Pressure and temperature: how they affect flavor?

[LeoZ]

i think ive reached intermediate level espresso making. lol..

ive been experimenting with different brew pressures, and it seems that there are pluses and minuses to each end of the pressure spectrum. my machine started at 11bar, which brought out quite a bit of the deeper flavors, but every shot seemed way too bitter for me. dropped it to 9 bar, and a great improvement, but still occasional bitterness. not sure if this was from the bean or my technique though. so, reduced to 8 bar, and zero bitterness, but flatness has started overtaking the taste components.

the biggest thing i noticed when dropping to 8 bar was the temp, and i wonder if the flatness is more from this than the actual pressure. at 8 bar, i still flushed the same ~6oz of water, which i think was way too much. I did a quick pre-work test this morning and didnt flush after a 20 minute warmup. the shot had much more body to it! it was still lacking a bit of the more complex flavors though.

maybe 8.5 bar is the next step? do you really notice these differences in pressure changes, or am i becoming insane from caffeine?

also, i know pressure and temp are directly related in closed systems; so does typical logic apply here? less opv pressure = cooler brew temps?

[another_jim]

Changes to the taste in the 7 to 11 bar range are fairly subtle; one needs good technique and a very familiar blend to be able to catch them easily. Increasing potency in all flavors as pressure increases has been reported by researchers, and this squares with my observations on the Tea.

There should, in theory, be no relation between brew pressure and temperature unless the volume of the shots you are making, or the amount you are flushing, changes. The OPV adjustments do increase the flow from the pump, but the increase is prior to the water being heated, so should not effect the temperature. Of course, "in theory" means if nothing idiosyncratic is happening in your machine. There's no way of telling unless you measure the temperatures.

[cannonfodder]

Versalab makes (or will be making) an espresso machine with a panel adjustable brew pressure. I have played with some brew pressure changes, but popping the side off the Faema to adjust the rotary pump is not a quick adjustment.

Some report a discernable change with as little as a .5 bar adjustment in pressure. To truly optimize brew pressure you would have to limit yourself to a single blend. I rarely use the same blend two weeks in a row so I run a generic 9 bar pump pressure. You would also need world class barista skills to keep your intra shot variables minimal to realize the fine tuning.

[timo888]

There's an interesting discussion of pressure in the Illy book (ch. 8 passim). Illy refers to the "time-dependent geometry" of the puck, citing a thesis by G. Baldini, Filtrazione non lineare di un fluido attraverso un mezzo poroso deformabile (Firenze, 1992). Illy writes:

Pressure on the cake does not remain constant during the percolation process, but varies as a function of the characteristics of the hydraulic circuit above the cake. This can be verified by continual pressure sensors: the recordings yield a function p = p(t) which exhibits an initial transient rise, increasing more or less slowly. The dry bed still lacks adequate cohesion and is susceptible to resettlement; therefore the initial phase appears to be decisive in percolation. After wetting, ground coffee particles swell up firmly interpenetrating each other, and percolation can follow its stationary course. This effect explains why...the flow is not constant during percolation, rather showing an 'asymptotic flow dependency'. .... Hydrodynamic experiments demonstrate that the average flow ceases to depend linearly on pressure in the proximity of 9 atmospheres and that an increase in pressure actually causes a decrease in the average flow....

(emphasis added)

Regards
Timo

[cannonfodder]

I have noticed something similar on my manual lever machine. If after preinfusion (10 seconds lever up) I exert a fast and forceful pull, the puck seems to seize up. The flow rate drops and the pressure needed on the lever for extraction goes up substantially. If on the other hand I use a slower, gentler pressure ramp it takes less pressure to pull the same shot.

I also added a delay on make relay on my pump machine so I could play with preinfusion using the mains pressure. While dialing in the preinfusion timing, I would prepare my shot as usual, lock in and engage the group. After the preinfuse cycle I would pull the portafilter off and knock out the puck to see if the water had fully and evenly penetrated the puck.

During one session, after I removed the portafilter, the puck actually swelled and heaved up in the center as the surface and sides absorbed water with the center base still dry. It was interesting to watch, the puck really does absorb a substantial amount of water and any unevenness in the absorption will defiantly cause a disruption in the puck. I would surmise that there is a minimum preinfuse pressure needed to keep the puck from swelling and channeling. To low a pressure and the top layers will swell and separate from the lower dry layer. I would imagine that a long preinfuse would cause similar problems.

[timo888]

A very good experiment and observation. We await the photo-essay

There's an optimal pre-infusion volume and pressure and time. Too little or too much of any of them and you get a cake with a dry bottom, or a cake with channels, or a cake so tightly swollen it's hard to get the brew water through it.

But if I'm going to err, I prefer to err with too long a preinfusion, especially if my water is on the cool side and I've not overdosed the basket, so there's room for the cake to expand.

Regards
Timo

[popeye]

Adjustable brew pressure on my zaffiro. Yea, i had to drill through the case, and my OPV to do it. Just began playing around, but for the hacker in me, it was a worthwhile mod. I primarily play with temp, but after i get the temp where i want it, I can play with pressure. But then after i get pressure where i want it, i go back to temp...Yea, and i'm trying not to drink too much caffeine.

[another_jim]

Love the garden faucet handle; combined with the PID panel it's either surreal or post-modern

[LeoZ]

There's an interesting discussion of pressure in the Illy book (ch. 8 passim). Illy refers to the "time-dependent geometry" of the puck, citing a thesis by G. Baldini, Filtrazione non lineare di un fluido attraverso un mezzo poroso deformabile (Firenze, 1992). Illy writes:

Pressure on the cake does not remain constant during the percolation process, but varies as a function of the characteristics of the hydraulic circuit above the cake. This can be verified by continual pressure sensors: the recordings yield a function p = p(t) which exhibits an initial transient rise, increasing more or less slowly. The dry bed still lacks adequate cohesion and is susceptible to resettlement; therefore the initial phase appears to be decisive in percolation. After wetting, ground coffee particles swell up firmly interpenetrating each other, and percolation can follow its stationary course. This effect explains why...the flow is not constant during percolation, rather showing an 'asymptotic flow dependency'. .... Hydrodynamic experiments demonstrate that the average flow ceases to depend linearly on pressure in the proximity of 9 atmospheres and that an increase in pressure actually causes a decrease in the average flow....

(emphasis added)

Regards
Timo

does this apply to preinfusing? i can see what cannonfodder is saying - that makes sense. but if the puck is preinfused, and a linear ramping of pressure is applied (how most e61s act), i dont see flow becoming an issue, only bitterness!

[cannonfodder]

Then you get into the water debit, how much water the group jet flows in a given time. Others are more qualified to speak about it than I.

The faucet is a nifty idea. When I do make a brew pressure change, I will go downstairs and adjust my regulator. One of the benefits of a rotary pump, brew pressure is directly affected by input pressure. So I can make an adjustment without having to open the side panel on the Faema. Unfortunately, that then changes the preinfusing timing because of the reduced or increased line pressure. I have the delay relay under the drip tray so I can make changes on the fly.

If the pressure is increased in a short ramp, the sudden build of pressure seams to compress the puck even tighter and reduces the flow rate. At least that is what I have noticed on my lever machine, which is easy to adjust brew pressure on. Just pull harder/softer. With a slower linear ramping, the flow appears to be faster and requires less pressure, I also got a better tasting shot.

From what I have seen, most E61's have a slow pressure ramp which makes them more forgiving. The Elektra being one that stands out in my mind as a fast pressure ramp. Whether you are running a pour over machine or a direct plumb will make a difference in the speed of the preinfusion as well. The slower linear ramp allows for a more uniform preinfusion and extraction in most cases (IMHO). However, Dans experience with the Elektra A3 and its rapid pressure ramp were enlightening.

Your original post question was 'less opv pressure = cooler brew temps?' no. Can you fine tune a machine's pressure to get the optimal flavor from a specific blend, yes. That is what a commercial cafe does. Since they only run one blend, the dose, temperature, and pressure are profiled for that specific blend. That is why it is nearly impossible to get the exact same cup from a home machine (or commercial machine at home) as the cafe.

[cannonfodder]

A very good experiment and observation. We await the photo-essay

There's an optimal pre-infusion volume and pressure and time. Too little or too much of any of them and you get a cake with a dry bottom, or a cake with channels, or a cake so tightly swollen it's hard to get the brew water through it.

But if I'm going to err, I prefer to err with too long a preinfusion, especially if my water is on the cool side and I've not overdosed the basket, so there's room for the cake to expand.

Regards
Timo

A video would be more telling. At first I thought I was seeing things. The puck slowly heaved up in the center to the point it was nearly even with the top of the basket. Then it cracked right down the center revealing the dry puck underneath.

I have noticed some apparent horizontal channeling while dialing in the preinfuse delay. After the extraction I would knock out the puck which would separate right down the middle. Almost like it was dosed and tamped, then dosed and tamped again. A very clear lairing between the top and bottom half's of the puck.

[another_jim]

does this apply to preinfusing? i can see what cannonfodder is saying - that makes sense. but if the puck is preinfused, and a linear ramping of pressure is applied (how most e61s act), i dont see flow becoming an issue, only bitterness!

Andy and I have been discussing this a bit. He's experimenting with a pressure profiling pump, and once the puck is preinfused, it responds "classically," i.e. less pressure = less flow. On the unregulated, no preinfusion Elektra Semi, one gets extreme Illyesque behavior -- an initial squirt, then a chocking off to dribbles, with classical flow only returning in the second half of the shot after the puck has settled.

Preinfusion certainly eliminates the non-linear, out of equilibrium phenomena that the Illy chapter describes. However, for overall shot timing, one still has to coarsen the grind when one raises the pressure. Perhaps the high pressure has more of an effect moving the fines into a blocking zone than in punching through them.

[erics]

My only comment before I go re-read Chapter 8 in the Illy book is that, to me, it would make a taste difference IF, after making adjustments to the OPV on a vibe pump or the relief valve on a vane pump, you then adjusted the grind "fineness" to cohabitat with that new pressure such that a 60 ml ± shot is produced in 25 seconds ±.

I.E. keep the bean type constant, keep the dose constant (weight), keep the tamp constant, keep the brew water temp constant (as best you can) and keep the shot volume/time constant.

I understand that this is one heck of an iterative process, and yes, unless you are modifying parameters of that new Versalab via the keyboard of your connected laptop, some temporary caffein related headache is bound to appear.

Eric S.

[King Seven]

Love the garden faucet handle; combined with the PID panel it's either surreal or post-modern

Post of the week for me! Made me chuckle...

[LeoZ]

i guess OPVs are different on each machine? i cant see an easy way to get to mine. outside of removing panels, theres a bolt with a water line connecting to it.
:/

[cannonfodder]

I.E. keep the bean type constant, keep the dose constant (weight), keep the tamp constant, keep the brew water temp constant (as best you can) and keep the shot volume/time constant.

Eric S.

And there in lies the conundrum. Not many home baristas are consistent enough in their technique to eliminate all of these variables. After a few years of practice, I am still not there. The problem is that the average home user pulls 3 or 4 doubles a day. A professional Barista pulls several hundred in a day. If I had a local cafe worth bothering with, I would even work for free every now and then just for the practical experience.

[LeoZ]

And there in lies the conundrum. Not many home baristas are consistent enough in their technique to eliminate all of these variables. After a few years of practice, I am still not there. The problem is that the average home user pulls 3 or 4 doubles a day. A professional Barista pulls several hundred in a day. If I had a local cafe worth bothering with, I would even work for free every now and then just for the practical experience.

LOL. thats crossed my mind on more than one occasion, for more than one trade!

re: the conundrum, im to the point now, where if a shot comes out really bad, i dump it and start over. when i first got my machine, a 20 minute process stopped me from doing this, but i think im down to 3-5mins now, so doing one more shot doesnt really bother me, and like you said, only doing a few a day, its too hard to know how to grind based on humidity and 'bean age' levels at any given point during the day. thats the part that really bothers me. :/

[timo888]

.. .hard to know how to grind based on humidity ...

I read somewhere that the next Versalab system (the M4Xh) will have an integrated hygrometer and provide grind-adjustment recommendations.

Regards
Timo

[LeoZ]

I read somewhere that the next Versalab system (the M4Xh) will have an integrated hygrometer and provide grind-adjustment recommendations.

Regards
Timo

last i checked the cost was what? 2 g's? heh... i dont want to get kicked out of the house just yet!

[timo888]

... if the puck is preinfused, and a linear ramping of pressure is applied (how most e61s act), i dont see flow becoming an issue, only bitterness!

Illy hypothesizes, and then refers to an experiment that would seem to validate the hypothesis, that pressure in excess of ~9 bar causes the fines to migrate towards the filter, where they create a sort of impenetrable barrier. He has some graphic illustrations representing computer-generated models of the puck's "time-dependent geometry".

I have been thinking about the effects of pressure in connection with the magneto-piston we were discussing over in the lunatic fringe of the Peppina Marries MiniGaggia thread. Gedankenespresso is the tastiest brew. After a hand-controlled preinfusion, pressure created by a magneto-piston would start off high (as the repelling magnets are closest together) and then, instead of ramping up, fall off exponentially, say from 11 bar to 7 bar. I am wondering how that profile would affect migration of fines and the flavor.

Regards
Timo