gscace wrote:Contact them directly if you're interested in a 1-group.
-Greg

I have, quite some time ago. Perhaps it is time to approach them again.

Thanks again for participating and allowing the videos to be published.

-Mike

So I finally managed to watch all of the videos. In essence, a couple friends spend an afternoon making a LOT of espresso and geeking out on machines and grinders.

Some random observations:
- I wish all the people who seem to think that there is an absolute ranking of grinders from best to worst would watch this series of videos;
- my Scace I thermofilter seems pretty indestructable. How can people break these things? The only thing that has happened over the years was a bit of the mesh broke off and got stuck in the exit hole.
- it was great to see the Leva X at work. Finding out these is/are cartridge heater(s) in the saturated group was illuminating; and being able to adjust the spring (and thus, I presume, peak pressure) is something every lever-head that has played with springs on conventional spring levers would dream of.

I was wondering:
- Does the choice of notch to engage the clutch pawl impact peak extraction pressure? In a conventional spring lever, it is always a full pull so the flexibility the Leva X offers is unique;
- the fixed orifice of the Scace thermofilter typically results in longer 'extraction' times (e.g., 60 seconds to get the target volume) on spring lever machines due to the declining pressure profile. There has also been some discussion that reported temps during initial wetting/pre-infusion is not representative, especially in dipper levers which pre-infuse at service boiler pressures - and with flash-boiling(!) water. In my usage, say for a WBC-ish test sequence, I only concern myself with maximum temperature - which usually corresponds to peak pressure, and seeing similar slopes of declining temperatures across pulls for the rest of the 'extraction.' I was wondering what your experience has been with the Leva X - assuming that you have done some measuring since the videos. (I'm not asking for specific Leva X performance numbers, but your experiences in obtaining them.)
- the little flush before locking in the portafilter was explained to me as needed in a pump machine for two reasons: 1) to clean off the shower screen, and 2) to get rid of colder water in the group. The first reason doesn't apply as there is no 3-way solenoid, but I was wondering if you had any further thoughts beyond what was expresed in the video about the usefulness of this flush.

Thank-you for allowing Jay into your house with video ... and that's a cool camera backpack you have!

Very cool stuff

I like the dosing funnel

Where can you get that

baldheadracing wrote:So I finally managed to watch all of the videos. In essence, a couple friends spend an afternoon making a LOT of espresso and geeking out on machines and grinders.

Some random observations:
- I wish all the people who seem to think that there is an absolute ranking of grinders from best to worst would watch this series of videos;
- my Scace I thermofilter seems pretty indestructable. How can people break these things? The only thing that has happened over the years was a bit of the mesh broke off and got stuck in the exit hole.
- it was great to see the Leva X at work. Finding out these is/are cartridge heater(s) in the saturated group was illuminating; and being able to adjust the spring (and thus, I presume, peak pressure) is something every lever-head that has played with springs on conventional spring levers would dream of.

So the cartridge heater thing is quite convenient, because you can use commercially available devices and adapt them. On the other hand, engineeringwise the choices are not entirely perfect because the relatively local introduction of heat means that the energy has to diffuse through some materials with really poor thermal diffusivity, so there are hotter and colder places around the system. Whether or not this is important is an engineering judgement and the fact the the Leva is as good as it is demonstrates that the use of cartridge heaters is an entirely adequate and convenient solution although perhaps not elegant. Elegance would raise the price still further with no real benefit and perhaps a serviceability detriment.

baldheadracing wrote:I was wondering:
- Does the choice of notch to engage the clutch pawl impact peak extraction pressure? In a conventional spring lever, it is always a full pull so the flexibility the Leva X offers is unique;

Yes, the peak pressure is adjustable depending on which notch one engages, and also by the position of the spring adjusting nut on the back of what I call the "Alien Head." I think La Marzocco's promotional publications refer to the body of the mechanism as resembling a shark, but to me it's Alien all the way. This gives one the opportunity to sample coffees at several different peak pressures with no adjustment at all, and the ability to globally adjust brew pressure up or down. Different notches produce different possible shot volumes, so holding shot volume constant requires a digital brewing scales so that one can know when to pull the cups. The amount of water used in the pre-infusion stage is variable, depending on time and the particular coffee being used. I had a long chat with Scott Calendar of La Marzocco, who has extensive Leva experience. His preferred work flow was to preinfuse until drops appear, then to engage the spring. This method vastly improves consistency. He was also a fan of using only the first (lowest spring pressure) notch, and adjusting spring pressure via the adjustment nut. For commercial use, ergonomics are a bit better this way, according to him. But if I don't know any better, my preference is to pull shots with brew ratios around 50% (38 gm shot from 19 gm coffee). So I use the second notch and it's very close to right and I often dispense with the scales. My workflow has evolved a bit from the videos. I use botttomless portafilters so that I can observe the preinfusion step more closely. I engage the spring when I see coffee appearing at the floor of the filter basket. I use the second notch, as mentioned above, and I pull pretty long extraction times. The Leva is so stable that one can do this, unlike a heat exchanger machine. I'm not afraid of high 30 second to 40 couple of seconds extraction times, which is something people were doing with EK grinders when I started playing around with it. I forget if I mentioned Noyes-Whitney or the basic mathematical model for extraction that is relevant for coffee, but time is one of the primary parameters in extraction, as you know. If temperature can be held stable and if the flow behavior through the coffee cake is reproducible and has minimal to no local channeling (thus nominally good axial uniformity), then you can get high extraction yields that taste really good. The Leva is extremely good thermally - Kent Baake told me it was more stable than a Strada, which is saying something. I don't have a Strada here, but I can tell you that the Leva is remarkably reproducible. You can see the degree of axial uniformity by observing the arrival of coffee at the bottom of the filter basket during preinfusion and by the degree of pour centering if one uses bottomless portafilters. With care, one can get brew time reproducibility on the order of a couple of seconds, which I think is pretty good. FWIW, discussion of brewing parameters in Illy's "Espresso, the Science of Quality" supports extended extraction times. The presence of fines in ground coffee seem to affect how successful one can be with axial flow symmetry and long extraction times. I have a Titus Nautilus here now with 93mm (I think) SSP high uniformity burrset. It's a bit more finicky to use with some coffees, but produces a different cup in the Leva - one that i sometimes prefer and sometimes not, depending on the coffee)

baldheadracing wrote:- the fixed orifice of the Scace thermofilter typically results in longer 'extraction' times (e.g., 60 seconds to get the target volume) on spring lever machines due to the declining pressure profile. There has also been some discussion that reported temps during initial wetting/pre-infusion is not representative, especially in dipper levers which pre-infuse at service boiler pressures - and with flash-boiling(!) water. In my usage, say for a WBC-ish test sequence, I only concern myself with maximum temperature - which usually corresponds to peak pressure, and seeing similar slopes of declining temperatures across pulls for the rest of the 'extraction.' I was wondering what your experience has been with the Leva X - assuming that you have done some measuring since the videos. (I'm not asking for specific Leva X performance numbers, but your experiences in obtaining them.
- the little flush before locking in the portafilter was explained to me as needed in a pump machine for two reasons: 1) to clean off the shower screen, and 2) to get rid of colder water in the group. The first reason doesn't apply as there is no 3-way solenoid, but I was wondering if you had any further thoughts beyond what was expressed in the video about the usefulness of this flush.

There are a lot of parts connected to the piston in the Leva group and while the heaters help, there are flow paths that while poor, still exist. I do flush a small amount. Unlike other lever machines that I have used, including traditional levers and Kees Vad der Westen machines, you don't have to play games to get the starting temperature close to right. The Leva incorporates the advantages of very orderly introduction of water with extremely reproducible and stable brewing temperature. It's quite unique.

I forgot to say (and why my post is edited) that flow rate through a Scace device with the Leva is a little faster than with a pump machine, so I use a smaller orifice to get extraction times near what I like for Scace 2 devices. Kind of the opposite to your experience.

baldheadracing wrote:Thank-you for allowing Jay into your house with video ... and that's a cool camera backpack you have!

michael wrote:Very cool stuff

I like the dosing funnel

Where can you get that

Weber Workshops. https://weberworkshops.com/products/blind-shaker

Best engineered single dose redistribution tool I've ever used.
-Greg

gscace wrote:... what I call the "Alien Head." I think La Marzocco's promotional publications refer to the body of the mechanism as resembling a shark, but to me it's Alien all the way. ...

From now on, cognoscenti will call the Leva X the "Alien Head" machine.

Hi Greg,

Excellent videos. I have been spending time thinking about the concepts.

When you state that the Carimali DivaPro gets 22% - is that for the same CBR that say - the Linea or the Aurelia II would get a lower EY (say 20-12%)? Is it a function of the espresso machine (e.g. a programmable PI, or a group configuration)? Or a puck phenomenon?

In regards to grinders "they all taste different" due to the ratios of "fines" to "boulders" (a flat paper vs. a crumpled one in a ball).

And yet whomever tries to sift (e.g. the Kruve sifter), and whomever tries to create a unimodal grinder (if that were possible) - it doesn't seem to work at all, and all grinder burrs designs seem to return to the boulder chutes design (and humped grind distribution graphs). Perhaps giving up that last 8% (by virtue of having fines and boulders) is what allows the espresso to emulsify?

In the first video you were showing a ppt titled "Understanding and Quantifying Espresso Extraction". Is that something you can share?

Thanks for the great work and for sharing it.

AssafL wrote:Hi Greg,

Excellent videos. I have been spending time thinking about the concepts.

When you state that the Carimali DivaPro gets 22% - is that for the same CBR that say - the Linea or the Aurelia II would get a lower EY (say 20-12%)? Is it a function of the espresso machine (e.g. a programmable PI, or a group configuration)? Or a puck phenomenon?

As a consulting engineer I was asked to improve the extraction yield of that particular machine. The improvements I came up with included some different designs for some of the group components, and also cherry picking commercially available components - brew baskets, gicleurs etc. to find what worked best. I used brew ratio of 50% and the same dry coffee weight for comparisons between other machines that i own and am intimately familiar with. Comparisons were apples to apples in that regard.

In regards to grinders "they all taste different" due to the ratios of "fines" to "boulders" (a flat paper vs. a crumpled one in a ball).

And yet whomever tries to sift (e.g. the Kruve sifter), and whomever tries to create a unimodal grinder (if that were possible) - it doesn't seem to work at all, and all grinder burrs designs seem to return to the boulder chutes design (and humped grind distribution graphs). Perhaps giving up that last 8% (by virtue of having fines and boulders) is what allows the espresso to emulsify?

I don't think that is the case (emulsification). Bimodal extractions are easier to keep consistent than unimodal ones. By keeping consistent I mean maintaining axial symmetry of water flow through the cake, without regions of higher or lower flow rate. You can see this when brewing with a bottomless portafilter - good symmetry with bimodal burrsets and more difficulty achieving it with unimodal. Regions of the cake with fast water flow tend to over-extract relative to regions of the cake that have reduced flow as a result of the nonuniformity. The reason for this is that the rate of energy transfer (heat) from the water to the coffee is dependent on the local velocity, and the amount of water and available heat. Regions with little flow remain colder and so the local diffusion coeficient is low compared to regions with high flow. Also, the concentration gradient across the boundary layer in fast flowing regions is higher, which increases diffusion rate of dissolved coffee into the flowing water. If you can get a shot produced by unimodal burrs or high uniformity burrsets that has axial symmetry, then you can get higher yields because the boulders are smaller and fines are less. You can grind the coffee finer which increases surface area. That all helps get high yields, but the problem is right there in the flow uniformity problem. It's easier to get that uniformity with bimodal burrsets. Some coffees seem to be better than others in this regard, so those high uniformity burrsets can work, but not for everything and not as easily.

AssafL wrote:In the first video you were showing a ppt titled "Understanding and Quantifying Espresso Extraction". Is that something you can share?

Lemme look for it.

I also have questions.
We all know what to expect with spring levers... but this seems to be a very different design. Does it taste like a spring lever machine? How does that large group stay hot?

gscace wrote: I don't think that is the case (emulsification). Bimodal extractions are easier to keep consistent than unimodal ones. By keeping consistent I mean maintaining axial symmetry of water flow through the cake, without regions of higher or lower flow rate. You can see this when brewing with a bottomless portafilter - good symmetry with bimodal burrsets and more difficulty achieving it with unimodal. Regions of the cake with fast water flow tend to over-extract relative to regions of the cake that have reduced flow as a result of the nonuniformity. The reason for this is that the rate of energy transfer (heat) from the water to the coffee is dependent on the local velocity, and the amount of water and available heat. Regions with little flow remain colder and so the local diffusion coeficient is low compared to regions with high flow. Also, the concentration gradient across the boundary layer in fast flowing regions is higher, which increases diffusion rate of dissolved coffee into the flowing water. If you can get a shot produced by unimodal burrs or high uniformity burrsets that has axial symmetry, then you can get higher yields because the boulders are smaller and fines are less. You can grind the coffee finer which increases surface area. That all helps get high yields, but the problem is right there in the flow uniformity problem. It's easier to get that uniformity with bimodal burrsets. Some coffees seem to be better than others in this regard, so those high uniformity burrsets can work, but not for everything and not as easily.

So if a fine powder would have higher and lower density areas (as would happen if the fine powder was not perfectly set up and then tamped) - some areas would stay dry and under extract - and some would over extract. I would assume then that the boulders keep sanity by ensuring most "uniform" particles would have a large boulder nearby that would get it wet. I can see that happening.

In ice cream (and other fancy texture making), when we distribute hydrocolloids (gums, gels), we mix them with larger particle powders (like sugar). Otherwise the very fine powders tend to easily clump up. Wet outside, dry innards. The sugar granules allow water to access the gum particles and hydrate them. Even then we slowly pour the sugar/gum mixture into a blender vortex.

So that makes a real case for a well performing shower filter that provides the homogenous flow for all the basket. Makes the sintered filters (which convert pressure to flow) even more compelling...

Also - since fines over extract - does this not mean that the 8% you were looking for in part 1 (up to the generally accepted limit of 30% EY) are stuck in the boulders? The centers of which are less exposed to water flow (and probably dry for some boulders)?

Basically, we are sacrificing the 8% to be able to get 22% (without the boulders we'd get even less)? Maybe getting to up to 24% (given the same CBR) is a matter of gradually reducing boulder size - or reducing the number of boulders - until an inflection point where flow starts to become non-uniform?