I am looking for a discussion about controlling the temperature of a commercial lever group. With the CMA group that I have, a bottomless portafilter, fine setting of the BV and a bit of distribution generally produces no channeling and a sweet and complex shot when the group is warmed up. Too cold the shot is sour and too simple. Through the winter the group did not produce bitter shots with my pressurestat set to cycle off at 1.25 bar as long as the pour appeared good. However, I am concerned that in the warmth of a summer day, outside things may overheat. Then at my local Farmer's Market, I could produce a shot I think looks extracted well, but not hold the ideal flavors I wish to share.

I read recently in another thread that 3 hours spent with Gwylim Davies on the Victoria Arduino did not produce great shots, just consistently very good shots.

So I tried once more to measure temperatures. My TC doesn't last placed between the portafilter and the puck and I didn't find the measurements made much sense previously so I measured in a different way. Like using the Richard Penney's thermometer strip on the Pavoni, the method I seek would allow me to know by temperature when the shot will drift into the overheated bitter zone. I tried a method today that seems to hold some promise. (A temperature strip is not calibrated correctly since the sink effect of the group has the outside quite cool but the superheated water entering the group and interacting with the coffee is nearer the ideal extraction zone.)

I placed a probe down in the plastic cowling blindly trying to make sure it is near metal. (TC1) I placed another probe against the middle of the bell and covered it with metal ductwork tape. (TC2) BTW, that stuff doesn't stick too well (a good thing since one of my extra groups will always have the firmly adherent thermometer colorstrip attached or lose some chrome - I would guess). Then I used my usual routine, pulling tasty shots TC 1 in the range of 60-75 and TC2 50-65 (all temperatures in Celsius). I found that the machine eventually settled into temperatures of TC1 low 70s and TC2 low 60s. However by running water through groups or pulling shots in succession TC1 climbed into the low 80s and TC2 closer to 70. In that temperature range the shots were bitter with deleterious effect on straight espresso and cappuccinos. (based on Zoka Espresso Quattro 15 grams app 10 sec preinfusion 28 sec extractions.) I was able to cool the group with a bowl of cold water drawn up into the group until the TC probe came back down to 73C and again found the espresso sweet.

I will need to repeat this testing to see if I can use my thermometer to notify me when to cool the group. I will test again tomorrow.

Anyone input as always is appreciated.

Hi Flint. You lost me on this one. I don't even know what a TC is. I just fiddle around with the temperature and the grind until it tastes right on my Astoria lever at the markets.

Hi Clint, Great! (TC is a thermocouple or temperature sensor for digital thermometer - abbreviation I've seen used on these fora.) Sorry if the post is disjointed.

I was hoping you might see the thread. As a barista one cannot taste the drink. So aside from visual cues, how can one be sure of the quality of the drink. The group is quite forgiving and the only drawback is that one cannot dial in the temperature profile. One can use stale coffee or grind too coarse or underdose and thereby overwhelm the heat sink to overheat the extraction or one can do the various steps right and end up with a really tasty treat, but when your groups are hot, I suggest you pull 2-3 shots in succession and taste the next one. Might be bitter...

Before the market gets busy, I pull three or four shots and get the temperature and the grind right so the shot tastes balanced. Today was really busy, so I just trust that my dose is pretty consistent and just keep going. I keep my eye on the bottomless and on the color of the stream out of the other group. As the day heats up, I can see I need to adjust the grind. Every time I do that, I stop and taste a shot and make sure it's still good. If I pull a bad shot (for whatever reason) I just dump it, tell the customer that it would have been crap, adjust on the fly, and usually the next one is okay. If I'm not totally sure, I'll give the customer or someone else a shot and ask for feedback. Taking cash, grinding, pulling shots, steaming, washing glasses, getting milk from the cooler into the thermos, pulling the manual fill........ I don't have time to do it perfectly or be super precise.

The feedback has been astonishing. I'm new at this, so I know I have a long way to go. I think the coffee served around this valley must be really bad to warrant the feedback I've been getting.

As to the temperature of the group. It doesn't seem to matter how much use it, the temperature of it seems to stay pretty consistent (based on feel). I think the machine is bombproof and well designed and for the most part, I just pull the levers and let it rip. I hope you're having fun with the market. I sure am.

This is of great interest to me as I rebuild my single group CMA lever.

Chert wrote: I placed another probe against the middle of the bell and covered it with metal ductwork tape. (TC2) BTW, that stuff doesn't stick too well

Kapton™ tape is what the Dr. Prescribes for situations like this. It has a strong adhesive, but comes up clean, is electrically insulating, and can withstand high (and low) temperatures. It's moderately expensive (per roll) but if you know an electronics lab technician, he may be able to give you a few inches of tape.

Regarding temperature control, I was contemplating adding a PID to the boiler. My thinking is that it will give better temperature stability and easier tweakability, versus the pressurestat. - the temperature would still put the boiler in the 1.2 bar range. I'd put the Solid State Relay in series with the pstat (set high) to keep the pstat as a safety cutoff. The Astoria boiler has that convenient drain plug - looks like an ideal spot for the thermocouple. But that doesn't give group temperatures.

But even with a PID, one needs to choose the target temperature.
I'm used to the HX flush on my other machine and adjust temperatures by feel.
This one is going to be interesting to figure out. At least I'm working in a controlled environment (indoors) and not lugging my machine around in all kinds of weather....

Hi Allon - is that a Lego machine in your picture?

Clint, No doubt the espresso is now quite tasty at your stand. I would surely enjoy seeing your routine and tasting. My first run will be (or is planned for) next Saturday.

Allon, My boiler is well insulated which hopefully adds enough stability. Perhaps a complicated PID could take input from the group itself to adjust the boiler temperature for a lower set point as the group reaches a certain heat. Thanks for the tip about Kapton.

I have a bag of Backporch blend for today's tests. 15.8 grams in a smallish double basket. The thermocouple positions are the same TC1 in the cowling, TC2 taped to the vertical middle of the bell. I pulled four shots: 1. TC1=49C TC2=45C; 2. TC1= 50C TC2=56C; 3. TC1=52C TC2=69C; 4 TC1=53C TC2=74C. Shot 1 was sour and undefined. Shot 2 had lost the sourness but lacked clarity. Shots 3 and 4 brought out the fruit flavors without sourness.

I'll pull another series later when the group reaches steady state.

After four hours I ran four more shots, same coffee, dose, grind settings as before. 4 seconds preinfusion, 30 seconds total extraction time, shots stopped 26 sec after the lever allowed to raise.

I recorded the two temperature probes before each shot.
1. TC1=83.4 C TC2=67.6 C
2. TC1=71.2 TC2=85.2
3. TC1=69 TC2= 85.8
Before the fourth shot, I placed the bowl of room temperature water beneath the group and drew the water up into it to cool it down.
4. TC1=68 TC2=74

The flavor results? Every shot was quite good, displaying the sweet fruit forward/floral notes in this blend. This time I found no advantage in cooling the group. These second group of experiments suggests that the shot parameters are key and the group design manages temperature quite well, just as Clint's experience, noted above.

Hi Flint,

I sent you a PM.