Versalab M3 Grinder - Page 3

[AndyS]

Thanks for the scale Jim, nice Job. I created a similar one using word, but I think a more appropriate one would be a scale printed on translucent tape. I'll see what I can do about that.

Good idea, print it on this:
http://tinyurl.com/abxaj

...although it's not self-adhesive.

[Abe Carmeli]

Good idea, print it on this:
http://tinyurl.com/abxaj

...although it's not self-adhesive.

Yeah, I was just thinking about that. I'll stop by Office Depot today to see if they have any self adhesive transparencies.

[Abe Carmeli]

DAY 4

Today I was trying to determine the optimal Barista technique for the Versalab M3 grinder. Let's see some pictures:



Uneven distribution half way into the grind

The distribution out of the grinder is not always perfect as we see in the image above. (I am nitpicking here, of course). The photo was taken half way into the grind of a 14 grams basket. This does not happen on every grind but on about 50% of them. I found that a good way to fix it is simply to turn the basket or P/F 180 degrees halfway into the grinding session.


Uneven distribution out of the grinder - Full Basket

Dosing

An under-dosed basket

The above picture shows an under-dosed basket. Leveling at this point will leave the center of the basket less dense and may cause uneven extraction.


A good dose

You don't want to get it any fuller than the above picture.

Leveling

Moving to the Barista's leveling technique, I produced the best results using a 1,2 North South flat sweep across the rim with no downward pressure at all. Avoiding downward pressure is key. Since the M3 does not produce any clumps, there is no need to apply downward pressure to break them, or to redistribute the grounds. You want to move the mound and keep its density as it fills up the little crater left in the middle.


Leveling the puck: one flat sweep north to the edge, and one south


I call it Schomer's Crater. Don't do that.

The above image is a result of bowing your finger when you level. It creates uneven density - the center of the puck is more dense. This will slow down the extraction from the center. That distribution technique which works with all other grinders, does not work well with the M3. The grounds are already at a state of almost perfectly even density right out of the grinder, and the Barista's job is to mess as little as possible with it.

My preferred method of leveling - no downward pressure

I found that tapping the p/f straight down against the counter, halfway or 3/4 way into the grinding session allows me to control the final dosage amount and still keep even distribution.

This is not the only way to work with this grinder, but I found it to be optimal for my usage at this point.

[Abe Carmeli]

DAY 5

After shamelessly drooling all over the Versalab M3 for four days, today is judgement day for that little engine that could.

The Shortcomings

Where is My Dial Scale?

Do you know of any grinder without a dial scale? Do you know of any $1,2500.00 grinder without a dial scale? Do you know of anyone who could not use a dial scale? (These are rhetorical questions, and if you do, please do not spoil my rant here ). It is a serious flaw and I must say, inexcusable. Can you fix it by creating your own paper scale and attaching it to the dial? Yes, kind of, but it looks hokey, it is prone to get dirty with grinds, and it is just unacceptable.

Can you Hold that Dial for Me?

This one is a first for me. The Mazzer Mini, and many other grinders have their dial spring loaded, so the threads are always under pressure. This makes it easy for the dial to hold its place, but more importantly, to consistently repeat its performance after a change. In other words, if I dial 3 to grind Harrar then move to dial 4 to grind Cerrado, and then back to 3 to grind Harrar, the grinder will accurately repeat the grind size for the Harrar, and my shot time will be as it was the first time around.

That little convenience makes all the difference in the world if you are a home user and enjoys drinking variety of coffees at a time. Creating a dial stop chart for each coffee saves tons of coffee and frustration. The M3's dial is not spring loaded, and it appears that the wiggle room between the threads is causing the dial to lose its consistency. I've waisted a lot of coffee that way. The dial locking knob pushes a little rubber "break" against the threads, and if you do not consistently tighten it at exactly the same pressure, you will get the dreaded dial all over the place.

These are the big two shortcomings on my list. Can they be fixed by us users? I invited Sean Lennon to Join this thread, and show us the light. Sean is a Robotics Engineer with 20 years of experience, and a fantastic contributor to the coffee community. He is going to post an interesting solution to both problems later today, so stay tuned.

[another_jim]

Can you fix it by creating your own paper scale and attaching it to the dial? Yes, kind of, but it looks hokey, it is prone to get dirty with grinds, and it is just unacceptable.

I'll have you know my strip is coated in plastic; and it looks a lot less hokey than the bubblewrap basket holder:

[another_jim]

Abe, Andy -- I have an odd request, please bear with me. I'm spending my time doubling up on all my espresso making, making a shot each from the mini and the m3, and tasting a bit of each. Since this is my daily consumption, I'm making macs and cappas as well. It's no sweat pouring the milk into both cups after taking the sips; so I've been comparing these too.

I don't want to give anything away until I've racked up about 20 to 30 comparative shots over the next week, but I'd like you to also try some milk drink comparisons if you normally make these.

[Abe Carmeli]

I'll have you know my strip is coated in plastic; and it looks a lot less hokey than the bubblewrap basket holder:

Jim,

Plastic or not, your strip's hokeyness remains intact and unchallenged

[lennoncs]

Hi all,
I was talking a bit to Abe today about the issue he was having with creep in the adjusting barrel after tightening the lock screw. I grew up in the family business of machine tools and precision tooling, so my take on statements of "precision grind adjustment" made by manufacturers may seem a bit harsh to some. I would only start to consider precision adjustments to be those that are able to be reproduced within a few tenths (.ooo1) time after time. I defer to the experts on grain size for optimum extraction to give an exact size of the grains that we want (any time now Jim )

I do not have an M3 to take apart and look at but,
Here are my observations to date:

What is the bearing configuration supporting the mainshaft (duplex pair? Angular contact W/preload?) And how stable is the burr from being moved off axis?

What is the composition of the "solder" holding the wiper together? (this is not nitpicking...some of us spent years using red-lead for gear rubs and are exposed to heavy metals too much already)

Why do we care if the angular displacement of the barrel is calibrated? The clearance in the threads makes this a ballpark figure anyways. What we really care about is the actual axial movement of the burrs relative to each other. Why not use a dial indicator on the top of the funnel? Or some other part that reflects this relative motion. You can buy precision in the form of a dial indicator for very little money that exceeds the precision of the machine that built the M3 (I am talking about a ~$50 investment).

Go out and buy a Starrett "Last Word" or similar unit and the appropriate one of dozens of clamps/brackets available to mount it and use it to read the ACTUAL position of the barrel not some approximation. Remember; a thousandth of an inch is a significant percentage of the particle size we are trying to reproduce.

I have included a really lousy job of approximating what I am talking about below (MS Word, sorry)


Cheers
Sean

P.S. Looks like a really nice unit, I am going to have to crash Abe's place to rip his apart

[lennoncs]

Federal makes a superb short range digital unit that allows the display to be rotated independently of the plunger. You can also set it for analog type display or digital with analog marks as shown below. I currently use one of these on my Mazzer SJ. I run the burrs down to contact and zero the indicator and the from that point I have an exact display of my actual burr clearance and all I have to do after re-assembly is repeat the zero procedure and I am set. No more fiddling for a number of shots for the perfect setting and I now have a chart that is accurate after complete disassembly for cleaning because it is an actual number.

Cheers
Sean

[Abe Carmeli]

I'd like you to also try some milk drink comparisons if you normally make these.

Jim,

I can't help you there, since I don't drink milk. But I'll be doing shot comparisons this week, and will wait for your report before I publish mine. I don't know anyone who can do as good a job at it as you.

[lino]

Hey Sean,

I don't have an M3, but I've been staring at the pics for the last several days...

Looking at the burr shaft, I was a little disturbed to see no lower bearing. That shaft is hanging way out there. The bearing housing (where I assume the bearing is, anyway) looks mighty thin too. Even if the upper pulley is hollow and the bearing is in there too, it's still not much room for two bearings, and they certainly can't be separated much. That doesn't leave much resistance to radial loads at the business end of the shaft. Even if they are good bearings with a mighty preload. Plus, those bearings appear to be housed in aluminium, making things even more flexible.

As a side note, the SJ uses two $3 bearings (size 6202, IIRC, shielded, not sealed), but they are 10" apart. They can't support much axial load, but there doesn't appear to be much...

One might ask, yeah, but when does the shaft experience side loads? (at least I asked that....)
One significant time that the case came to mind. When the last bean goes thru, there isn't (necessarily) an even distribution of grinds around the burr, side loads could be significant.

Also, does the shaft move or flex if you tug on it?

Anyway, just a thought....

The proof of the pudding, however, is "in the cup", as they say. And it seems to be doing quite well in that respect.

How long have people had these in use? I know you got yours recently Abe. How about Jim and Andy, how long have you had them?

ciao

lino

[another_jim]

I defer to the experts on grain size for optimum extraction to give an exact size of the grains that we want (any time now Jim

Not something most of us are up to measuring: here's a frequently reproduced graph (fair use, I hope) from Illy & Vianni showing grind particle distributions:

The horizontal scale shows particle diameters, The vertical scale shows percentages. The graphs show the number of particles, and the percentages of total volume and surface area at each particle diameter.

The graph's empiric distributions have been successfully simulated by finite element simulations of the grinding of hexagonal cells of simulated cellulose. These tend to break into fragments of single cells (fines) as well as particles consisting of multiple complete cells

The grinding chapter also addresses the underlying design of the M3 explicitly:

1: Grinding takes place in two phases - a crushing phase breaking particles to 1mm sizes, and a shearing phase reducing them to grinds.

2: "Cheap" home and coffee shop grinders use a single set of burr-wheels with unevenly spaced cutting teeth. "Industrial" grinders use either two rollers in sequence, or conical followed by flat burrs to do each grinding phase separately.

If you look at the mini's burrs, you'll see the "cheap" arrangement of widely spaced crushing burrs on the inside of the wheel, and finely spaced shearing burrs on the outside:

Abe's pic of the M3 flat burrs on page 1, shows they have evenly and finely spaced "shear only" cutters.

3: Industrial grinding equipment always finishes with a homogenization step that thoroughly unclumps and mixes the grinds.

I never read the chapter thoroughly before. My take is that the commercial burr grinders we respect are somewhere between substandard and barely adequate by the industrial criteria used by Signor Petracco, the author of the chapter. In contrast, the M3 has each of the three grind phases -- crushing, shearing and homogenization -- covered by an explicit design element -- the conical burrs, the flat burrs, and the sweeper-funnel arrangement.

Perhaps John Bicht read this chapter a bit more thoroughly than we all did.

[lennoncs]

Thanks for the info on the grinding, very enlightening indeed!

One of the customers I used to visit on a regular basis is a spice grinding concern...I almost wish they would have robot problems so I could check out their cryo-grinders for spices with my new-found interest in grinding. I remember relating to one of the plant guys how nice I found my Mazzer and his only comment was that coffee was another universe away from their grinding. I would like to talk to him more about that comment.

I see the majority in volume is particles of .019" followed by .0013" if I read the chart correctly, do you suspect this distribution holds true for the M3?

Sean

[lennoncs]

Hi Lino,

I probably need to temper my penchant for precision in some things but I do agree about the bearing spacing...the shaft might not get enough loads in operation to deflect much...I just dont know.
this is most likely a dead issue...

I remember making a cabinet years ago for my girlfriend out of wood...not a succsessful venture if you apply the same standards to wood that you do metal...the first change in humidity and she couldn't even pry it open!


Sean

[malachi]

You should look at pharmaceutical grinders.

I talked to an engineer working on these products. He said that (in his opinion) the current commercial espresso grinders are what was industry standard for his industry around 50 years ago.

Basically, he laughed at the design and described it as "one step up from big heavy rocks and a mule."

[another_jim]

I see the majority in volume is particles of .019" followed by .0013" if I read the chart correctly, do you suspect this distribution holds true for he M3?

Yes, it holds true for all grinders and according to Marino Petracco, who also wrote the brewing and tasting chapters (7 & 8 ), the bimodal distribution is required to get proper packing of a perculation mass (perculation is coffee scien-tese for any brewing process, including espresso, that sends water through the grounds). Think about packing with uniform spheres -- can't be done. However, several people have done measurements of the dust from different grinders, and the good and new ones produce less than old or poor ones.

A simple way of doing this, for those who have 1/10 gram scales, was proposed by Steve Ackerman. Brew the same weight of coffee ground for FP (using a pinch test), on several grinders. After you're done, run the grounds through a Swiss gold, then a paper filter. Dry and weigh the grounds in the paper to compare grinders dust output. This is very rough, however, according to Petracco, using a set of gradated precision sieves doesn't work repeatably for coffee, and laser diffractometry needs to be used to generate graphs like the one reproduced above.

I find my collection of tools required by this hobby frightening enough already, and don't want to convert to a lab. So I'll leave the grinds measuring, rough with paper, or smooth with lasers, up to someone else. I know what my tongue tells me, the M3 produces less dust than (my) mini; bank on it.

[Abe Carmeli]

P.S. Looks like a really nice unit, I am going to have to crash Abe's place to rip his apart

Anytime Sean, and if you do it within the next couple of weeks, I'll let you break apart the new espresso machine I'm testing too .

I love the meter solution. I'm going to look into one that hopefully will fit nicely on the M3. There are quite a few out there. I think an analog dial will go better with the design style of the machine, it is 18 century English style. But a digital meter would be more functional.

[AndyS]

Even if the upper pulley is hollow and the bearing is in there too, it's still not much room for two bearings, and they certainly can't be separated much....
...does the shaft move or flex if you tug on it?

I haven't disassembled it, but I was told by John from Versalab that there are two bearing in there.

How long have people had these in use? I know you got yours recently Abe. How about Jim and Andy, how long have you had them?

8 days. A long time, no?

[AndyS]

I'll let you break apart the new espresso machine I'm testing too .

Hope it's a good one.

[AndyS]

I don't want to give anything away until I've racked up about 20 to 30 comparative shots over the next week, but I'd like you to also try some milk drink comparisons if you normally make these.

OK, will do.