Looking for a Simple Measure of Preinfusion

Beginner and pro baristas share tips and tricks for making espresso.
User avatar
Team HB
Posts: 13824
Joined: 19 years ago

#1: Post by another_jim »

The Probem: More and more espresso machines are offering preinfusion. But each one does it in its own way. Some have a chamber that fills slowly, others turn the pump off and back on, others have power settings for the pump or delays before engaging the lever, and some have pressure profiles. So how does one describe a preinfusion in a few numbers that works for every machine?

I was contacted by members looking to create a public database app where people can store and compare shot data. This sounds like a great idea if the data entry is pleasant or at least painless. But the success of such cross-machine projects depends on machine independent ways of describing espresso prep. This is a big, juicy problem; so lets kick it off with preinfusion.

In the old days, before profiling, dwell-time, that is, the time elapse before the first drop appears, was a good measure. But on the newer machines, you can get a low pressure preinfusion that creates a slow drip over a long period of time. The ideal would be a graph of shot weight versus time; but that is hardly a single number. So we need something else.

My suggestion, to get the ball rolling, is time to 5 grams of shot weight. That would include the time the long, slow drip preinfusions take. But I'm sure the members can come up with something better than that.
Jim Schulman

User avatar
Posts: 3690
Joined: 15 years ago

#2: Post by Peppersass »

I think of "preinfusion time" as the time it takes for the basket to fill with water and become pressurized.

With my gear-pump modified GS/3, I lower the line pressure to 1-2 BAR and preinfuse with the pump off. I don't see drops on the bottom of the basket until the basket is full and begins to pressurize. There are two ways I can tell when preinfusion is over: 2) When drops appear on the bottom of the basket, and 2) When the GS/3 boiler gauge begins to rise from wherever it was when preinfusion started.

If I don't turn on the motor immediately, I call the next segment "preinfusion soak" time. Yes, drops are coming out, but the pressure remains constant at line pressure. Actually, I never do this, but it's a possible profile.

Once I turn the motor on, preinfusion is over. I call the time it takes to reach maximum pressure, nominally 9 BAR, "ramp time". I can control the ramp time with the gear pump motor speed.

On a stock GS/3 AV, and many other non-profiling machine, the pump is on from the start, so "preinfusion time" would be the time before pressure begins to rise, which is also the time before the first drops appear. Even with af flow restrictor (gicleur) inline, it's so fast that I don't think of it as preinfusion at all. "Ramp time" is determined by the pump and gicleur.

The stock GS/3 AV has a preinfusion feature that I never use. As I recall, it turns on the motor for a predefined amount of time, which probably should be the time it takes to fill the basket, and you can optionally program in a motor-off "soak time" before the "ramp time" begins. To me, this isn't really preinfusion because the flow rate when filling the basket is way too fast.

I believe "preinfusion time" on the Slayer is the same as my modified GS/3 -- time it takes to see the first drops or the first sign if pressure rising. AFAIK, the Slayer has no concept or capability to do a motor-off "preinfusion soak". Since the pump is on from the start, once the basket fills it's the start of the "ramp time", which is determined by the motor speed and needle valve settings, which are preset before the shot, not varied during the shot.

My sense is that first drops or first sign of pressure rise can be used to measure preinfusion time on most, if not all, pump machines. I've never used a lever, so that may be different.

User avatar
Posts: 2588
Joined: 14 years ago

#3: Post by AssafL »

My observation is that PI on a 18gr dose is when 28ml water has passed into the puck. Obviously the time it would take depends on volumetric pump speed/debit/lever piston depression.

PI timing may not be important as long as the puck drinks up the PI volume. To get it to fill up one may have a minimum time and that would depend on the grind. A finer slayer shot just needs more time to absorb a nearly equal volume of water.

Not everyone has accurate flow meters though.

So I like Jim's idea, as it creates a consistent platform for discussion. The 5gr is somewhat past PI, and is therefore very consistent demarcation point. It alleviates inconsistencies such as finding agreement on which drop constitutes first drop - what if one holds back a few seconds on powering the pump?
Scraping away (slowly) at the tyranny of biases and dogma.

User avatar
another_jim (original poster)
Team HB
Posts: 13824
Joined: 19 years ago

#4: Post by another_jim (original poster) »

The 5 grams is prompted by two things I have found on the Bianca.

-- The first confirms the general experience: If you use Turkish grind, e.g. from an EK43, and the puck isn't soaked at low pressure, the high pressure shot will choke. This adds one more reason to the usual ones for soaking the puck at lowered pressure, an easier route to high extraction and less channeling

-- The second is about preinfusion becoming akin to a steeping extraction. I found I disliked the taste of Turkish grind shots that preinfused longer than about 20 - 25 seconds; they tasted overextracted, even soapy. On the other hand, regular espresso grind preinfused for about 30 seconds, and as a "push it beyond sane limits thing", French press grind preinfused for 90 seconds, made perfectly good demitasse coffee. They no longer had the mouthfeel of espresso, and the flow portion of the shots were at very low pressure; but the mechanics are the same as making steeped coffee, longer steeps for coarser grinds. FWIW, these brews taste a lot better than Americanos, or demitasse strength coffee from mochapots, Clovers, or AlphaSteampunks.

In any case, regardless of grind used or preinfusion time, the drip, drip, drip, didn't add up to five grams in the cup.
Jim Schulman

Posts: 665
Joined: 9 years ago

#5: Post by walt_in_hawaii »

I've found the same thing, the long low pressure PI allows finer grind without choking and has all but gotten rid of my tamp... tamp is now more like level and polish more than pushing down. But 5g? I usually try to start the PI clock when I see 2 or 3 of the first drips into the cup... no idea what that is in grams, but feels like 1 or 2g? at that point i'm satisfied the puck is saturated.

User avatar
Posts: 2588
Joined: 14 years ago

#6: Post by AssafL »

So what does the end of PI signify: IMHO it is when the last milliliters of (compressible) air has been displaced by (incompressible) water - which enables flow.

If you don't let the PI continue to the end, the wet part of the puck (the top part of the puck), will act as a piston on the bottom part of the puck - the dry part. Being dry, the air (being compressible) will be pushed out and density will increase. Either the puck will be sealed, or water will be forced to find channels or, if the layer is thin enough, continue PI SLOWLY. Obviously SLOWLY depends on the size of the grinds. Boulders will be easier for the crushing water to permeate.

So how do we measure that PI is over (somewhat scientifically, or at least repeatedly)? By looking at flow rate. Once flow rate is up and slowly increasing, PI is over.

Jim's 5g is a good starting point as it basically assumes that at 5gr in-the-cup flow has been established. It is no longer drip-drip. Conversely, if it is still drip-drip at 5gr, you are really no longer espresso....

Is it 5gr? or is 4gr enough? I don't know. I am not sure it matters. As the % of time for the PI is significantly higher than the % of time to get to 5gr. That error is mostly negligible and if we all agree - is cancelled out between recipe users.
Scraping away (slowly) at the tyranny of biases and dogma.

Posts: 234
Joined: 7 years ago

#7: Post by maki »

AssafL wrote:if it is still drip-drip at 5gr, you are really no longer espresso.
may i ask - why?
LMWDP #630

User avatar
Posts: 1464
Joined: 11 years ago

#8: Post by happycat »

How about...

The period in which water is added to the coffee puck between when pressure is less than 6-9 bar until pressure is raised to 6-9 bar to create espresso extraction.

So if pressure too high at start, no PI
if pressure never hits espresso extraction, no PI just a coffee shot
LMWDP #603

User avatar
Posts: 2588
Joined: 14 years ago

#9: Post by AssafL »

maki wrote:may i ask - why?
Actually a good question. I don't really know. I guess I made an unsubstantiated assumption.

In my humble experience, having tried (more than once) to salvage good coffee ground too fine by letting it drip. I can't remember if it ever came out good.
Scraping away (slowly) at the tyranny of biases and dogma.

Posts: 1200
Joined: 7 years ago

#10: Post by JayBeck »

As I understand it, the Decent Espresso DE1 defines the end of preinfusion as pressure rises. Preinfusion is considered "successful" when water in and water out during the pour are the same (I.e. the puck is fully saturated from the preinfusion and cannot absorb anymore water).

Most profiles that have a long (10-30) second preinfusion set with a pressure limit of 4 bar. Once the machine detects the beginning of the pressure rise, the machine detects it so the timing of full pump pressure to compress the slurry occurs at the user defined pressure. So the machine automatically detects this for you, being the automatic machine the DE1 is.

Note: if doing a Slayer shot, this pressure threshold is 1 bar for the end of preinfusion (but the flow rate is 1-2ml/s instead of 3-4ml/s so it takes 45 or so seconds for pressure to build).

So that's the standard way. In advanced shot mode, you can mimic anything, including adding in dwell times like Jim is doing with the Bianca where he cuts off flow. The DE1 comes stock with the "Scott Rao" shot with a looooong dwell (30-60 seconds after a fast preinfusion) and then pulling the shot. Scott got a 27% extraction using this profile on his DE1 with Baratza Forte.

TL;DR - the DE1 uses a pressure threshold (usually 4 bar) as the end of preinfusion. With advance editor, dwell times and steps can easily be added prior to full pressure and any decline at the end.