Best flush routine for Bezzera Strega? - Page 3

#21: Post by **JohnB.** » February 5th, 2012, 12:39 pm

sekihk wrote:My boiler started when pressure gauge readings dropped to 0.8 bar. The boiler would stop when the gauge readings reached about 1.3 bar.

If that is correct it might explain your temp findings. Was that suppose to be 01.18 cut in or is 0.8b correct?? In the owners thread most if not all of the settings posted run within a .15b deadband. Mine originally cut in at 1.07 & out at 1.22b. Now it cuts in at 1.23 & out at 1.38b.

If the numbers you posted are correct you have a .5b deadband which would affect your HX temps & indicate a defective P'stat.

February 5th, 2012, 2:52 pm

@dennis - I think both my charts and jim's charts had drawn to the same conclusion: cooling flushes give you inconsistent starting brew temperature while no flushing would gives you temperature same as brew head thermostat setting. Read carefully Jim's last chart and you'll know if you intended to do a 2s flush, end up you would get brew temperature higher or lower than you had after a 1s or 3s flush. That's unpredictable basically.

@eric - thanks for the sectional drawing. After reading your drawings and reconsidering the implications behind all the testing results, I think the effective way to control the brew temperature is to control the thermostat at the brew head. Seems to me that when water left the HX and entered the head, it's temperature should be below 95'C. Then the head's heat capacity would heat up the water to close to 95'C (or above, depends on the actual thermostat setting). It's evidenced by the fact that if you flush longer, the head and brew water temperature would drop more. If I'm to control the peak brew temperature, i should PID the brew head temperature. If I'm to flatten the slope of the declining brew temperature profile, I'd have to increase the pressure stat so that the water leaving the HX should be hotter than factory setting and closer to the temperature of the brew head.

Eric, do you have any thermocouple that could replace the factory thermo sensor used in the Strega brew head? I'd like to try PID the temperature of the brew head.

#23: Post by **tekomino** » February 5th, 2012, 3:09 pm

sekihk wrote:I think both my charts and jim's charts had drawn to the same conclusion: cooling flushes give you inconsistent starting brew temperature while no flushing would gives you temperature same as brew head thermostat setting. Read carefully Jim's last chart and you'll know if you intended to do a 2s flush, end up you would get brew temperature higher or lower than you had after a 1s or 3s flush. That's unpredictable basically.

I see much more variations in your graphs. That's why I am questioning your methodology. Did you always start with the fresh ground coffee at the same dose and same thermocouple position or did you simply reuse spent puck?

#24: Post by **another_jim** » February 5th, 2012, 3:19 pm

sekihk wrote:The Strega is a different lever machine than the traditional ones. There's an additional temperature controlled heater installed by factory to heat up the brew head. The factory must know that there's some design needs in it to justify such additional cost.

As far as I know, it's a cost saver.

---------------------- Begin boring narrative mode ----------------------

HX groups need to idle at roughly 90C, so they cool off the overheated HX water. Repeated use is supposed to heat them up to 92/93C, while the HX drops to that temperature. With proper tuning, flushes have no large effect, and long flushes hurt, since they are designed to be flushed as you make shots.

For lever groups and the simpler pump groups, the heating is done by bolting to a boiler. For more complex groups, a thermosyphon is used. Home E61s are delivered with poorly tuned thermosyphons and require excessively long flushes.

-------------------- End boring narrative mode -------------------------------

Instead of thermosyphons or bolt to boiler designs, Bezzera now uses group head heaters, keeping design simpler and more compact.

#25: Post by **tekomino** » February 5th, 2012, 3:28 pm

sekihk wrote:My boiler started when pressure gauge readings dropped to 0.8 bar. The boiler would stop when the gauge readings reached about 1.3 bar.

And this is also huge spread. At 0.8 bars water temp is at about 200°F. At 1.3 bars its 224°F. In my experience boiler water temperature has big effect on brew temperature on levers and Strega is not different. This is 25°F swing you have and depending on when you pulled the shot in relation to boiler pressure, the brew temperature will be different and if you did not track the boiler temperature the brew temperature will appear all over the place and inconsistent... Methodology Peter, methodology

#26: Post by **farmroast** » February 5th, 2012, 3:51 pm

Just to add, my factory setting, 1.02-1.22 with a .2 deadband
The group heaters are nice for helping during warm-up.

Where is the temp profile at after a long idle no flush?

http://users.rcn.com/erics/ · #27: Post by **erics** » February 5th, 2012, 4:57 pm

Eric, do you have any thermocouple that could replace the factory thermo sensor used in the Strega brew head? I'd like to try PID the temperature of the brew head.

Yes. It would be done with a "baby" RTD having the same thread sizing (M3) as the group heater tstat. The temperatures on this grouphead vary, of course, depending upon where you measure and each location varies very similar to that shown below. These temperatures were measured at the very front of the group at the mid-point of the piston's stroke.

#28: Post by **allon** » February 5th, 2012, 5:19 pm

tekomino wrote:And this is also huge spread. At 0.8 bars water temp is at about 200°F. At 1.3 bars its 224°F.

You forgot to take gauge pressure vs. absolute pressure into account.
0.8 bar gauge pressure is 1.8 bar absolute pressure; the water is 242°F
1.3 bar gauge pressure is 2.3 bar absolute pressure; the water is 258°F
Still a spread of 16 degrees fahrenheit.

February 6th, 2012, 11:19 am

erics wrote: These temperatures were measured at the very front of the group at the mid-point of the piston's stroke.

Thanks for doing the work -- the group is actually cooler than I expected. This may explain the memory effect on shot temperature when doing multiple shots.

It's looking like Bezzera achieved a rough 90C to 95C range temperature stability; but that looking to get a more precisely repeatable profile with fancy flushes is going to be difficult.

#30: Post by **mtn hack** » February 6th, 2012, 5:48 pm

Ok, help me out here.

It seems to me that the forum's thoughts on the reasoning for the addition of the two cartridge heating elements are to a) minimize warm-up time and b) maintain group head heat between shots due to the group being separate from the boiler. However, I believe the other reason is to stabilize in-shot temperature, not the starting temperature which is solely the HX.

Once the pump quiets, you are done filling the piston chamber with pressurized HX brew water, and as soon as you cut the pump, your piston water begins to cool. I am sure there are many factors that would contribute to how much your piston water temperature changes, including ambient temperature, pre-infusion pressure and also puck density. I think Bezzera setting the t-stat on the cartridge elements to 95C just gives us the best starting point for the group's temperature, as long as there is sufficient recovery time between shots. As far as I have seen reading through all of these threads, no one has mentioned anything regarding the cartridge heating elements' thermal responsiveness. When do they kick on, 90C, 94.5C? Where is the cut-off; 95.0C, 97C? And most importantly, how does the group respond?

Therefore, the idea of PID controlling the cartridge heating elements would be good if you wanted to control the negative temperature drift during the shot, but IMO it is not a viable solution because the time required to make substantial and precise changes to the group would exceed your shot time. Not to mention, no one seems to mind this negative temperature drift. The only way to truly adjust brew temperature, would be to adjust the starting point via the p-stat... correct? It appears from the graphs provided thus far, the starting HX temperature is at or near the 95-96C range.

Also, it seems to me that the flushes on the Strega do not accomplish what they do on levers with the group attached to the boiler. Flushes on the Achille cool the group to HX temperatures, especially after it has been sitting idle for a while. With the Strega, it appears we are flushing heat exchange water (~96C) through the group head that is theoretically 95C. The problem with flushing, according to OP's graphs, seems to be that we can then introduce colder water into the group head because of this flush. Longer flushes may introduce so much cold water that it actually cools the HX tubing so much that we begin to see starting brew temps down around 90C. Similarly, if the boiler gets low during this process and calls for a fill, the effective boiler temperature lowers, and we see ridiculously low starting temps from the HX.

I think I am going to try the zero flush and see if I get any "hot shots". [Edit: If so, adjust the P-stat down until I can get starting temps that taste good... 92-93C? / ]

It seems to me, the Strega was designed to be a no-flush instrument.