dsc wrote:I haven't but I don't know anything about the method used for getting those distribution graphs. Was this done using sifting or microscopy photography or a mixture of both?

Also was this done a single grinder with a VFD, something custom built or on different grinders with different burrs sets?

Presumably this is laser diffractometry, done on a single grinder (as noted by the OP). If they used different grinders, then any comparison of the three distribution curves is worthless. No idea how they varied the rpm, or what slow/medium/high speed means.

If they only swapped between 60Hz and 50Hz, this is a 20% change on the motor speed (less / more cooling etc.).

Not knowing how the distribution was checked and what speeds the grinder was running at makes this a bit useless imho.

Regards,
T.

I wonder if burrs are designed to run at a certain RPM. That is, they are optimized to produce the best grind for espresso within a certain RPM (tight?) range. (I don't know anything about burr design and so I may be totally wrong here.)

Paul

another_jim wrote:but you need to compare low and high speed grinds that actually flow the same at the same dose. I very much doubt the illustrated grinds do that.

Yes, and even if it turns out that slower speeds do produce more unimodal distribution, we still have the unanswered questions of whether unimodal distribution results in higher extraction yields and whether those higher yields generally improve taste in the cup! EK43 all over again.

It should be noted that two of the most lauded big conicals among members of this site, the Mazzer Robur and the Compak K10, operate at relatively slow speeds compared with other grinders. Yet the Robur is accused of producing low extraction yields due to the greater variation in particle distribution compared with the Mahlkonig EK43 and even the Mahlkonig K30. But the Robur's RPM is about 70% lower than those two grinders, and if low speeds produce more unimodal distribution...

Well, you can't have it both ways!

The burr action will be the same regardless of speed (provided the motor horsepower changes in proportion).

The beans themselves are another story. Every bean experiences a decreasing gap between the burr walls as it travels through the burrs. As the space decreases, the bean fractures; and then the pieces fracture, etc, etc. Faster rotation means the bean passes through more quickly. I have no idea if that will alter the fracture properties or not. These type of materials engineering questions remain among the most difficult and intractable in the sciences, since there are no general mechanical models for answering or even thinking about them.

Peppersass wrote:Yes, and even if it turns out that slower speeds do produce more unimodal distribution...

The plot indicates that slower rpms produce coarser grinds. Coarser grind settings invariably produce less fines, and a more unimodal distribution. The low rpm grind setting, with its peak at about 1mm, needs to be adjusted much finer for a decent espresso grind. This finer grind setting will produce more fines and a more bimodal distribution, regardless of the rpms. Before even thinking about extraction yield, you need to adjust the grind setting along with the rpms to produce a comparable particle size distribution.

The more interesting question, in my mind, is whether the finer setting low rpm grind and the coarser setting high rpm grind turn out to be significantly different.

RapidCoffee wrote:The more interesting question, in my mind, is whether the finer setting low rpm grind and the coarser setting high rpm grind turn out to be significantly different.

Yeah, that's the interesting question. We need a test.

I don't know enough about AC motors to know whether its safe to test this with a variac. Could lowering the voltage risk higher current draw as the burrs meet resistance, and could that damage the motor windings?

Isn't one problem with such a test that there may be other differences between fast and slow RPM, such as heating of the grinds? I recall reading posts that this may explain differences between the slow-RPM big conicals and fast-RPM big flats.

I have seen no evidence that grinder-related heating is of any consequence in a home environment. And anyway, it sure would be nice to have more info before worrying about ancillary issues. Nice plots, though: http://bwissue.com/xe/grinder/25753

Peppersass wrote:I don't know enough about AC motors to know whether its safe to test this with a variac.

I'm not an EE; but without one's guarantee, I'd stay away from it. At the very least; I'd start the grinder first before turning the Variac down, since you don't want to mess with the starting capacitor. But the Variac's inductance will still interact with the running capacitor. Larger AC motor's inductance and their running/starting capacitors are usually tuned to one another; adding the Variac will change that in ways that's in EE territory for figuring out.

Can you please post the link? Thanks
Joe