Owner experience with Bezzera Strega - Page 90

Need help with equipment usage or want to share your latest discovery?
User avatar
doublehelix
Posts: 470
Joined: 9 years ago

#891: Post by doublehelix »

Wow! Very impressive mods and data--I think you'd have a rapt audience if you posted your findings.

Hide
Posts: 15
Joined: 9 years ago

#892: Post by Hide »

Hi everyone, I've just received my Strega, I am reading the manual and it prompt for first-start up to run the machine filled and turned on for 12 hours and drain afterwards, is this a recommended practice?

DanoM
Posts: 1375
Joined: 11 years ago

#893: Post by DanoM replying to Hide »

I don't think I did that procedure on my Strega, but I could be wrong. If it's in the manual though it is by definition "a recommended practice." It only costs you a few cents worth of water and electricity, so I'd follow the manual.

Have fun!
LMWDP #445

Hide
Posts: 15
Joined: 9 years ago

#894: Post by Hide »

I'd actually skip the step but went to flush out at least 1 reservoir worth of water from all outlet and went straight to make 2 shot just to find the grind setting and how does the strega works.

Strega is really need some time to get used to! (Coming from a ex-E61 user, dialing in is so much easier!)

User avatar
another_jim
Team HB
Posts: 13871
Joined: 19 years ago

#895: Post by another_jim »

radudanutco wrote: ... with a direct connected pump, the final ramp starts after cca. 7 sec and takes 4-5 sec, with a dimmer, the final ramp could start after 17-25 sec and is slower, up to a pressure which yes, depends of the dimmer setting ...
I have installed (on pump output), a pressure gauge for preinfusion, and also I've made some measurements with a portafilter with pressure gauge and a valve;
also, I have a PID installed on the group, and a lot of session temp profiles (HX output and group point 1), and some pressure profiles, for those interested;
Thanks for the info. If you can post time versus pressure graphs of the preinfusion, it would add some detail to what you've said. One question. Do you need the dimmer at all? If you have the gauge, can you stop the group fillup at, say, 3.5 bar? (the pump can be stopped by adjusting the microswitch so a move of the handle, while it is down, to the right, turns it off). This way you can emulate the shots you would get with the machine plumbed to the mains and operating with no pump.
Jim Schulman

User avatar
radudanutco
Posts: 147
Joined: 11 years ago

#896: Post by radudanutco »

another_jim wrote:Thanks for the info. If you can post time versus pressure graphs of the preinfusion, it would add some detail to what you've said. One question. Do you need the dimmer at all? If you have the gauge, can you stop the group fill up at, say, 3.5 bar? (the pump can be stopped by adjusting the micro switch so a move of the handle, while it is down, to the right, turns it off). This way you can emulate the shots you would get with the machine plumbed to the mains and operating with no pump.
Thank you Jim, I didn't notice that lateral movement of the lever; however, at my first attempt, I found a micro switch with no apparent adjustments with those two holes fitted exactly for those two screws; perhaps an extension plate...
anyway, I am thinking to put another switch on the pump control circuit (I already have one to short-circuit the dimmer, restoring the original function for the pump); all these mod's are located externally, on a metal plate fixed on the front left leg of the machine;

back to PI pressure - it appears to me as a flow rate problem;
the flow rate in a flush with full pump is around 10 ml/sec, while the dimmed pump has a max of 2.5 ml/sec (mid position, in my particular type of dimmer); and the tap water in my apartment has a minimum of 120 ml/sec;
during flush, the flow is almost at nominal maximum for the Ulka EX5 and the pressure is practically zero;

roughly the same flow rate is when filling the inside space in the group up to a point when water, steam and air start to oppose resistance and the vibratory pump is decreasing the flow while rising the pressure;
hence, with such a low flow rate, the main part of the pressure profile is at about 0 bar diff, and the final steep pressure ramp, up to the OPV set limit, or in the case of a dimmed pump, a slower rising profile, up to a lower than opv set pressure, as corresponding to the low flow at the start of extraction, or to the pump in a marginal status;

with a properly prepared puck, a non-mod pump would make a 7 sec "unpressurised" preinfusion, then a ramp PI of about 4-7 sec to full pressure, or less if the pump is stopped to a lower pressure;
a dimmed pump would make quite a longer, about 14-17 sec unpressurised PI, then a ramp of 15-25 sec to, say, 6 bar;

a plumbed Strega version should have, with the high flow rate from the network, a very quick fill-up (less than a second? just a speculated estimation), then a PI at 2-4 bar as stated;
also, a dipper architecture would provide a very quick fill up, then a PI at the 1-1.5 bar in boiler;
apparently, most part of the Strega PI is a kind of soaking, then a rather quick ramp PI; both a quite particular to this machine;

however, I didn't have notable results with the dimmed pump, or at least not at all better than by stopping during the PI pressure ramp;
currently, with the blend I made (city plus mostly, with longer development) the most appreciated shots are those made with full pump (that is, dimmer shortcut) stopped at cca. 2 bar on the ramp, then a 3-4 secs of the residual pressure kept, until the shower is covered, followed by a rather long extraction - 30-40 sec - until 42-45g in the cup, from 18,2 g dry;

I only have recordings for some simulated shots, with a DiY thermofilter having a higher than usual hydraulic resistance;
therefore, the values are less significant, only the profiles:


this is an example of a shot, as recorded for group in point #1and for the HX output:


this is a session temp profile:


and this are the overlapped temp profiles for the shots in a session:

User avatar
JohnB. (original poster)
Supporter ♡
Posts: 6579
Joined: 16 years ago

#897: Post by JohnB. (original poster) »

a dimmed pump would make quite a longer, about 14-17 sec unpressurised PI, then a ramp of 15-25 sec to, say, 6 bar;
No idea why you are seeing such extremely long fill times with the dimmer mod. My times ran a few seconds longer then stock. I also saw the ramp with undimmed pump start several seconds sooner then you are. Maybe the vibe pumps aren't all created equal or maybe there is difference between the 230v & 120v models.
LMWDP 267

User avatar
another_jim
Team HB
Posts: 13871
Joined: 19 years ago

#898: Post by another_jim »

Thanks for the graphs. I think my undimmed preinfusion also runs a few seconds faster; but it's in same the ballpark. I've replaced lots of Ulka pumps, and their pump curves are not all that consistent. Your temperature readings, on the other hand, are very consistent. My readings showed more variation. PIDing the group heater is definitely paying dividends.

I think Bezzera has done very well with its group heater designs; it's an inexpensive way of stabilizing temperatures. But your data shows they could do even better, without adding a lot of cost, if they replaced the mechanical stat with a PID.
Jim Schulman

User avatar
radudanutco
Posts: 147
Joined: 11 years ago

#899: Post by radudanutco »

another_jim wrote:...
I think Bezzera has done very well with its group heater designs; it's an inexpensive way of stabilizing temperatures. But your data shows they could do even better, without adding a lot of cost, if they replaced the mechanical stat with a PID.
I am also contemplating their reasons not doing so, yet;
I understand that there are anyway temperature compromises - for ex., a long pump cycle, a long preinfusion, means heat loss, not to be compensated in a couple of seconds by the heating element; I am trying to accept temperature profiles as being not so important here;

however, if the water from HX is coming with the same temperature profile while pump run (each time a recovered HX), it is not the same resulting shot profile with the group at 84C (203F) or at 87C (208F), as mantained by the thermostat control;
the thermostat sensor is placed to allow for a rather insulated heat source mode, maybe to allow for a certain degree of shot temp adjustment by operator, as thermostat and the group cycle is a order of magnitude greater then the shot duration; but this could be in certain condition counter productive: having heat on, while the group starts to be heated from the HX water - meaning inconsistency;
while group heater was not PID-ed:


I am lazy enough to postpone an 'experiment' (in the room I have now 88F...) with the PID sensor placed in the original hole of the thermostat;
now, the K probe is placed externally, on the group surface, in point #2;

User avatar
radudanutco
Posts: 147
Joined: 11 years ago

#900: Post by radudanutco »

(an 'addendum' to the previous posts:)

Strega has a group chamber rather large and, in the whole circuit, from the pump, via the check valve, HX and pipe connections, up to the group/piston chamber, there is no notable flow restriction;
so, after the initial slow start-up (1-2 secs) of a vibe pump, the flow of the pump is rising to almost its maximum value, about 10 ml/sec, quite similar with what happens in a flush;


for a notable interval, during chamber filling, the pressure (I measured it with a pressure gauge portafilter, and now, with a pump pressure gauge) remains almost zero (gauge);
including the initial start-up ramp of the pump in the first 1-3 secs, it takes 5-7 secs for the Ulka pump to fill enough volume to start meeting some resistance;
as flow decrease, the pressure will increase now - at the end - in a ramp;
for the first part of the pump run, the water was just accumulating above the puck and soaking into the upper layer of grind by gravity and by hygroscopic and capillary infiltration;
then the pressure increases gradually, as the water is penetrating an increasingly thickness of the coffee grind; then appears first drops under the screen etc.

if so, as above:
- only Strega among spring levers would have such a "hybrid" kind of preinfusion; longer no-pressure PI (filling the chamber), followed by a rather fast ramp PI;
- in both plumbed and dipper preinfusion, the flow rate is much higher than the one of a vibration pump, so what prevails is a real linear, low pressure PI;
- in the case of a dimmed pump with Strega, the flow rate is lower, filling the cylinder chamber lasts much longer, but only to have a wet upper layer of the puck;
the main benefit in the case of a dimmed pump seems to be a longer pressure ramp at the end; and also this slower ramp would allow for a more precise stop at the desired pressure; provided there is a gauge to monitor preinfusion pressure;
the zero gauge bar pressure wetting of the puck should have some effect, I suppose, but if there is not a significant one, the length of the dimmed pump PI until pressure ramp, would have not much relevance (?); which I am not sure!

I tried to combine both worlds of the Strega pump PI:
- lever down, full pump start, for a fast fill of the most of chamber volume;
- as soon as the pressure builds up, and manometer needle starts to move, I switched on a low setting dimmed pump, to have a longer (pressure ramp) preinfusion;
- stopped the pump at, say 3 bar, by clicking another switch, connected in series with the lever microswitch;
- it was the start of the extraction flow, so I raised the lever to free the springs and continued with kind of a small lever retention, mostly at the end of the extraction;

results? confusing, for me and my abilities, especially...
I tried something similar at the end (3 bar reached with dimmed pump), but keeping the dimmed pump setting all the way from the start (hence a longer interval for filling); all the others being the same; and it was a difference, with a more balanced, friendly result in the case of a full dimmed pump;
on the other hand, the best shots I made with Strega, were without dimmed pump at all, but stopping on the final ramp at 3 bar, short 2-3s pause, then the springs;

I am measuring the group temperature at the front/lower part of the cylinder surface, which is a rough estimate for the inside thermodynamics;
but as it is, it helped me to decide switching from thermostat to PID for the group heating control; and as in Bezzera Magica case (E61 group with Eric's thermometer), I found a useful complementary info in HX output temperature profile; (in E61 case, the HX output is reflecting the thermosyphon flow too);
when the pump is started, the hot water from HX pushed towards the group gives a fast rising profile, up to a peak at 110-115C (230-239F) then the temperature decreases as pumped cold water reaches HX;
this happens quite slowly in Magica, where the flow rate is for the extraction (1-2 ml/sec), but is much faster in Strega where the flow rate is for lever chamber filling (10 ml/sec);
the HX output temperature profile reflects the pump running duration, and any other change in flow value;
the HX output peak value indicates the level of HX recovery and
the HX lowest value reached, is an indication for the water volume pumped;


the group temperature is increasing on the pump run, being flat roughly while the flow is stopped (manual preinfusion), and is descending while there is a flow (during extraction);

ref mod's: IMO, the most significant are (1) PID on the group (still to find the best location for the probe), and (2) a presssure gauge on the pump, for the PI pressure; the first, for improving consistency, and the second, for a better feedback while a nonstandard approach (Jim's procedures);

still remains for me to wonder when is the best moment for stopping the shot (usually, by a scale, and a target extraction rate), but it is not so critical for me, because if working with early stopping of the pump, leaving the shot up to the end is affecting the extraction yield, but still for a good taste;

Post Reply