One week with the La Marzocco GS3

[HB]

With the holiday coming and some of the staff on vacation, the roasterie at Counter Culture Coffee is super busy. They certainly won't have time to work with this... So why not put it to use? Day 1 of 7 starts now.


PeterG generously agreed to a one-week loan

[cannonfodder]

I don't suppose that is a Mazzer Kony I see peeking out on the right?

[HB]

No, it's a Super Jolly. I tried to cajole a Robur that was in for repairs out of their technician's hands without success. Plans are afoot for a mini-jam at my house; with luck I can sneak one out of their espresso lab on Friday for the weekend.

I am thinking grinder upgrade again. Based on Greg Scace's glowing comments, the Kony is on my shortlist. Unfortunately the height is an issue until I find a dedicated espresso workarea. Oh bother! However, there is comfort in using one's own equipment. The GS3 dialed in nicely at home while I struggled at Counter Culture's espresso lab the other day. Peter later informed me that the "mystery beans" in the grinder were weeks old. That happens frequently, i.e., the coffee is so fresh it's still warm, or it's leftovers that have sat for weeks.

[HB]

No espresso machine in recent memory has been lavished with such attention as the La Marzocco GS3, despite that only prototypes are available for evaluation (according to reports, US imports aren't expected until 2007). Besides the "star" factor that drew my interest, another reason I wanted to check it out more closely is because after sampling its espressos at the SCAA conference, I was left unimpressed. In the GS3's defense, the samples were prepared using a Swift "auto-everything" grinder.

A few weeks back, Peter at Counter Culture Coffee invited locals to check out the GS3 that was recently brought in-house. I thought that a few sessions in their cupping lab would be enough to form an opinion about the espresso machine's capabilities. But during two casual Friday morning sessions, it wasn't coming together. As I mentioned in my earlier post, that's the problem with ad hoc espresso jams: You're dealing with unfamiliar equipment and coffees of unknown age and composition. After all, it's an "espresso lab" and constant change is the norm. If you want any consistent results, it requires preparation, including having known test coffees on hand.

Peter mentioned that the GS3 was a short-term loaner. With the July 4th holiday approaching, the crew at Counter Culture is shorthanded and the chances of any experimentation happening this week are low. I jokingly noted in the site's announcement box "CCC has a GS3 in their espresso lab... 2nd session and so far, it's good but hasn't come even close to rocking my world. I think a loan is required to investigate." Peter picked up on my message and asked:

Do ya wanna borrow?

To which I replied:

Do ya have to ask?

Woohoo! In my own kitchen using my own kit and familiar coffees! In the morning, I stopped by to pick it up. It's heavy, but luggable if you're highly motivated. Trundling back home with the GS3 in the back, I mulled over a reporting angle on this machine. Needless to say, the temperature / clarity angle has been discussed ad nauseum. Given the short nature of its stay in my kitchen, I'm going to focus on initial impressions and the "forgiveness factor." I'll document more thoughts about the first day tomorrow.

As I cleaned it up for the night, I noticed how easily the grouphead surfaces wiped down. The underside of the grouphead is a smooth disk with a center hole for the screw holding on the dispersion screen:


Grouphead with dispersion screen removed.
Nothing but smooth surfaces

The dispersion screen is slightly offset from the disk and its center screw is drilled down the center for most of its length; four tiny side-holes act as water jet breakers to disperse the water evenly across the top of the puck.


Dispersion screen held in place by no ordinary screw

[HB]

(As you might imagine, last night was a late one. Before continuing with today's report, allow me to wrap up yesterday's first impressions.)

After setting it up, I took a step back to admire the GS3. You know, compared to some other espresso machines that have passed through our doors, the GS3 isn't a looker. Rather boxy. Others have made the same observation; I recall seeing a photo collage comparing it against the Speedster, suggesting the GS3 was modeled after a Ford LTD and the Speedster after an Italian sports car. That said, cramming all these goodies into a package this small was a minor miracle. I was able to carry it myself from the car by gripping the sides. Although it's heavier than the Elektra A3, it's a lot easier to carry. Not that I recommend it; I was waiting for my spine to pop out when it came time to heave it onto the countertop.

If you're desperately looking for reasons not to buy a GS3, I would put poor ergonomics at the top of the list. Was the product designer left handed? Or is the machine's compact nature force compromises due to interior room constraints? The workflow from brew-to-steam is clumsy:

• Steam arm is on the left, toggle control is to its right
• Brew controls are on right, opposite side of the grouphead.
• Steam arm doesn't reach the driptray (must purge into a steam pitcher)

And while not part of the ergonomics proper, the programming could be politely described as "Byzantine". I was able to piece it together with the help of Chris Tacy and Andy Schecter. Inexplicably, the least-often changed settings are first in the program cycling. Brew temperature is buried near the end... let's hope they revisit the ease of use in the production version.

Tonight I focused on extraction diagnosis and confirming La Valentina was tuned to the same pressure for some side-by-side action later this week. I was pleased how quickly the GS3 dialed into nice 27 second extractions. Before heading upstairs to bed, I noticed the back of the house was running 5F degrees warmer than normal. This machine puts out a beastly amount of heat. What do Italians have against insulation, at a minimum for the steam boiler?

[HB]

It's rare that our household is quiet enough in the morning to make an espresso-related video. The audio tracks were removed from the few that I have made to save viewers from the roar of children in the background. But this morning was different... my lovely wife offered to take all the boys to the park. Below is the first series of extractions of the morning. Since this writeup is about first impressions, I haven't "cherry picked" the best videos. A couple espressos to adjust the grinder, then I recorded all subsequent extractions, starting with Counter Culture Coffee's Toscano.

Note: I placed a lamp with 100W lightbulb less than 18 inches from the grouphead to show details more clearly. My apologies for the washed out colors.





Approximately 18 grams of coffee, LM double basket. Both extractions were direct brew, i.e., the start-stop preinfusion was not used. Final volume was approximately 1.75 ounces.

[HB]

This next series is Intelligentsia's Yemen Sanani. I wanted to try the start-stop preinfusion and pull a slightly tighter extraction. You will hear distinct clicks as the preinfusion solenoid closes and reopens. The pump continues to run during the pause that follows the initial presssurization.

The first extraction was a little slower than desirable for this espresso, so I loosened the grind for the second extraction. Overshot it slightly and the third extraction was the best of the three. Spicy red wine with cream.







Yemen is denser and has a higher moisture content than Toscano. It compresses to a thinner puck, but I kept the weight close to 17 grams. Updosing Yemen tends to yield an overly bitter espresso. By the way, of the four tampers I tried, the best fit was the HB Compressore (convex). No, really:


Photo courtesy of EspressoParts.com

[Nick]

Ummm, did you actually taste any shots? Because you haven't mentioned doing so.

[framey]

Ummm, did you actually taste any shots? Because you haven't mentioned doing so.

Overshot it slightly and the third extraction was the best of the three. Spicy red wine with cream.

Kind of brief for you Dan

[HB]

Ummm, did you actually taste any shots?

Yes, I tasted quite a few and they were as good or better than most cafes, including yours. But I don't consider the coffees I had on hand particularly difficult (Toscano, Black Cat, Yemen), so that comes as no great surprise. Peter, Miguel, and Jim have some coffees that will offer more challenge for this weekend.

What was a surprise to me, given my prior experience with La Marzoccos, was that the GS3 produces such good results with less difficulty than its siblings. Later in the week I plan side-by-side comparisons of the GS3 and La Valentina. It's not a fair match-up, but I hope that will give some food for thought to those considering the really big upgrade.

By the way, I'm organizing a mini-espresso jam at my house this coming weekend. Those in the area are welcome to contact me for time and directions.

Kind of brief for you Dan

Chris, Mark, Andy, Greg et al have covered the question of taste performance at length. I started this thread in Overextracted in hopes of finding a new, unexplored angle. Whether it will be interesting remains to be seen...

[another_jim]

If you're in a debunking mood, Dan, or at least a tough one, serve some blind paired shots during your weekend jam. Generally a series of 5 shots per taster will reveal whether there is a big difference. For small but real differences, you have to go out to about 20 to 30 shots, not really possible for a weekend.

Your parallel test at CCC showed no large differences between the Linea, Junior and E-61 box; so if the GS3 does show a big one, it would be major news.

[AndyS]

What was a surprise to me, given my prior experience with La Marzoccos, was that the GS3 produces such good results with less difficulty then its siblings.

Bill Crossland describes the GS3 as an "easy" machine. I believe he says that it may or may not make better espresso, but it definitely makes it easier.

Question: do the Lineas have the same drilled-out dispersion screw? Or is that new with the GS3? I think it's a very good design because water enters the area above the screen so well dispersed.

On my tricked-out Silvia I can set the water debit and extraction pressure wherever I want, but I still often get flow weighted towards the front of the puck. This appears to make sense, because the water path into the group is pointed to the front by the way the head is drilled. Not so with the GS3.

[darrylr]

Yes, Lineas use the same kind of dispersion screen.

Darryl

[miKe mcKoffee]

Nice video work Dan. Takes naked porn to a whole new level compared to glossies. Might be nice to have both the PF and clear demi in the videos if possible.

[HB]

If you're in a debunking mood, Dan, or at least a tough one, serve some blind paired shots during your weekend jam.

That's definitely on the agenda, though I don't expect a big turnout given the holiday and short notice. Lino is coming by tonight to hook up the Mr. Wizard Kit (temperature and pressure profiling). The GS3's temperature performance has been exhaustively documented, but I don't recall anything about the start/stop preinfusion and the resulting pressure profile.

Might be nice to have both the PF and clear demi in the videos if possible.

OK, next time I'll get out the Illy Nudes and put them on a "booster" to get them in the frame. In the meantime, below is a demonstration of my feeble photography skills:


Washed out by the flash, bad exposure,
sat for a minute... but still quite tasty

[Teme]

If you're desperately looking for reasons not to buy a GS3, I would put poor ergonomics at the top of the list. Was the product designer left handed? Or is the machine's compact nature force compromises due to interior room constraints? The workflow from brew-to-steam is clumsy:

• Steam arm is on the left, toggle control is to its right
• Brew controls are on right, opposite side of the grouphead.
• Steam arm doesn't reach the driptray (must purge into a steam pitcher)

And while not part of the ergonomics proper, the programming could be politely described as "Byzantine". I was able to piece it together with the help of Chris Tacy and Andy Schecter. Inexplicably, the least-often changed settings are first in the program cycling. Brew temperature is buried near the end... let's hope they revisit the ease of use in the production version.

I am under the impression that some of these have been fixed for the production model. Among other things:

a) the production model's steam arm moves freely and can be placed over the drip tray
b) the production model's drip tray is larger
c) the keypad and programming should be more user-friendly

This doesn't help regarding the location of the items, though.

Br,
Teme

[HB]

the production model's steam arm moves freely and can be placed over the drip tray

The arm rotates in a horizontal plane but is several inches above the driptray. I doubt they would want to make the steam wand reach lower than it already does. Despite the disappointing steam arm ergonomics, the quality of the microfoam is excellent. You would have to try really hard to screw it up.

[Teme]

The arm rotates in a horizontal plane but is several inches above the driptray. I doubt they would want to make the steam wand reach lower than it already does. Despite the disappointing steam arm ergonomics, the quality of the microfoam is excellent. You would have to try really hard to screw it up.

I was at the factory two weeks ago and they had one of the prototypes there. From the discussions there I did understand that also the movement of the steam arm was to be redesigned for production. I may have misunderstood, though...

Br,
Teme

[HB]

Few rotary pump espresso machines have an inboard water reservoir, but this is just one of unique characteristics of the GS3.

As Ken and Jim documented in How to Preinfuse; Extraction Pressure Redux, there is arguably a benefit to preinfusion under mains pressure ("At 3 bar, good stuff is happening, the water has enough pressure to go through the whole puck in a reasonable time."). That's not an option for espresso machines drawing water from a reservoir, but the GS3 does provide start-stop-start preinfusion by closing / reopening the grouphead solenoid in the early seconds of the extraction.

Lino stopped by this evening to help plot the GS3's pressure profile resulting from this design:


The "speed bump" is the on/off pulse

The thread Pressure profiles, preinfusion and the forgiveness factor shows the pressure profiles of several other espresso machines. This profile resembles the Elektra A3's no preinfusion design, except for the small speed bump and a less inclined rise, presumably thanks to the GS3's smaller diameter gicleur. Lino and I were surprised to see that the group depressurizes and repressurizes after the preinfusion cycle. I programmed in one second of pressurization followed by two seconds of dwell time. While I haven't tried it yet, I assume that allowing much longer would mean a higher rise and fall, which could disrupt the puck's adhesion. Given the GS3's already easy nature, I wonder if it would improve further still with a higher pressurization to 3 bar and no depressurization as How to Preinfuse advocates?

With the Wizard Kit tucked away, Lino and I turned to pulling shots for the next hour or so. Both of us noted an ashy finish from the get-go and went about diagnosing the cause.

A brief interlude before continuing today's report...

Peter mentioned they had not changed the brew temperature since the GS3 was delivered and that it might be running a little cold - I believe the display said 200.3F when I picked the machine up on Monday. Lino had borrowed my thermofilter, so I couldn't confirm the temperature by direct measurement. The taste of the Toscano indicated the brew temperature was a little too hot, not too cold, and over the course of two days I've worked down ~1.5F to 198F, as displayed on the control panel. Earlier today, the discrepancy between the displayed and apparent brew temperature prompted me to e-mail two GS3 beta testers to ask if there was a programmable offset, and if so, how does one adjust it. [The Expobar Brewtus has an offset to compensate for the difference between the measured boiler temperature and the (displayed) grouphead brew temperature; for the Brewtus, it's a 6-7F delta, depending on who you ask.] Neither was aware of a programmable offset. No problem, I would continue to treat the displayed temperature as "N" and adjust by taste until Lino arrived with the thermofilter.

Sorry... back to Lino, Dan, and the inexplicable ashiness. At first we blamed the extractions; they started a bit dark and ended a bit light. Changed the grind, updosed, downdosed. We even changed baskets to the our more familiar Faema-style instead of the stock La Marzoccos. Still, the ashy aftertaste remained.

Finally the 15W lightbulb that was hovering over our heads increased to 25W: What about the temperature? Afterall, if someone reported problems with ashy flavors in the forums, what would be the diagnosis? From The Home Barista's Guide to Espresso:

Ashiness: Usually a flaw in rapidly dark roasted, low grown coffees. Drop the temperature to the low end of the espresso range. Dial in to the lungo end of the optimum crema range and dark stop the shot (you may be under 20 seconds when you do this, that's OK). These measures will not much reduce the ashiness, but will mask the problem with a little more brightness and crema. The real solution is to change blends.

Down, down, down went the temperature as we retried shot after shot... 198F... 197F... 196F... 195?!? At that temperature, the ashiness was nearly gone, the espresso the most pleasant of the series. But c'mon, we asked each other, this temperature readout can't be right, can it? A few minutes with the thermofilter confirmed a three degree offset between brewhead temperature and displayed temperature.

The lesson we learned tonight: Comfort in digital readouts is false comfort. Our tastebuds were yelling the answer and we ignored their calls because an LCD display said it wasn't so. "Fool me once, shame on you. Fool me twice, shame on me."

[annp]

First, I'm glad you have the GS3 to play with, Dan.

I always appreciate your viewpoint because while you have all the graphs and gauges, you also approach stuff from a non professional perspective. You have solid data but you don't neglect usability impressions.

If you're desperately looking for reasons not to buy a GS3, I would put poor ergonomics at the top of the list. Was the product designer left handed? Or is the machine's compact nature force compromises due to interior room constraints? The workflow from brew-to-steam is clumsy:

• Steam arm is on the left, toggle control is to its right
• Brew controls are on right, opposite side of the grouphead.

This machine puts out a beastly amount of heat. What do Italians have against insulation, at a minimum for the steam boiler?

Those of us who are left handed aren't too upset about the location of the steam arm...

I'm not sure about the brew controls as I'm used to a sexy little levetta. I'll miss the sexy little levetta, but you cannot have your cake and eat it too.

The heat thing sounds a little scary. So will we have an article pretty quickly after the GS3 reaches market called "How to Insulate the Boiler of your GS3?" We get hot enough up here in the woods of Georgia without more electronics in our home putting our AC in overtime.

Interesting situation with the temp gauge. I guess taste is the ultimate judge and I'll need to keep that in mind.

Just out of curiosity... What size (diameter) is that portafilter?

Ann