Is there a too-fine grind barrier, really?

Beginner and pro baristas share tips and tricks for making espresso.
GDM528

#1: Post by GDM528 »

Curious if others have observed the following:

Espresso shot seemed to run a little too fast, so grind a little finer next time. Concurrently, I've been experimenting with longer preinfusions (up to 20 sec post drip), and lower-pressure (6ish bars). Sometimes those experiments would cause the shot to run faster, so I adjust the grind a little finer... see where this is going? Now my espresso shots never choke; in fact I have to throttle back the flow/pressure to keep the shot time within a 20-30 second window, even as my grinder approaches burr-chirp.

I'm now at a point where it's my grinder that's choking, not my espresso machine - cool. The grind is not-quite-but-approaching a flour-like texture, and I've been slow-feeding the beans into the grinder to (theoretically) minimize fines and suppress the burrs from seizing up.

I don't think I'm channeling, as I insert filter paper disks in the puck top/middle/bottom and they create telltale signatures in the paper of any channeling, which I don't see. The filter paper is from AeroPress filters, which I think are designed to not be prone to clogging up.

Have I managed to navigate through some sort of shot-pulling 'choke barrier'? Perhaps I've developed a set of compensating behaviors to pull a decent-tasting shot from just about any grind. However, I don't think that's a bragging right, as I ponder if that may be leaning towards exaggerated bitterness and lost access to other flavor notes.

I suspect the crux may be in the thorough soaking the puck gets before pulling the shot: 3-4 bars for about 20 seconds after the puck has wetted through - sorta like a "blooming espresso", but with pressure applied. The first 10g is crazy-dark and there's a pretty sharp blonding between 1.5 and 2:1 ratio. The pucks that knocks out is soft but not gooey.

Has anyone else found themselves here, and did you leave or stick around to explore the space?

Nunas
Supporter ♡

#2: Post by Nunas »

Interesting post. Somewhere in these pages, you'll find discussions among the early adopters of flow control devices on e61, among whom I count myself. I don't recall anybody referring to a barrier, but long low-pressure preinfusions/soaks, resulting in the ability to extract from much finer grinds, was discussed. I began to push this envelope a bit, but didn't find the espresso to be to my liking, so I reverted to my previous technique.

iyayy

#3: Post by iyayy »

yes.
at least i found that some beans takes fine grinds better than others.
you can easily test this by underdosing. ur puck prep needs to be very good and very level.

common indicator of a limit that i hit is that it gets extremely bitter even at 0.8 ratio, despite puck still looks dark.
i have yet to choke my grinder with ssp mp, but there are beans that i did settings 1 before chirp. 0.5 doesnt feed well and takes too long so i avoid it. it'll also likely just bleeds through vst basket. for those kind of bean i updose more than 10g.

below is a geisha beans i tried.
df64, ssp mp, grind settings 2 before chirp, 10g in vst15, 8.6 bar shot, 13s preinfusion up to 4bar. 1:1.2ratio.

the coffee has a lots of fines that passes through the basket, and even fresh beans barely have any crema. straight is super intense, needs about 1:1 dilution to make out proper taste and flavor layering.
easily cuts through 1:8 as is milk while still retain flavor complexity.
biscuit puck ftw~



GDM528 (original poster)

#4: Post by GDM528 (original poster) »

Nunas wrote:... but didn't find the espresso to be to my liking, so I reverted to my previous technique.
Exactly what I'm wondering about. Do you recall any details about what you didn't like about it?

Jonk

#5: Post by Jonk »

I have experimented some with extreme underdosing in a Robot basket. First with a Niche; got bitter, unpleasant shots (a few numbers above 0 though).

More recently, using the 64mm SSP cast v1 I needed to grind just above chirp for light roasts to get enough resistance for a 1:2-3 shot. I believe it's finer than what I used on the Niche, very much approaching flour. A normal dose with that setting and a medium roast turned into clay when I tried to pull a shot :lol:... and underdosed light roast just tasted strong in a bad way, with no defined flavor. Much preferred a slightly coarser grind for an allongé.

So for me, yes, there seems to be a limit to how fine I can grind and still like the resulting espresso.

GDM528 (original poster)

#6: Post by GDM528 (original poster) »

iyayy wrote: the coffee has a lots of fines that passes through the basket, and even fresh beans barely have any crema. straight is super intense, needs about 1:1 dilution to make out proper taste and flavor layering.
easily cuts through 1:8 as is milk while still retain flavor complexity.
biscuit puck ftw~
Wow, that puck looks like clay! Respect!

The filter paper I put in the bottom of the basket very effectively keeps fines out of my cup. However, when I was experimenting with making a Turkish-ish coffee using my espresso machine, I found that grounds in the cup actually improved the flavor: "Turpresso"? Turkish coffee with an espresso machine.

I too have been noticing a gradual reduction in crema as I tailspin into fine-grind oblivion, but thought my roasting was the culprit. My tiny-batch roaster has allowed me to tweak my roast every 2-3 shots, and I've found myself tweaking my roasts to improve grind behavior.

As you point out, the intense shots are well suited for milk drinks, which is what I mostly prepare. Perhaps this is the right path for caps and lattes?

iyayy

#7: Post by iyayy »

GDM528 wrote:Wow, that puck looks like clay! Respect!
thank you.
The filter paper I put in the bottom of the basket very effectively keeps fines out of my cup. However, when I was experimenting with making a Turkish-ish coffee using my espresso machine, I found that grounds in the cup actually improved the flavor
yep. it seems to add layers in the flavor, altho i dont find all beans actually taste well at this level.. i seem to hit the extreme bitterness on some. i think some beans arent porous enough and actually ends up restricting flow instead. for those the puck usually looks very pale on top and very dark at bottom, with noticeable color gradient if u cut it.
i then started down dosing to see if i can make the water exit faster, hence how i ended up with extreme fine grind and underdosing my vst15 to 10g.
As you point out, the intense shots are well suited for milk drinks, which is what I mostly prepare. Perhaps this is the right path for caps and lattes?
maybe? i mostly do lattes too. and my original goal was to see if i can efficiently maximize use of the whole puck.. instead it lead me to very tasty and enjoyable desert.
this method seems to always brings out the floral notes, even on those stubborn beans. i really enjoy this lingering fragrant aroma post sipping the latte.

recently i tried bleeding water for end of shot (pseudo flow profiling), so maintain the same start method, but just add another step to slowdown the flow.
this seems to suprisingly round out the sharpness and produce more drinkable shot while maintaining most of the intensity. i didnt have to change my milk ratio.

i'm still practising and eyeballing the pressure gauge. am considering diy the knob into a lever for easier control.

GDM528 (original poster)

#8: Post by GDM528 (original poster) »

iyayy wrote:i then started down dosing to see if i can make the water exit faster, hence how i ended up with extreme fine grind and underdosing my vst15 to 10g.
I found the same thing with underdosing, only now I'm wondering if I've unintentionally/organically figured out how to do the same thing with 18g in a VST 18g basket. Finer grind, higher water temp, longer soak, lower pressure, and no sudden ramps.

My theory:

1) When I open the valve on my GCP, the initial few grams of water are hot, 96-98C, so, higher solubility.

2) As the water slowly infuses into the puck, several grams of the coffee dissolves into the solution, effectively reducing the 18g dose to maybe 15-16g. I've been concurrently slow-dosing my grinder, which I suspect is also increasing solubility.

3) As the puck continues to soak in its own juices with very low pressure and a drip-drip flow rate, so do the gases: air trapped around the grounds, CO2, and any other gasses released <-- This is one of my key concerns, that these icky gasses are getting infused into the brew rather than being quickly pushed out during a more conventional pull. This reminds me of a recent research paper that asserted that an optimum shot is closer to 15sec, very fast.

4) By ramping pressure very gradually, I avoid the grind packing into the bottom of the basket and clogging the flow. I'm gunning for the Nobel Prize for creating a caffinated colloidal superfluid...

I even tried updosing from 18g to 19g in the same basket - no problemo, invincible. Upside is I can work with any grind level, but downside is now I have to rely on other signals that my grind is too fine (like, my grinder freaking out). I suspect the above process may also be extra-sensitive to the roast level, post-roast resting time, and bean origin/processing/age. I'm using this workflow with year-old ASD processed greens roasted to Full City, and experiencing sour notes I wouldn't otherwise expect under these conditions. So, this thread may simply be another installment in my service as a warning to others :)

GDM528 (original poster)

#9: Post by GDM528 (original poster) »

iyayy wrote:i mostly do lattes too. and my original goal was to see if i can efficiently maximize use of the whole puck.. instead it lead me to very tasty and enjoyable desert.
this method seems to always brings out the floral notes, even on those stubborn beans. i really enjoy this lingering fragrant aroma post sipping the latte.
Have you described this workflow in a separate thread? It sounds like something I want to try myself.

iyayy

#10: Post by iyayy »

thats some cool theory you have there.
GDM528 wrote:Have you described this workflow in a separate thread? It sounds like something I want to try myself.
havnt really discussed it.
in some threads already a mere mention of 9 bar shot with light roasts gets attacked with it being bitter old unproven standard while 6bar is the new norm and correct.
i got tired of these 6bar zealots.
personally it feels like an excuse or escape because of frustration from failing to use a unimodal grinder for 9 bar espresso.

my coffee background starts with filter, so most of the roast i bought are light and filter. its normal to find these beans taste boring as espresso, but magnificient as brews.
i hoped to have that taste in lattes, because its not impossible to have the complexity from filter + milk, but its also too diluted as a milk drink (mostly low fat/skimmed only). however i always found it to easily best any modern type shot from my bdb. i actually gone up drinking from 50/50 brew/espresso to 90% brew for awhille because of ssp mp. and the better i got with my espresso, my brews becomes even better and consistent.

so i started to attempt 9 bar shot. with ssp mp, the settings usually choke the machine instead, and a flowing grind will speed up too fast, failing to reach 9bar. adding pi helps to prevent choking while still reaching 9 bar.

fyi bdb allows min 55% pump preinfusion, which in my case actually rise up to 5.4bar with blind basket, and 8.9 (opv limited) at full. bdb has a gauge that reads grouphead pressure, so at least i can work out whats happening to the puck.

a few issue is
1. the pressure rise during pi is very slow, it takes quite long to fill head space and starts building pressure. i feel it takes too long to fill headspace and preextracting the top puck too much. no idea how to fix this without flow profiling.
2. its always a rising profile until it hits max pi pressure of 5.6bar at my settings. i cant do a static pressure preinfusion to saturate puck without constant compression, slowing of flow and eating the top puck.
i have found underdosing to somewhat help counter this.
3. theres an easily identifiable difference if i stop pi during the rise and full pressure ramp starts when gauge reads 2bar, 2.5~3bar, or 4+ bar. 2 is usually complex, 2.5+ is very bright acidic, 4+ is more balanced but can be less sweet vs the earlier. pressure usually starts rising at 10s, with 12~16s to reach those numbers, and bottom basket will be somewhat saturated by then, but not yet dripping. there is still much that can be tested or used for tuning here. i havnt even tried to figure out flow rate yet.

these are machine characters. urs may be different in shot pulling, but i'd figure if the shot character is similar, the result should be too.

also dose was kept, grind was adjusted to ensure all shots reach min 8bar, otherwise the profile of the shot would not be similar in how it flows. i feel keeping the behavior of the shot would be a best representation of adapting the settings correctly. this should be a clearer target method to achieve rather than just individual numbers.

yield was also adjusted by swapping cups to get maximum sweetness and complexity, usually ending with 1.2~1.5 ratio yield. more yield will tend to slighthly reduce the flavor, adds more aroma and aftertaste, but those aftertaste may have a hint of dryness. going further will eventually loses the character and becomes too diluted to taste, with bitterness being the most obvious.
if i use my 1zpresso k, it'll be closer to 1:2, but never as sweet as ssp mp.

also if pi is short, then shot will take too long and extract more at top and less at bottom, not a good balance at all. too long pi and it will start flowing and not reach max pressure. for this cases underdosing (+ grind tightening to maintain profile) usually works to gives more even extraction top to bottom.
similarly if grind isnt fine enough it will actually start flowing early and i wont get full 8.6bar (flowing) at full pressure ramp.

for getting the acids to blend well with milk, so far i found that i need to reach at least 8 bar, even if just for 1-2s. reaching max 8.6 (at least, on my setup) would be ideal. so maybe theres more rules, but this has been a definite and defining one in my experience. coffee:milk ratio are commonly 1:6 ~1:9, depends on how milky u prefer it.

else it will likely fight the milk and turns into sour mess, especially when it starts cooling. my target is to get the drink work at any temp, otherwise i feel like im not cooking it right. hot drinks or food tends to mask defects anyways, and most flavor and aroma are at peak when its just below hot, but still far from warm.

so about grind limit, some beans i didnt hit it when dialing with these method. for those it will become sweeter with tighter grind and more floral with longer pi. but on some, when continue to down dose and go finer it starts to build a bitter profile that exist in any yield, even very small ones, and only goes more bitter as i went finer. for those i will reverse to go coarser and start updosing to maintain the flow behavior.
to have enough grind headroom, i play between 10‐15g on my vst, but i usually starts at 12g for new beans. my range has been mostly 1.5~5 on df64 ssp mp. i adjust about 0.5 for each 1g dose.

i also usually serve this drinks to my sister in law and niece, they tend to be very picky on food. they wont eat non premium donuts at all.
but so far i've only been getting praise with my coffee, and that's good enough to convince myself that im doing something right.

interestingly my attempts and ideas that came up at trying to optimize the shot and puck efficiency seems to suggest that im heading towards a plumbed/dipper type spring lever profile. perhaps the londinium shot isnt just a favorite profile. i'll likely go down this route later on.. if i have budget :roll:

sorry if im repeating a few things, im just not sure how to write and pass the understanding properly. if you attempt it then good luck.