From Favorite Espresso Blends 2011:

cannonfodder wrote:<snip> I moved up and down the temperature scale and found 198-201 to be the working range of the coffee. I pulled most of my shots between 199-200. At 201 some bitters start to work into the cup...

I settled in on 19 grams pulled in 22 seconds at 199-200F for 22 grams extracted.

As a fellow Elektra owner/operator, I'm curious as to how you know you are hitting these temperatures precisely with this machine? You've never mentioned running a PID to record grouphead water temps, so how are you differentiating 1 degree temps?

Just curious...

Easy, I have a Scace. After many hours of cooling flushes followed by Scace measurements I know that flushing x seconds after flash boil and x seconds or recovery I get Y temperature. There is a certain amount of variance but I can regularly get within half a degree of my desired temperature. That holds true for any heat exchanger.

Thanks Dave, wish I had a Scace to help dial mine in that tightly.

cannonfodder wrote:... certain amount of variance but I can regularly get within half a degree of my desired temperature. That holds true for any heat exchanger.

I have a Scace and a Bezzera BZ07. I have not been able to achieve anything like that consistency, even after many trials. I get about 1 to 2 degree variance either way from the nominal temp, along with a rising curve. i wonder if the machines are that different or if it's something else. My results seem sort of like the way Jim S has described his Elektra Semi, if I understand him correctly.

Alternatively it's possible that it has to do with the tight fit of the Scace in the Bezzera head, as the Bezzera has very little headspace. I had to machine off a mm or two from the top of the Scace puck to avoid interference.

Different machines behave differently. Keep in mind my machine is the A3, not a semi. I have a 6 litre boiler and big heat exchanger. That is a lot of thermal mass and helps to stabilise everything. It does have an inclining temperature profile. I will typically get a couple tenths variance during the shot but at around 25 seconds the temperature starts to tick up a tenth of a degree every second or two. So if you run long ristrettos the temperature can creep up at the end of shot. It is also plumbed in so I have a relatively constant cold water supply. The input water has not been heating up in a reservoir. The water in a machine housed reservoir can get upwards of 95F depending on the machine while my HX supply water is in the 60-70 range which helps to keep the temperature from ramping up too fast.

I guess they are just very different machines. I have checked and my reservoir temperature is quite constant during testing. I typically see 1 or 2 or more degrees rise during a 25 sec shot, along with plus or minus a degree or two from shot to shot.

But the coffee generally tastes good...

And that is all that really matters. I think folks get to caught up in temperature measurements. It is one of the few things that can be measured relatively easy so we tend to focus in on that. Personally, I think there are many other factors just as important or more so than uber flat temperature profiles. As long as a machine produces a reasonable temperature profile, the key is the ability to reproduce those temperatures. You can do a lot on a HX machine to tweak the temperatures. You can add flow restrictors to the thermosiphon, change the pressurestat, change the injector tube in the heat exchanger, change the ambient temperatures around the machine, etc... I have used PID controlled single boilers, heat exchanger, steam pressurised levers, etc... and they all produced good shots, just different. It is just a matter of learning the nuances of you kit and that only comes with time and experimentation.

Forgive me for my ignorance, but considering chain of variances in the process, including roast variability within one lot, is it practical to worry about minor temperature variance during extraction? I am not trying to be a smart alec in this business while I am a novice at best, but from my other life where I make my living I have learned a thing or two about process, and therefore I have been merely wondering how far it is practical to be "exact" in coffee business? I though it's more off (like in JJ Cale song) little bit of this, little bit of that, and then you get the espresso...

It may not matter much, but I would feel better knowing that the shot to shot variation is less than +/- 2 degrees. It usually is, but not always. Probably best not to know. TMI as the kids say.