Rough temperature guide (continued)
To give you a start for dialing in shot temperature, I measured maximum and minimum recommended temperatures and did this with and without preheating the cylinder. "Maximum" was as hot as I could get it for a light roast. "Minimum" was targeting approximately 175°F/79°C as needed for some dark roasts (some of which brew well at "standard" shot temperatures of 202°F/94°C.)
Here's a summary of the plots that follow. The starting temperatures are from the reading in my gooseneck kettle:
This is a "rough" temperature guide, as my measurement gear was limited, as described below. Please note that measuring temperature in the coffee is an extra step because the accepted standard is to measure the brew water only.
I tried here to get an overview of what happens with and without preheating the cylinder. I was surprised to find that brew water temperature declined gradually whether the group was preheated or not, as long as the water poured in was hot enough that it exceeded the absorption capacity of the metal. Pouring cool enough water into a group that isn't preheated yields a much greater temperature decline, which may be helpful when brewing dark roasts where you want to reduce harshness. This post completes with Artisan plots of the shots in the table, with commentary.
Measuring temperature inside the brew group of the Flair Signature PRO "should" be easy, but it was challenging in practice. I offer estimates of temperatures in the brew water and coffee bed, given these limitations:
- The Artisan software sampling rate at 3.5 seconds for my computer was slow enough to miss peak temperatures - I compared this to eyeballing the thermometer during one of the maximum temperature trials
- Probes measured 2°F high when immersed in water at a rolling boil
- Probe placement was approximate because the thin wires are flexible
- Shots were pulled manually, so they aren't identical
- The portafilter holding the coffee and the shower screen aren't preheated
I drilled the bottom screen of one portafilter to insert thin wire thermocouples ("probes," or "TCs"). One TC was inserted into the middle of the coffee cake. It appears as BT (Bean Temperature) in Artisan. The other TC was inserted further and threaded through a tiny hole drilled in the shower screen. It appears as ET (Environmental Temperature) in Artisan. These photos will help you visualize probe placement.
Other measurement details
- Brew water was pre-poured for 3 seconds to fully heat the gooseneck kettle spout before pouring into the cylinder top, slightly filling the well where the stem inserts
- 18 gm home-roasted coffee ground for a shot of 25-35 seconds following preinfusion
- Artisan 2.0 software
- MacBook Air 2015 version running MacOS Mojave
- Amprobe TMD-56 data logger and thermometer
- Typical shot profile of 10 - 15 second preinfusion at 1.0 bar, 25-35 second shot at 5 - 6 bar pressure that declined to about 3 bar before remaining pressure was released at the end of the shot
Here are Artisan plots of several shot profiles, with commentary. The brew water temperature drop at the end of each shot indicates that water has been drained from the cylinder.
Artisan missed the peak here, where brew water peaks at 202°F/94°C. With a fully preheated cylinder, temperature decline is very gradual, going to 188°F/87°C one minute in. Keeping in mind the lag time for Artisan sampling, low pressure preinfusion ran for 15 seconds, followed by full pressure. So this covers about a 45 second full pressure pull until the shot completed. Temperature in the coffee cake peaked at 189.5°F/87.5°C (after adjusting for TC calibration). Remember that SCA guidelines measure brew water, not water in the coffee bed, but I thought readers would find that reading interesting, because the ground coffee, the portafilter and the shower screen are also absorbing heat.
Here, brew water is estimated to reach a maximum of 183°F/84°C and declines gradually to 172°F/78°C as the shot nears completion at 49 seconds. The shot fully completes at 55 seconds. The maximum coffee temperature is 147°F/64°C.
In this plot water was added at 211°F/99°C to a brew group that wasn't preheated. Brew water temperature is estimated to peak at 186°F/86°C and declines to 170.5°F/77°C at 49 seconds when the shot is nearing completion. Coffee temperature peaks at 152°F/67°C. The shot is fully complete at 1:05. With water just off the boil poured into a room temperature brew group (about 74°F/23°C), the brew group is sufficiently heated that temperature decline is gradual. Compare this plot to the next one, where cooler water is poured into a room-temperature brew group.
This plot shows a 193°F/89°C shot where the cylinder was not preheated. It was made before receiving my new thermocouples, so there was only one TC in the brew water, and it calibrated correctly in boiling water. Assuming that Artisan still truncates peak temperature, brew water starts at 175.5°F/80°C and steeply declines to 153°F/67°C 25 seconds in -- actual pour-in started 10 seconds after Artisan started recording -- and settles into a more gradual decline as the brew water and brew chamber temperatures approximate each other. The shot ran long because the grind was too fine. The key point here is that this shot tasted good with a coffee home-roasted into 2C (second crack) where oil was starting to show on the beans. The combination of low start temperature and rapid decline reduced harshness for a very mellow, traditional shot.