My long and rambling path to preinfusion/pressure profiling - Page 11

Need help with equipment usage or want to share your latest discovery?
User avatar
Jake_G (original poster)
Team HB
Posts: 4295
Joined: 6 years ago

#101: Post by Jake_G (original poster) »

AssafL wrote: Oh - and Jake - it sort of does prove your analysis... sort-of-ish. :wink:
Hey, for an analysis based entirely on the premise of -ish, I'll take that endorsement as scientific proof!

Thanks buddy!
LMWDP #704

User avatar
AssafL
Posts: 2588
Joined: 14 years ago

#102: Post by AssafL »

**edit #2**
Just read your update to Chimera. Noisy flow meter may vary well be an overactive PID loop, eh? :P
Drop you kp way down and kill kd and ki until you get a nice coil spring oscillation output and then temper that by adding kd to limit overshoot. You will then have a steady state error that you can nix with ki. Then it's iterative to bring kp up, while keeping the response in check with the other two until you're satisfied with the response speed.
Jake - yes... I did the above process with a sealed puck (blind basket). Worked like a charm but the params did not work given a porous puck. I was too low and the pressure did not build up.

My problem is rather silly - I don't like wasting coffee. So maybe the next step is to setup the "scace like" device at a flow that is the average of a flow (now that the Digmesa is calibrated) and tune the PID using the Scace. If then it still misbehaves at least I'll have a closer set of params.

One point about stabilizing the loop - the sort of flow rates are borderline (almost clogged - but not), and micro-channels seem to open up and close as the pressure sensor shows - so it may be sensitive to exact flow rate. And it may be hard to stabilize it for all flows and all pressures.
Scraping away (slowly) at the tyranny of biases and dogma.

User avatar
AssafL
Posts: 2588
Joined: 14 years ago

#103: Post by AssafL »

Jake_G wrote:Aren't you glad you keep me around to bug you? 8)
Check my edit #2 above your post for notes on PID tuning. I've used this process many a time to get good results on many different systems.

Cheers!

- Jake
Yes indeed. I find I learn more from mistakes than from being right...

Now onto my next mistake. Where is it I wonder? Please bug more.

PID calibration over the next few days - and then start tackling flow profiling. That would be fun!

BTW - I really wish I could say Chimera leads me to prefer a specific pull mode - or at least be able to say "this type of pull profile for this coffee". I now alternate between the two manual modes: Slayer_PI&PID_PP and Manual_PWM. I can't say any method results in substantial differences.
Bean sourcing and roasting profile make much more of a difference!
Scraping away (slowly) at the tyranny of biases and dogma.

Pino
Posts: 117
Joined: 19 years ago

#104: Post by Pino »

At this point, water begins accumulating on top of the puck. I had always envisioned that the water comes down like a piston, pushing the air out of the headspace through the puck,...
I may have missed something yet the headspace (space between ground coffee and the dispersion screen) above the puck I assume is there intentionally to allow for expansion of the coffee as it becomes wet. So how does the expansion of the puck fit into your description? Or does it fit in?

User avatar
Jake_G (original poster)
Team HB
Posts: 4295
Joined: 6 years ago

#105: Post by Jake_G (original poster) »

Head space is definitely there to allow for puck expansion.

There are two primary schools of thought regarding this and I don't think they are mutually exclusive.

The first concept is that when the shot comes to an end, the abrupt drop in pressure causes the puck to swell. This has been verified in glass portafilter all motion videos, so I think it's safe to say this does happen.

The second concept is that the coffee blooms when given the opportunity to do so with very gentle and prolonged preinfusion. I suspect this is happening as I run into major channeling if I over dose when doing slow preinfusion from the puck hitting the screen and fracturing.

I think a fair amount of this blooming/expansion early in the process is likely canceled out when the puck is compressed by the rising pressure. Likewise, I think traditional espresso shots that hit the puck hard within a few seconds of starting the shot either don't allow much expansion at all as the puck is compressed before the grounds can swell, or the grounds are compressed as they swell. Either way, I suspect the result is that excessive up-dosing is more tolerated by shorter preinfusion...
Pino wrote:So how does the expansion of the puck fit into your description? Or does it fit in?
My short answer is that I certainly think it happens but I don't think it effects the dynamics of pressure and flow much, but I could definitely be wrong! What are your thoughts on this? One thing that definitely does happen is that if you dont have any room, the dispersion screen will provide a super-tamp when you lock the portafilter on and you will need to grind coarser to compensate.

Cheers!

- Jake
LMWDP #704

User avatar
redbone
Posts: 3564
Joined: 12 years ago

#106: Post by redbone »

Pino wrote:I may have missed something yet the headspace (space between ground coffee and the dispersion screen) above the puck I assume is there intentionally to allow for expansion of the coffee as it becomes wet. So how does the expansion of the puck fit into your description? Or does it fit in?
Without the clearance that allows for puck expansion you would get 1) Dimension or no water flow resulting in negative pressure within the group and coffee and grounds pull back. This would cause P.F. sneeze of such as levers and pump machines without a three-way solenoid valve. 2) Dirty shower screen 3) Increased strain on pumps or springs. 4) Makes for a dryer puck post shot.
Between order and chaos there is espresso.
Semper discens.


Rob
LMWDP #549

User avatar
AssafL
Posts: 2588
Joined: 14 years ago

#107: Post by AssafL »

Pino wrote:I may have missed something yet the headspace (space between ground coffee and the dispersion screen) above the puck I assume is there intentionally to allow for expansion of the coffee as it becomes wet. So how does the expansion of the puck fit into your description? Or does it fit in?
My hypothesis is that the difference between low flow (ie slow) PI and quick PI is the depth of water penetration in the puck before the puck compresses. Obviously, the slow PI is a fully wetted puck (perhaps partially or fully bloomed) whereas a quick PI is a partial bloom/wetted coffee at best.

Why the compressed layer of dry coffee matter is hard to penetrate I don't know. But it is.
Scraping away (slowly) at the tyranny of biases and dogma.

User avatar
AssafL
Posts: 2588
Joined: 14 years ago

#108: Post by AssafL »

AssafL wrote:Why the compressed layer of dry coffee matter is hard to penetrate I don't know. But it is.
Actually I have a hypothesis. We learn that coffee fractures during roasting and has fissures and Has various shapes and sizes of air pockets in it. Illy has microscope pictures of this.

Now, remember that water doesn't compress but air (and other gasses) do - and quite a bit.

So PI coffee that blooms, at least some of the fissures get filled up with Water. So when pressure is applied this wet coffee retains it shape better than any dry coffee below it.

Conversely, The fissures in the dry coffee matter should collapse under the 9bar, increasing the density of the dry puck and virtually sealing the puck.

That is my hypothesis.
Scraping away (slowly) at the tyranny of biases and dogma.

ira
Team HB
Posts: 5497
Joined: 16 years ago

#109: Post by ira »

AssafL wrote:One point about stabilizing the loop - the sort of flow rates are borderline (almost clogged - but not), and micro-channels seem to open up and close as the pressure sensor shows - so it may be sensitive to exact flow rate.
As I recall, John from Decent also mentioned this and commented that he had adjusted the rules to back off pressure to let the channels close as soon as he sees them happen. I think he also said that made a significant difference in shot quality, shot consistency or both.

Ira

User avatar
AssafL
Posts: 2588
Joined: 14 years ago

#110: Post by AssafL replying to ira »

I think the volumetric aspect of gear pump also does this. Unlike a rotating vane pump which is all to happy to dump water onto a microchannel (making it into a macrochannel ;) ?) The gear pump has very little spare water available in the gear teeth.

Actually - that is one VERY good reason to choose 4mm rather than 9mm teeth...
Scraping away (slowly) at the tyranny of biases and dogma.