Buyer's Guide to the Gaggia Achille - Page 12

[cannonfodder]

Yes, but in Jim's humpbacked whale of a profile, the temperature does not spike into boiling water territory.

The temperature profile is a bit deceiving. I can assure you that the shots it produces taste much better than that graph suggests. Going strictly by the numbers, the shots the Achille produces would be a melange of sour and bitter swill, not the sublime cup it produces.

Like Jim's experience with his Elektra semi autos pressure settings, based on the numbers it should produce a harsh shot but instead it gleefully produces a wonderful elixir.

Dan has the uber geek espresso kit but I do have a Fluke 52-II two channel thermocouple thermometer and a Fluke 189 datalog meter in transit. I have to have a few toys. I was going to play with an over the lip and up through the puck thermocouple measurement when they arrive and I return home. I am out on business through the weekend starting tomorrow evening.

If you reduce the Pstat setting down to 1 bar you would get much more recovery time at the expense of steam. The shot in the video was too hot and went down the sink, but I went slow just to show the process for the video. I find the machine works best with lighter northern Italian roasts which tend to prefer a hotter extraction. My one dark roast was a challenge.

To answer hbuchtel, yes, adding cold water will draw out the recovery. That was a trick I noted in a previous post. When dealing with blends that prefer a lower temperature, I would fill the reservoir with water from the refrigerator. That tasted like it lowered the temperature by a couple of degrees. However, if you normally drink dark roasts that work best in the 198-200F range, lowering the Pstat may be a better option, or try the power off, flush and pull trick HB mentions.

[another_jim]

In a manual lever, one can use the lever to vary the pressure, or more importantly the flow rate, to compensate for the temperature. If the temperature starts low, a longer preinfusion will still transfer the flavors from puck to water, if it gets hot in the end, one can finish extra fast (or perhaps extra slow? I'm amazed to find myself without any intuition on this). These are adjustments any competent home barista will make the first time she or he takes the machine for a spin just by tasting successive shots.

[timo888]

In a manual lever, one can use the lever to vary the pressure, or more importantly the flow rate, to compensate for the temperature. If the temperature starts low, a longer preinfusion will still transfer the flavors from puck to water, if it gets hot in the end, one can finish extra fast (or perhaps extra slow? I'm amazed to find myself without any intuition on this). These are adjustments any competent home barista will make the first time she or he takes the machine for a spin just by tasting successive shots.

That one of Dan's "adjustments" was to turn the machine off during an extraction tells me that either his tongue or the data or both were telling him the machine's intra-shot temperature can sometimes get too hot for a good extraction, despite the pre-pull cooling flush.

My hypothesis is that the double-pump on the preinfusion is what may be causing the heating element to come on mid-shot, driving the temperature out of the sweet and into the bitter range. I would like to see a 26-second or 30-second plot where the lever is raised, depressed gently part of the way for the preinfusion ~5 seconds, and then, instead of raising the lever again to draw more water, press it down to pull the shot. This will reduce the volume of water, of course, and so the dose should be reduced accordingly.

Regards
Timo

[cannonfodder]

Just got home from a week of business travel. My bottomless portafilter is here, my K type datalog meter is here and my other Fluke 52 (J, K, T and E type two channel thermocouple meter) will be arriving soon.

I will try to shoot a couple of videos and I am going to botch together a two probe LM basket to try to get some in the puck measurements. The buyers guide editing is also in full swing so time is tight. I only have a week at home (which includes my daughters 12th birthday) and then back on the road for business.

[timo888]

Dave,
It's hard to tell from the videos, of course, but since you're no bantam-weight, when you really lean on the lever with your upper body, you are applying quite a lot of force to the piston and creating considerably more than 9 bars of brew pressure. I'd estimate you've been extracting at 11 bar maybe even 12 bar. The Achille's piston is relatively small at 44mm and yet the mechanical advantage of its lever is greater than the Cremina's, for example.

Achille Piston Diameter: 1.73 inches (~44mm)

Piston Surface Area: PI * (.865 * .865)

9 bar = ~ 130psi

130 psi = REQUIRED POUNDS OF FORCE divided by ( PI * (.865 * .865) )
130 psi = REQUIRED POUNDS OF FORCE divided by ( 2.35)
305 # = REQUIRED POUNDS OF FORCE

Judging from the length of the lever and the distance between the pins, and adjusting for the fact that your grip on the lever handle is not precisely at its extreme tip, let's say the lever gives you only eightfold or ninefold mechanical advantage instead of elevenfold.

305 divided by 9 = ~34 pounds of force required on the lever handle to produce 9 bar

It seems you're putting a lot more than 34 pounds on that lever.

If the lever's mechanical advantage is only eightfold:

305 divided by 8 = 38 pounds of force required on the lever handle to produce 9 bar

It seems you're putting quite a lot more than 38 pounds of force on the lever handle too.



Regards
Timo

[cannonfodder]

Looks can be deceiving. While I do lean into the handle, it is controlled. I am not doing an 800 pound gorilla one arm hanging in the air (BTW I am 210lbs not 800). When I originally received the machine I used a high tech bathroom scale to measure my pressure. According to it I am getting somewhere in the 40lbs range give or take 5lbs. I simply leverage my weight to obtain a smoother more controlled pull rather than relying on sheer strength.

Quite honestly, the exact amount of pressure is immaterial. I measured it at around 40 lbs, then tossed the scale back in the bathroom and adjusted the machine based on taste. Just like tamping, the 40 lbs mark is merely a suggested starting point not a rule. Each person will adjust their pressure based on blend, taste and physical ability. Just like the thermal measurement Dan made, which shows the intra shot temperature spiking, I don't always believe in the numbers. I can assure everyone that the end of the shot is not hitting 210+F. A bumble bee should not be able to fly given its large body to wing ratio, but it does. The Achille or Elektra semi auto should not make a good shots based on the numbers, but they do.

I can assure everyone that Dan and I have tried to present the machine accurately and openly. I have pointed out its strengths and weakness without bias and you the reader must decide for yourself if you trust our observations and judgments. Realizing that this review will effect many purchasing decisions I have thoroughly thrashed the machine to within an inch of its life to test the boundaries (read this as used hard but not abused).

The technical side of my vocation is interested in the science and mechanics behind the machine but the artistic side of my brain screams who cares enjoy the magic, and I am. While I do plan on tinkering with some measurements, I would not trust them beyond what I taste, and what I taste is good.

[cannonfodder]

Blonding sucks, but was only my second shot from that blend, it needs more dialing in. Diagnostics will be posted in the Videos of espresso extractions thread.

I promised the good bad and ugly. This one definitely started good, went bad then turned ugly.

[cannonfodder]

I had an interesting thought a couple of days ago while cleaning the group on the Achille. Since the machine works like a typical heat exchanger pump machine, why not do a portafilter wiggle and backflush of sorts?

Since you can partially open the group to relieve pressure, why not use a backflush basket with a little JoeGlo to spoof the machine into a pressurized wash. So I tried it while the machine was cold just in case things went awry. If you have ever been splashed in the face with 190+F water you understand, if you have not than heed my warning. Always try something new with a COLD machine, trust me, hot water in the face hurts.

So I took my backflush basket and popped it into my portafilter. I gave the lever a 1/3 stroke until it firmed up and stopped. Then I slowly rotated the portafilter until I was greeted with a whoosh of cold water. The water jets around the top of the portafilter and lug raceways in the group. The water runs down the portafilter and into the drip tray. Hmmmm, interesting.

So I repeated the above procedure with a hot machine. I added just a bit of JoeGlo to my basket, gave the lever a little tug to fill the blank basket with hot water (that dissolves the detergent so it flushes through the group better). Then I lock the portafilter into the group and tug until the lever stops. Give the portafilter a quarter twist and the hot water and detergent blows out some coffee residue around the portafilter and into the drip tray. I repeated the flush twice then finished with a portafilter wiggle. While holding the portafilter and partially inserting it into the group, I pulled the lever and wiggled the portafilter back and forth to further clean the gasket and raceway.

Wipe off the group, rinse out the blank basket and repeat with clean water to rinse. I would still remove the shower screen and dispersion block for monthly cleaning. This may work well as a weekly cleaning regiment to keep the machine fresh.

[another_jim]

I had an interesting thought a couple of days ago while cleaning the group on the Achille. Since the machine works like a typical heat exchanger pump machine, why not do a portafilter wiggle and backflush of sorts?

Um, because there's no 3 way valve to keep clean? If you want to clean this machine's entire waterpath, some cleanser dissolved in water in the tank along with some lever action will flush the entire system. If you do that with a machine that has a three way (boiler autofill disabled), the "forward-flush" would never get at the crud accumulating on the other side of the three way valve.

[cannonfodder]

I actually meant to call out the fact that it does not have a 3 way but thought that was understood so I did not.

I am not cleaning the water path as one would have to do with an E61. My thought was to flush some of the buildup out of the nooks and crannies in the group. If you use the partial portafilter removal method to depressurize after a shot (as I showed in a video in a previous post) a lot of grounds end up blown into the lugs and the little crevices around the shower/dispersion block.

In order to clean it, you need to take the screen and block off. If you flush/PF wiggle like this, you essentially power wash those nooks and crannies out. While not a substitute for removing the screen and block for a good cleaning and scrubbing, it does help cut down on the grunge between cleanings.

[cannonfodder]

In the pursuit of knowledge, I decide to take some temperature measurements on the Achille to see what is going on in the basket. These were taken with a type K thermocouple on a calibrated Fluke 189 meter. The bead TC was simply placed on top of the puck using the popular 'over the lip' placement of the thermocouple. I settled on an average rebound time between flush and preinfuse of 12 seconds simply because that was as fast as I could remove my flush container, put the TC on the portafilter, lock it into the machine and start the datalogger. The rebound is not exact, one shot may be 12 seconds the next may be 14 which may account for the slight variation in temperature from shot to shot. However the general temperature curve remains the same.

I used a standard dose of 17 grams, a three pull flush (to help compensate for the added time needed to fiddle with the meter and thermocouple), a half pull preinfuse which you can see in the stair step around seconds 4-5, followed by a full lever pull for a total of 1.75oz dispensed. I pulled three shots back to back pausing between each shot to key in the datalog numbers on the computer.

One big surprise was how low the temperature readings were. I will partially attribute the low reading to the thermocouple bead laying on the surface of the puck. Once I removed the portafilter, the thermocouple had slightly imbedded in the pucks surface. It was just enough to scrape the surface of the puck as I unlocked the portafilter. As a reference point, I pulled a single shot form my Faema two group for comparison. On that shot, the thermocouple did not make an impression on the puck. It rode between the pucks surface and the shower screen giving me a reading that I expected.

More shots will be needed. The next series I will shorten the flush to two pulls. Until then, here are my cursory graphs.

[timo888]

In the pursuit of knowledge, I decide to take some temperature measurements on the Achille to see what is going on in the basket. These were taken with a type K thermocouple on a calibrated Fluke 189 meter. The bead TC was simply placed on top of the puck using the popular 'over the lip' placement of the thermocouple. I settled on an average rebound time between flush and preinfuse of 12 seconds simply because that was as fast as I could remove my flush container, put the TC on the portafilter, lock it into the machine and start the datalogger. The rebound is not exact, one shot may be 12 seconds the next may be 14 which may account for the slight variation in temperature from shot to shot. However the general temperature curve remains the same.

I used a standard dose of 17 grams, a three pull flush (to help compensate for the added time needed to fiddle with the meter and thermocouple), a half pull preinfuse which you can see in the stair step around seconds 4-5, followed by a full lever pull for a total of 1.75oz dispensed. I pulled three shots back to back pausing between each shot to key in the datalog numbers on the computer.

One big surprise was how low the temperature readings were. I will partially attribute the low reading to the thermocouple bead laying on the surface of the puck. Once I removed the portafilter, the thermocouple had slightly imbedded in the pucks surface. It was just enough to scrape the surface of the puck as I unlocked the portafilter. As a reference point, I pulled a single shot form my Faema two group for comparison. On that shot, the thermocouple did not make an impression on the puck. It rode between the pucks surface and the shower screen giving me a reading that I expected.

More shots will be needed. The next series I will shorten the flush to two pulls. Until then, here are my cursory graphs.

At what point in these 35-second extractions, following a three-pull cooling flush, does the Achille's heating element kick on?
Regards
Timo

[cannonfodder]

Just got back from a business trip, spent a week on the dark side of the moon.

Normally, I would not run a 35 second extraction, but I ran the shot volume longer than I normally would just to see what happens.

I was not watching the heating element on those extractions so I can not say. The element was still heating as I started the shots. The three pull flush drops the boiler pressure to around .9 bar.

[cannonfodder]

I spent more time with the Achille tonight pulling shots and trying different timings and flush amounts. I also tried moving the over the lip thermocouple around but never got what I would call a reliable temperature. I believe the general temperature curve is a good representation of what is going on in the portafilter.

I settled on a pressurestat setting of 1.0-1.2 (or 1.98 but close enough to say .2). A two pull flush works best followed by an 18 second recovery. That was just enough time for the boiler to recover and the heating element to shut off. The half pull preinfuse is made just as the heating element clicks off. That allows me to get through most of the shot before the element kicks on at around second 23/24. At around 27 seconds I start to get a slow rise in temperature as the energized element starts to flash heat the water in the HX line.

This is the flattest profile I achieved from a dozen different flush/recover combinations. Don't be fooled by the low temperature reading, I can assure you this was not a 191F shot, but due to the over the lip on the puck thermocouple placement all of the temperatures read low. Do not view this graph as an absolute temperature profile, but as a temperature trend profile. I am confident based on my tasting that this profile is representative of the general temperature curve of the Achille.



Based on the rise at the end of the shot, using Dan's idea of allowing the machine to recover after the flush then powering off the heating element, you should be able to produce a nice, flat temperature profile. For darker roasts, allowing the machine to heat, then power off the heating element, do a two pull cooling flush followed by an 18 second recovery should produce a flat and cool temperature for those blends that work best in the sub 200F range.

For those of you wondering just how hard I have run the Achille over the past couple of months.

This is the new portafilter.


This is the portafilter after lots of coffee. I have worn a good bit of the chrome out of the PF.

[HB]

Don't be fooled by the low temperature reading, I can assure you this was not a 191F shot, but due to the over the lip on the puck thermocouple placement all of the temperatures read low.

That's a larger delta than I've seen with other over the lip measurements.

I'm reminded of one of the interesting discussions of the thermofilter Bench review that centered around the meaningfulness of temperature profiles. I've concluded they are excellent for tuning technique to increase reproducibility, but I don't think either the thermofilter or puck TC placement accurately captures the coffee's "temperature experience." Seeing the Achille's profile and thinking about its aluminum dispersion block versus others' brass dispersion block, I suspect that laying the TC across the top of the puck is measuring the dispersion block temperature more than the brew water temperature.

[cannonfodder]

I have used the over the lip TC on my Faema and Isomac and gotten readings that I expected based on taste and knowledge of the machines. But this Achille keeps me guessing as to why I get those low temperatures.

I wonder if I way under dose the basket (12g) and over the lip meter the results if I would get a more accurate water temperature. The mini puck would be providing just enough resistance to regulate the water flow while suspending the TC in a water cushion.

I may give that a try tomorrow just for fun. I have down dosed, but not that drastic.

[cannonfodder]

After a couple of months I believe I have pushed the Achille to its limits and exposed both its strengths and weaknesses. I have not had any surprises for a couple of weeks now and I can pull shots with consistent quality. The thermal measurements are the only thing that vexes me. For some reason I just cannot get a good temperature map. I am confident in the general temperature profile, but the numbers are nowhere close. She wants to keep at least one secret from us.

I will continue to use the machine for another few weeks until my time with it is up but do not anticipate any other revelations. But then again, if they were anticipated, they would not classify as a revelation would they.

So I will wrap up my 'official' bench review by simply saying that the Achille is a joy to use. The machine is quick to learn, more forgiving than most levers, superb materials in craftsmanship with one or two idiosyncrasies. Overall, it is a solid lever machine that produces a wonderful espresso.

One final parting video of a bottomless double ristretto...