For Silvia owners confused about temperatures surfing, these are results Ive obtained on my machine which has a 100C thermostat.
Room temperature: 22C. *Machine pre-heated for 30 minutes before times and temperatures taken.
Temperatures taken with a K-type thermocouple attached to the top of the boiler, readings with a digital multimeter.
The heating light is forced on by bleeding water through the group, with portafilter attached, but basket empty, until the light comes on.
Heating element light comes on at *0 seconds, *temperature 92-94C degrees
Heating element *light turns off at 50-52 seconds, temperature 102C *degrees
***PULL SHOT AT 62-65 seconds, temperature 108C degrees. (108 at boiler roof = water at 95C
exiting shower screen)
At *30 seconds, the top of boiler temperature is 88-90C degrees --far too cold.
at 40 seconds the top of boiler temperature is around 95C --still far too cold. Remember, the water exiting the shower screen is around 13C colder than the top of boiler. It is qute likely that those suggesting pulling shots at 40 seconds confused the boiler temperature of 95 degreees, with the supposed water temperature. *They may also have used the older machine with a 110C thermostat.
The maximum *temperature reached is 115C degrees, at around 80 seconds. Thermal dynamics mean the water, and boiler, keep heating for 30 seconds after power is no longer delivered to the heating element.
The temperature stays within +- 1 degreee of 115C for another whopping minute before it begins the descent. *
Therefore the Silvia is at its most thermally stable from 120" to 2 20" after the light comes on --- *but the downside is that the temperature then is 115C --probably a tad on the high side for a shot, *but Im going to give it a try to see what it would taste like.
The most obvious variables which would influence temperature surfing are:
--ambiant air temperature
--how cold the water poured into the tank is
--length of pre-heat time
--accuracy of test equiment!
--accuracy of the thermostat
Hope this all helps brew a better coffee from a pretty good machine.
I think you have to be careful how you interpret what youre reading. For instance a thermocouple is not a linear device, so one or two temperature spot calibrations can still result in errors of up to 10%, but more likely around 5%. I have calibrated my K-type devices against a calibrated reference using around 50 points over the range of 0-100 C and need to use a nonlinear curve to fit the data. Then I use the curve as a calibration curve for future work. I regularly see my K-type TC reading high by 4-5 C at brew temp, as determined by the DMM. Everything is corrected on the computer using the calibration curve before being presented.
Ive had a 100 C Silvia in my hands for a few days. Enough time to thoroughly thermolog it, whilst leaving it running. I found for this machine the most stable brewing region was approximately 130" to 2 after the element turned off, which gave the temperature of the water exiting the group around 92 C. I didnt monitor the boiler temperature, just the group temperature and water temperature under the shower screen. I wouldnt say these results are gospel, as every machine is different. Without recourse to this measuring equipment, Id say use your taste buds to determine the timing that is best for you. Bitter = too hot, sour = too cold. These measurements are just a guideline, but are good for at least finding some place to start.
Bearing in mind the boiler cycle time is around 8 minutes, Id start my timing from the 2 mark (after the element turns off) and move in 20 sec intervals either way, letting your tastebuds do the talking. Also a machine hat has not been on all day may require less time, as the group will probably be cooler.
Just a few extra hints and tips for those without measuring equipment.