil, that may explain your too cool calibration. Just a thought; otherwise I'm stumped.

Jim,

I have two other indications about my machine. First, CC tech says that they calibrate the offset using a Scale and their offset is 27 F (15C). Second, DavecUK posted on CG that his offset is 11 C (20 F) based on a method very similar to yours. People using the flash boil calibration were consistently getting offsets in the range of 10-12C. So I could definitely be off, but not by more than 1 C or so. There was a little back and forth between CC and DavecUK, in one thread on CG but nothing conclusive. BTW, neither of these calibration techniques involves a GH thermometer.

The temp difference of 4C or 7F seems rather too large to attribute to changes in ambient temp or variation between samples of the same machine.

Just a dumb question here, but does the group pressurize and operate during the flash boil calibration as it would when pulling a shot or using a Scace-like device?

Thanks,
Alan

Alan, roughly 750 people are following this thread, and have spent maybe 15 minutes each reading your five posts and my responses. On the fifth, you tell us we've been spending all this time on your rehash of points from an old CG thread. Wouldn't it have been simpler to state the flash boil technique isn't new, and given the links to its earlier description in your first post? I for one would have been pleased as punch to move the references and credits to the top post, and go on to more productive things. I'll make the changes to the top post once you post the links; other than that, I'm done with this ...

Tonefish wrote:It looks to me that they all reach the point of steaming, bubbling, and boiling, with the only significant difference being the time it takes, and the hotter you get the more aggressive it is.

It is interesting too that they all coalesce to a single stream. My E61 never does that with a Lelit or IMS screen. It looks like the bubbles are causing the coalescence.

Nope. Not if you actually do it! There are quite distinct differences between the slow boil at around 99C and the flash steam and turbulent energy in the output as the temp moves 1-2 degrees above local boiling point.
This method worked well on my Bianca and is a reliable way to set up the offset. As my machines offset turns out to be 8C rather than the factory setting of 10C, its been useful for getting recipe sharing and referral into a good place.
There is a genuine potential benefit for people who don't have accurate measurement devices but need to set up or check an E61 offset or brew water temp.
It'll be interesting to see if other machines are as easy to use this method with, but I can't see why not, with minor adaptation.
Thanks Jim!

Graham J wrote:As my machines offset turns out to be 8C rather than the factory setting of 10C, its been useful for getting recipe sharing and referral into a good place.

Mine clocks in at 14C; Dan's was good at 10C. Dan has a two group Strada, and the offsets are 1.4F and 4F. Thermosyphons and saturated groups rely on thermally driven circulation; which I guess is inherently chaotic, so that small machine to machine changes within the same model create appreciable offset differences. This means people who get thermosyphon or saturated group machines should all do some kind of calibration check. I wonder if groups that have metal to metal connections to the boiler also have these kinds of variations?

This is all kind of new to me. Is it news? I know people have discussed problems with offsets on their machines, but all in terms of believing the manufacturers have screwed up. It seems instead we should expect these variations in perfectly well designed and built machines.

Graham J wrote:Nope.

Huh? What I said was very obvious in the video. You said it yourself ... slow boil versus flash steam and turbulent energy. Maybe you need to watch the video again?