Menu ≡

David's Blog

The Persistence of Crema, (Espresso Theory)

Sweet espresso extractions feauture unstable crema, such as this coffee at 15 seconds after finishing.

Sweet espresso extractions feauture unstable crema, such as this coffee at 15 seconds after finishing.

Crema at about 10 seconds after shot finished.
Crema at about 10 seconds after shot finished.

When I toured Italy in 1989 and 1991 I was reminded by roasters and food critics that one sign of an expertly made espresso was long lasting, thick crema.  And at the time in our shops, crema remained thick and persistent for 90 seconds or more.

Then when we stabilized brewing water temperature to within one degree F in 2001 I noticed that along with all the sweetness we had preserved into the cup, crema became a light, delicate chiffon that dissapated quickly in the cup.  If you were quick, however, it had a lovely, silky mouth-feel.  Years have passed since then and I have learned that the sweet extraction we can now produce has a fundamentally different quality, probably due to the presence of the sugars.  I trust it is here to stay.

Foam is composed of what are called long-chain surfactant molecules.  Persistent foam is present when the molecules remain in unbroken chains, trapping the gas in the liquid. (Perkowitz “Foam”)  I telephoned food scientist Carl Staub when this phenomenon first occured, and he suggested that the presence of more sweetness in the espresso might be the reason for this breakdown occuring more quickly in the crema. 

It is important to consider this in enjoying caffe espresso as a culinary art, for the full flavor and silky texture: you must enjoy it immeadiately.  Two quick sips from the hand of the barista, at the bar.  The first sip is bracing, all the sass with lighter body,  in the final sip are the sugars,  which invariably sink to the bottom of the cup.

Flow Rate revisited 2012 (theory)

Hello Readers,


As I mentioned some years ago, improvements in espresso equipment or technique often reveal something new about the tricky coffee. Like Ukranian nested dolls, open one and, ha, there is another tiny doll. In the case of temperature-stabilized machines such as the Synesso or LaMarzocco we can see now that flow rate, the speed at which the espresso flows, acts exactly like temperature on the coffee.  Fast pours, under 25 seconds,  are sour just like under-temperature brewing water, and slow shots 28 seconds or more, are hollow. We kind of knew that….but the degree of sensitivity is what is so surprising. 


At Vivace we pull shots in 25 to 27 seconds to reach our target volume.  (Remember, establishing your own volume is an artistic choice you make for your self at your own bar, but Vivace’s double is just under one ounce now.  Did someone say “ristretto”?) Within the 2 second window there are dozens of distinct espresso coffee flavor profiles, none of them defective.



In a balanced shot of  Dolce I am tasting caramel, dark chocolate, leather/salt, blueberry,  and toast.  Depending on almost undetectable changes in the flow rate for the shot, different flavors vie for dominance.  A little slower shot has more caramel.  Another shot may feature more unsweetened dark chocolate or be more on the leather/salt note.  Only the most balanced shot  reveals the toast note.  .  (the notes…the notes…I must have the notes in the cremas….”The Coffee Experience”, you tube, OMG).  Anyway…what effects flow rate? 

1. The grind of course is first and foremost. Not only the coarseness but the particle mix. (See micro-particle blog).

2. Packed volume for the shot- dosage. (This is one of the things we work at the hardest at Vivace).

3. Pump pressure.  On almost all machines there is one pump for all three groups.  So as you activate group #2 with one running pressure drops 1/2 bar or more. Mark has created the Synesso Hydra, his new machine with a dedicated pump for each group head, eliminating the problem.  However, you should still operate out of a static water tank to eliminate changes in the incoming water pressure which will be passed on through the pump to the coffee.

Refractometry and Espresso Quality

There has been some good work done on using digital refractometry to asses total dissolved solids in coffee.  Naturally I was curious to see if a refractometry index (RI) reading could actually be a use ful number to asses the quality of espresso coffee.

I bought the Atago Pal 1 and read up on it.  Refractometery is the measurement of how much light is absorbed by your sample.  For wine making  they measure grape juice  and a high number correlates to a high sugar content, measured in “brix” units.  Hence the term, “brix meter”.   However, that has no bearing on the instruments usefulness to measure sugar content in coffee.  For example, if you measured a sample of India ink it would read 100 because it is very dense and blocks all the light from passing through the sample.

I went to work measuring the RI of many espresso shots, both excellent pours  and highly flawed shots.  Remember that excellent to me means that the shot preserves the fragrance of the ground coffee through the brewing process to be enjoyed as a flavor/aroma sensation.  Here is the test data:

1. Espresso Dolce,  ristretto pour prepared with 17 grams of freshly ground coffee and extraction time of 25 to 27 seconds.  Flavor profile is caramel, leather, toast, dark chocolate cocoa powder, with blueberry notes.               RI= 32.2  (average of about five shots, high 34.0  low 30.2)

2. Stale Espresso Dolce, same pouring parameters.  Flavor profile metallic/ vinegary.   RI= 31.0 (two shots 28 and 34)

3. Espresso Dolce ristretto prepared on a flat burr Mazzer (all other work on conical/flat DRM grinders)  Flavor profile, thinner but similar to #1.    RI= 28.0 (two shots, 26 and 30)

4. Espresso Dolce ristretto , prepared with 200 degree brewing water instead of 203. Flavor profile sour/astringent.  RI= 28.0 one sample

5. Indian varietal  ristretto, Northern Italian roast (my usual roast).  Flavor profile, very sweet caramel. RI= 31.0 one sample

6. Indian varietal same as above, ristretto, very dark roast, oily beans. Flavor profile, burnt rubber, bitter.  RI= 30.6 one sample

7. Espresso Dolce lungo pour 3.5 oz. in 25 seconds. Flavor profile, thin metallic, slightly sour.  RI= 20.7 one sample


It is easy to conclude from this brief study that flavor development and integrity to the fragrance are not measured in any way using digital refractometry.  The only interesting finding is that a coarser grind, yielding 3.5 oz liquid in 25 seconds, reduces the measured RI.  This tentatively  supports the assertion that digital refractometry might be a useful way to measure total dissolved solids in a brewing method.  So perhaps it is useful as ameasure of brewing efficency but will not tell you anything about the flavor profile.

Geneva Sullivan my partner at Vivace

Geneva SullivanDear Reader,


Often I have written about how small business is so tough it will polish you to your best self, or grind you up. And, in twenty four years it provides ample opportunity to run smack into your limitations. Now in my case I’m a wee bit of a hot house flower and I wilt in the presence of excruciating detail or repetitive managerial tasks.  So, I have major help.  It has been my extrordinary good fortune to have Geneva Sullivan  as my partner.  She is my front line general in the war against entropy that is the daily reality of a coffee business. (If it was just me I would be one small shop with very arty espresso but constantly running out of napkins or the teeny light bulbs that go under the bars.)

Geneva is a founding partner at Vivace, having listened to my schemes for improving espresso since 1988.   We got to know each other doing arabic belly dance gigs together with me on flute, Geneva doing cabaret as well as ethnic classical dance. She is a natural performer with a beaming, happy stage presence.

The front line on accounting, inventory, and computers is bravely held by Geneva.  When we met she was a main frame computer repair technician  at Unisys, working nights and arriving on jobs where the CEO is leaking money at $500,000 dollars per hour because the system is down.  This was at a time when hardware repair had the whiff of a male hegemony and she would have to prove herself again and again.  Usually in two hours she had them eating out of her hand.  To say she has moxie is to refer to Mozart as a piano player. She also worked at Digital before joining Vivace full-time in 1991.

During this early time she earned a degree in vocational education at night by commuting all the way over on Bangor Navy base on the Olympic Peninsula.  She finished, while working, and assisting in the beginning of Vivace.  Might I just add serious work ethic to the moxie mix…one descriptive quip I have always liked is to say that if I was going to get stuck on a desert island with one person, it would be Geneva because we would make it back alive.

And over the years I have learned a lot about the nature of intelligence from her.  Im all flash and bang, making intuitive leaps, and abohring process, a fire mind.  Geneva is a stately intellect seeping into every aspect of a problem at her own pace, and illuminating each corner of complexity in an operation or process until it comes out whole and finished.  We like to call it her water mind.  And it is very formidable.

Recently she created a program to calculate what effect any price change on any item we sell, would have on the bottom line at the end of the year.  Her program employs our actual sales hsitory to isolate the effect of, say, raising the price of a short Americano by 15 cents.  So if we have to raise prices we just plug in options and tune until cost of goods sold and net profit line up exactly where we want them to be.  In an early audit by the State of Washington resulted in a $100 refund for Vivace and the auditor pronounced us as one of the most organized small businesses they had ever reviewed.

Her nickname at work is “Hurricane Geneva” for the sheer energy she brings to any project.  And of course, to improve a workspace or a system, you need to tear it apart first.

Dolce Blend- “Schomer’s Opus”

Dear Reader,

After twenty years of roasting and blending I feel that this year’s Espresso Dolce has all the elements I am seeking, balance, sweetness, and complexity. Please allow me the conceit to label it myself as my opus, the best art I can do.

In tasting, the perfectly pulled shot of this blend reveals salted carmel, and toast over a dark chocolate base, with leather, blueberry, and a spice note up high like Cascade Hops in the finish.

Call “1 800 get some”…..just kidding but my web site has it

Micro-particle Migration in Conical Grinding Systems

Micro-Particle Migration in Conical Grinding Systems

Let me start with an unusual apology this time….A few years ago I announced that I was partnering with LaMarzocco to design a direct dosing, conical-burr grinder. I am very embarrassed to admit that we failed to produce any machine. The “S-Grinder” project has ground (hee-hee) to a halt. LaMarzocco and I parted company on the issue of micro-particle control in the dosing system. They took the approach of eliminating static, I favor managing the coffee powder so static cannot affect the micro-particles.

Here is some background: It is well established that grinding coffee with conical grinding burrs(or mixed-burr systems such as the DRM) produces micro-particles along with larger flakes of ground coffee, macro-particles if you will…. The heavy vanes of the conical burr compress the beans until they shatter. An “exploding” bean produces all sizes of particles to enter the fine burrs for shearing. The smaller particles, pass through the shearing and make it into the powder. Then, during extraction, some of the micro-particles actually make it into the cup, giving the espresso made from conical grounds a heavier mouth-feel and more pungent aroma/flavor characteristics. So our choice at Vivace has always been conical burrs for espresso.

However, the beneficial micro-particles present us with two broad sets of problems: controlling the flow-rate of the espresso as it oozes into your cup, and preserving the integrity of the mix of micro and macro particles through the dosing process. (In the interest of scientific rigor, I admit that all my conclusions are based on the DRM mixed burr system, but I believe the following pertains to pure conical grinding heads as well.)

Flow Rate

When the barista hits the switch and pressurized water begins to percolate through the packed coffee, it begins to carry micro-particles with it. Towards the end of the shot these particles can form a mat across the bottom of the coffee basket and slow down the flow-rate of the espresso entering the cup. And as you know, with the high-tech, temperature stabilized espresso machines, such as the Synesso, controlling the flow-rate has become even more critical to capturing the fragrance with any fidelity. Preparing espresso with conical burrs make the control of flow rate more difficult for the barista on the bar because of the micro-particle migration within the packed coffee..

COffee ground with a flat burr grinder. (Magnification is x60, Mazzer M-100) Coffee ground with a flat burr grinder. (Magnification is x60, Mazzer M-100)

Coffee ground with the DRM conical/flat burr grinder. (Magnification is x60) Micro-particles are visible between larger clumps. Coffee ground with the DRM conical/flat burr grinder. (Magnification is x60) Micro-particles are visible between larger clumps.

Note: The grind on the machines used for comparison was set to achieve the same flow-rate for the espresso.  It is interesting just how different the particles are in each photo.

(It used to be, on the “old-tech” espresso machines where temperature wandered around a five degree range that shots from 24 to 30 seconds were all OK, never great. With precision temperature control, shots less than 23 seconds are sour-astringent, and shots over 27 seconds are hollow, almost flavorless. In our system, the best shots tend to pinch off just as the volume is reached and begin dripping between 22 and 25 seconds.. Shots that fall into the 25 to 27 second range, have dozens of flavor profiles. None are defective, hollow or sour, but within the range of acceptable there are so many different espresso coffees. Shots brewed for 27 seconds tends to be heavier on the carmelized sugars but with less varietal nuance. Shots brewing for 24 seconds feature a lighter body and perhaps more varietal nuance such as hopiness from the Brazils or blueberry from the Ethiopian. Ahh, but Vita shots that are in the bull’s eye, (25 or 26 seconds to reach our volume) taste of caramel, toast, blueberry, dark unsweetened chocolate, hops, leather and salt…each flavor competing for your attention in the aftertaste. A perfect balance of these flavors is very rare, and usually one of the varietal flavors becomes more prominent with minute differences in flow rate.

In fact, flow-rate, temperature, and pressure are a intimately connected when it comes to seducing the fragrance into a cup. If the temperature is a little low, say 201 instead of 203, and you extraction time for a shot is 25 seconds with pump pressure at nine bar, you get sour/astringent flavors from a Northern Italian roast. Similarly, if the temperature is right at 203, pressure is nine bar, but flow is fast, say 22 seconds to hit your shot volume, viola, sour astringent. And, if pump pressure is low, say seven bar, and flow rate temperature are spot on you have the same defective espresso: sour/astringent. )

Dosing the Powder

The Italian dosing hopper has within it an impeller system that forces the grounds through a small apeture into the dosing hopper. It is a series of paddles mounted on a flat disc, that spin within a circular chamber around the burrs and impact the coffee in a few milliseconds after it exits the burrs to push it towars the apeture. The effect is to grab the powder in a gentle clumping and once fixed, the distribution of micro and macro particles is preserved. (I really do not know if this is a happy accident or if Italian engineers were aware of particle migration).

However, if a grinder design is attempted to eliminate the clacking doser mechanism you must avoid “free-fall” of the powder. If the coffee is allowed to drop in a “cloud state” without being clumped, even for less than one second, the micro-particles separate and begin to migrate towards any metal surface. The resulting powder, lacking the micro-particles, makes a sickly-yellow espresso without full aroma and flavor.

To address this problem and eliminate the Italian dosing hopper you need to retain the paddle system to impell the ground coffee towards the apeture and into the doser. But that is not enough by itself. About a year ago after giving up on Marzocco I took it upon myself to modify our fleet of DRM grinders. I followed the lead of Malkoenig and Mazzer, and combined a simple polished cone with a wire mesh screen on the exit from the apeture leading to the grinding head. . I found that without the screen the coffee sprayed out of the grinder and micro particles adhered to the vertical, stainless steel wall.

If coffee "sprays" out of the apeture into the cone, the smallest particles adhere to the side of the cone due to the effect of static electricity If coffee “sprays” out of the apeture into the cone, the smallest particles adhere to the side of the cone due to the effect of static electricity

This is the coffee that is clinging to the cone on the left.  (Magnification is x60) This is the coffee that is clinging to the cone on the left. (Magnification is x60)

Simple two-wire grate provides just enough back-pressure on the ground coffee exiting the apeture to gently hold the micro-particles in place within the powder.  It tumbles straight into the porta-filter. Simple two-wire grate provides just enough back-pressure on the ground coffee exiting the apeture to gently hold the micro-particles in place within the powder. It tumbles straight into the porta-filter.

Ground coffee pushed through the grate retains it's even distribution of micro and macro-particles. Ground coffee pushed through the grate retains it’s even distribution of micro and macro-particles.

(It is worth noting that micro-particles can migrate within the powder without sticking to the wall of the doser. The problem can be very subtle and requires a barista that really knows the coffee to detect when the migration problem is slight.) But with a wire grate the powder featured a very nice consistency and tumbled straight down the cone, into the porta–filter. I had to play with it a bit. At first I begin by installing a three wire grate, (two vertical and one wire horizontal in a rectangular apeture of about 1.5″w x 1″ tall) and got very cakey, hard-powder clumps. I then tried two wires and the powder became perfect. This back-pressure gate is the key to allowing the powder to tumble into the porta-filter, through a polished cone, with no particle migration or sticking to the side.. In fact you are tuning the powder for optimal density and manageability. Too much back-pressure and you get dense “caking” of the powder. Too little back-pressure and it can spray in low humidity conditions.

How Does It Work?

We have used the system for over a year with the simple two-wire gate pictured and a couple anomalies appeared. First, we learned that if the grinder burrs get dull the powder enters the hard cakey stage. Easy to fix with new burrs. But we also have a wee mystery…the grinders enter the cake stage on their own for no apparent reason. They will cake-up for 20 minutes, then resume delivering a beautiful consistent powder And this certainly bears out what I have learned in the last four years. Espresso is infinite in it’s complexity but grinding is easily the most complex part of the lovely process. And in this case, I think the machines might be accumulating micro-particles in the paddle area, due to wide manufacturing tolerances creating gaps. Then these particles break loose and clog the chute leading to the dosing cone like point release avalanches….maybe…


Dear Reader,

Something new under the sun…the “Foam Knife One” from Espresso Parts NW is a steam tip with a slit on the end, about 6mm in length, and a 1.2 mm hole in the center of the slit. It fits on LaMarzocco and Synesso machines. I recieved it in a sample pack of three innovative tips and it is the best tip I have ever steamed milk on.

In my micro-pitcher (1/3 L) I was able to foam very small quantities of milk with the best texture I have ever gotten off that wand. It is my training station at our Brix location,  I steam there every day and this tip produced the the finest foam texture I have ever seen come off that wand.

Thanks to Terry Z. and Espresso Parts NW.

Espresso Theory-Water Purity and Mineral Content

Dear Reader,


Let’s revisit a factor I have not written up for 15 years.  Water purity and mineral content.  For about a year in 2006 we had water formulators to “tune” our brewing water to a specific mineral content.  The effect on the flavor of our coffee and tea was sublime.  But what is the best water for espresso,  saturation method coffees, and tea?  In my opinion, it is water with a TDS of about 150ppm.  Ok, that is a mouthful…TDS means total dissolved solids and is also refered to as mineral content, and ppm means parts per million. (Water tuned to this mineral content will have a slight oily taste in the back of the mouth when you drink it. Water in most E. Coast cities will weigh in higher, around 350ppm and up, and features a very oily feel and taste and usually will be filtered with a reverse osmosis filter just for drinking.)

As I said we worked with this water for about 12 months in 2006 and 2007, using a very sophisticated mixing computer and mineral additives to formulate the mineral content in our water at all our bars..  Unfortunately, the company was owned by a brilliant scientific visionary with no idea how to build a company to service these high tech gizmos and eventually we had to remove them.  But during this time I discovered many characterisitics about espresso and tea brewed with this water that I will share with you.


Really the effect on espresso and tea was remarkably similar.  Let’s make a comparison with Seattle water that has been run through a carbon filter only and has a mineral content year round of about 50ppm TDS.  First the flavor of the espresso lacked a sharp point over the top of the pallette that is similar to making espresso on equipment that is new, or recently cleaned with detergent without seasoning the machine. Literally it is metal molecules that leach out of the boiler (or teapot) when water is heated up and it tastes metallic.. Molecules leave the metal through osmosis because of the low concentration of minerals present in the water. (This water tastes great for drinking as does most mountain states water. In Vancouver BC the water often contains as little as 10ppm of TDS.) So the flavor was softer with no harsh point. The same was true for macha, the concentrated green tea made with only the baby tea leaves and whisked to a foam during many tea ceremonies throughout Japan. The macha tasted sweeter with no harsh point.

The other effect on the flavor of espresso was to open up a multi dimensional layering of distinct flavors. I could taste the blueberry from the Harrar, caramel, dark chocolate, honey, anise, and “hoppiness” from the Brazil, very clearly, stacked up on my tounge in distinct flavors that were much easier to perceive than espresso brewed on 50ppm water.   This was quite remarkable and made cupping a very clear affair even using the espresso method to sample coffees.  (….and I struggled for months before ripping out these systems..floods, lack of supplies, it was a plauge of amateurism and marks the first time Vivace has ever backed off on improving a factor.)

A similar effect took place with the macha, bringing out a rich “umami” with a note in it, believe it or not, like a great steak.  The flavor of the macha was layered as well like a springtime breeze with underlying richness.  (Interestingly enough, my friends favorite old iron tea pot produced the desired TDS of about 150ppm when Seattle water was boiled in it for a few minutes. I got the biggest kick out of that.  The old tea masters in Japan had intuiitively chosen this style pot long before the modern era,  as creating the best tea, or perhaps it was a lucky shot.)


So, how to you measure your water?  get a  TDS meter from the Myron L Company like the one pictured below. It is called an Ultrameter II and I recommend picking up the calibrating fluid shown as well.


How do you tune your own water?  For areas like Western Washington, that have pure mountain runoff water with a very low TDS I do not have an answer.  I am hoping someone knows of a water tuning computer made somewhere that is supported by a professional company.  If you do please e-mail me at attn: David.

For most of the country where the mineral content is high, I would filter the tap water through reverse osmosis, producing literally zero ppm of TDS,  and then re-mix to the desired mineral content.

I consider water formulating a final factor to bring into an espresso program that is already very tight.  The effect, while delightful, is subtle on espresso, perhaps more noticeable for saturation coffees and tea.

Espresso Theory-pressure profiling

Dear Reader,


Pressure profiling essentially describes the abilityof the espresso machine to change the brewing pressure during the extraction of a shot and then save your changes in a program to be repeated on the bar.  The hope, of course, is to improve espresso coffee.

I see no increase in espresso quality through pressure profiling.  My opportunities to play on these new machines have been for just a few hours here and there. However, espresso requires extreme temperature stability for these fragrant molecules to be preserved into your cup. I hereby postulate that stable pressure is also extremely important in the preservation of these unstable molecular structures that constitute coffee flavor/aroma. It is purely an espresso geeks intuition…but I bet I am right.

Also, it seems like the machine companies just want a new bell or whistle on their equipment.  This is not led by better results in the cup but by manufacturers wanting to market additional complexities in their equipment.  Kind of like Illy’s blizzard of scientific analysis, and chemistry terminology, in an effort to obscure the simple truth that only fresh coffee is good coffee.  Preparing espresso is complicated enough without pressure profiling.  C’mon machine gods….give us a grinder.

Equipment-direct dosing grinders

Dear Reader,


After many efforts to assist in the manufacture of a new grinder with no Italian-style dosing hopper (read: air/coffee mixing device that contributes annoying clacking sounds to the espresso bar) I decided to just outfit my own DRM grinders with Mazzer cones. (I use the mixed-burr model with the rubber drive belt.) A few months into it I will cautiously announce success.



The trick is to manage the powder to avoid micro-particle migration as it tumbles into the chute. The powder coming from the DRM mixed burr does contain substantial micro-particles along with larger particles. These micro-particles give the espresso crema a heavier body and more intense flavor, but if  the powder is allowed to “float” for even a fraction of a second after it exits the grinder port, static will draw the micro-particles to the edges of the “cloud” of ground coffee ruining the extraction.  The trick is to provide just the right amount of back pressure on the stream of powder as it exits the grinder port so it tumbles into the chute in a manageble, cake-like consistency.  Too much back pressure and it is too hard of a cake, not enough back pressure means it sprays into the chute and micro-particles stick on the sides.  I had to try several different designs of gate systems on this port to achieve just the right powder. Even among my eighteen DRM grinders different gate designs are required due to the individual characteristics of each grinder.   And, I suspect that throughout the year we may experience the need to change the gate configuratiuons to accomodate different weather.  Complexities upon complexities…I love this.


The advantages are many: 

1. We see better espresso due to a more homogenous powder.  Forcing the powder through my gate system breaks up the natural lumping that occurs in fresh coffee as it is  pushed out the port.  Flow rate inconsistencies are also minimized in the pour. 

2. We can go faster on the bar.  Grinding is now hands-free for the barista. It takes some getting used to….installing the empty porta-filter into the mount before hitting the grinding switch changes the rythym of the barista but my staff is unanimous that it is a much faster system once you get used to it.

3. No gawd-damn clacking sound in my shops.