#navbar-iframe { height:0px; visibility:hidden; display:none; } Espresso Passione Espresso Passione: Grinder analysis

Wednesday, September 20, 2006

Grinder analysis

I had an opportunity to be a part of an experiment where the the grind of four coffee grinders was analysed. We used the same blend from the same roasting batch of coffee ground on the same day with four different grinders at roughly the same fineness (espresso). First 20 grams out of each grinder was discarded and the sample was taken from the ground coffee thereafter. The ground coffee was then run through an industrial particle analyser to see the coffee particle size distribution in each sample.

I do not know how conclusive the results are scientifically, but nevertheless, I found this quite interesting. One caveat, though. The machine used for the analysis was only able to analyse particles smaller than 0.6 mm, therefore cutting out the larger sizes. However, I still think that the results give an idea of the differences. To the results, then:

Ascaso (Innova) i1 flat burr
Ascaso (Innova) i2 conical burr
Mazzer Mini E
Mahlkoenig K30
Looking at the graphs, the first thing I notice is that the flat burr grinders have similar particle size distribution while the conical burr grinder does really stand out. Taking a closer look and comparing the flat burr grinders, one can see that the Ascaso i1 has more fines as well as a wider distribution in the size of the larger particles. The Mazzer and Mahlkoenig have a more uniform grind, the Mahlkoenig very slightly more so, but on the other hand it aso has more fines than the Mazzer. As fas as I understand, you do need some fines to hold the puck together and achieve a balanced cup, so is this a good or a bad thing? I guess this is something that the graphs won't tell us and we will need to evert to our tastebuds :)

The graph from the Ascaso i2 conical burr grinder is a surprise. The grind is a lot less uniform than that of any of the flat burr grinders. Logic would tell me that this is a bad thing and hitting the targeted optimal grind would be more difficult. On the other hand, you might get in the ballpark more easily - i.e. the grinder would be more forgiving? This is just speculation and I may be off in my analysis. After all, aren't conical burrs supposed to be better due to the longer cutting surface of the burrs etc? Perhaps the wider particle size distribution would bring more complexity to the taste in the cup?. I hope to have an opportunity to try and actually taste the difference in the near future... ...it would also have been interesting to see the graph from a higher quality conical burr grinder for comparison...

Like I said, this was all in good fun and interesting, not a serious scientific experiment. I'd be interested in hearing any comments you may have, though. Oh, I'd also like to take this opportunity to thank Jesper, Michael and especially JP for making this happen.



Anonymous Lukas said...

Wow, thanks for this one, Teme. Very interesting! Finally somebody casted some light into this discussion :)

Thursday, September 21, 2006  
Anonymous rice said...

Great post Teme, thanks for sharing.

Did you analyse multiple samples from each grinder? Or was it one batch?
Would it be better to go with a single origin bean? That way the variation of bean density and moisture content between batches can be kept to a minimum.

Just a suggestion on my part, by no means a put down and I hope my attitude towards your testing doesnt come across as negative. I'm looking forward to more analytical posts.


Thursday, September 21, 2006  
Blogger Teme said...

Thanks guys.

Rice, we used just one sample from each grinder, but made sure that the coffee used and day that it was ground was the same. Now the latter of these probably does not matter in this type of a test. Interesting and probably a very good idea about using a single origin. We just went with the coffee (blend) that we usually use and didn't give it much more thought. Then again, this was not a serious scientific experiment...


Thursday, September 21, 2006  
Anonymous NeilA said...

Nice post, thanks.
Another thought: I wonder if anyone has looked at the issue of grounds being heated by the grinding process, and how much that affects the end result...

I guess you can keep on finding more things to test and wonder about! :)

Thursday, September 21, 2006  
Blogger Teme said...

There was a nice article on grinder in the most recent Barista Magazine. I think that heat is definitely an issue in a heavy commercial environment. However, I also think that it is not a concern for a home-barista who grinds for one or two shots at a time - there are usually other things to focus on (such as achieving decent distribution).


Thursday, September 21, 2006  
Blogger rice said...

Teme, if you still have access to the particle size machine, hows abouts a test to ascertain the variation in grain size with regards to weight above the burrs (i.e. amount of beans in the hopper)?

re: heat
would more heat be built up by the home-barista, using a doserless grinder, who pulses the grinder and distributes? The stop-start shearing action on the beans?

Friday, September 22, 2006  
Blogger Jason Haeger said...

Regardless of these results, I was quite shocked at how small the Ascaso burrs are.

They're tiny, and at 700rpm's, with those tiny burrs, I found it took close to a full minute to grind for a double. (~50 seconds.. dpending on dose)

I also found that the grind consistency wasn't even as good as a Gaggia MDF.. an entry-level home machine.

Not good.

This thing's going back.

Saturday, September 23, 2006  
Blogger Teme said...


I cannot comment on the pulsing in terms of heat, but I do not think it would be an issue if you just do a shot or two. Then again I do not do this and I'm not sure why I would.

With regards to the grind quality depending on the amount of coffee in the hopper (or other weight), it may be a variable and an interesting though. Unfortunately, I do not have personal access to the analyser and I do have to limit the number of experiments that will be run going forward (if any).

By the way, could anyone give me a price on such an analyser? If I get one, I'd be happy to run all the analysis people want :)


Saturday, September 23, 2006  
Blogger Teme said...


Interesting comment. I do not have exhaustive experience on the conical Ascaso, but based on your comment (and that of Jim Schulman on CG), I am definitely interested. I he to get some more first hand experience on this grinder over the weekend. The grind quality doesn't look too good on the graph here and the grinder is actually almost new. Hmm..


Saturday, September 23, 2006  
Blogger Jaime van Schyndel said...

Interesting stuff.

Friday, October 27, 2006  
Anonymous nobbi4711 said...

In my experience the I-2 creates a very forgiving espresso. Compact, not very bitter, a mostly smooth taste. But the different notes of a blend are better shown by a I-1 or Mazzer (or big conical grinders).

Ralf (espressosorten.de) and me did several testings of grinders, including Mazzer Super Jolly, Compak K3/K6/K10, several Ascaso, Obel, Mahlkönig etc...the more precisly a grinder is built the more different notes the espresso will show.

Greetings \\//


Wednesday, January 17, 2007  
Blogger Teme said...

Hi Marcus,

Thanks for sharing. You must have had quite a coffee buzz after testing all those grinders! I assume you didn't do them all in one testing session ;)

I personally haven't had the opportunity to do side by side taste comparisons in a controlled environment and with various grinders. The only exception has been the Mazzer and the Rocky - and the difference between these was minimal in my opinion (fairly good flat burr grinders both of them). It certainly makes sense that the precision of the build - especially the rigidity of the burr carriers and the quality of the burrs - would make a difference. I hope to have an opportunity for such testing in the future.


Thursday, January 25, 2007  
Blogger Jakub said...


This is all very interesting. Could you explain what the difference is between the black and gray colors in the graphs? What do they represent?


Thursday, December 09, 2010  
Blogger Teme said...

Hi Jakub,

Sorry for the very late response.

The black graphs are the coffee particles describing the particle size distribution of the grinder. The grey area of the graph is a reference substance and for the purposes of this (unscientific) comparison it is irrelevant in my opinion (I did not run the tests myself but so I understand).


Thursday, March 03, 2011  

Post a Comment

Links to this post:

Create a Link

<< Home

Web www.coffeegeek.com