Why do bulk grinders produce a superior grind for non espresso preparation? - Page 11

#101: Post by **Bane** » April 13th, 2011, 3:18 am

another_jim wrote:I think the laser sizer works for the application for which it is most frequently used -- adjusting bulk grinders so the peak is at a given mm level.

i wouldn't say that this is the most frequent usage of laser diffraction particle sizer, but yes: for that purpose it works well...

another_jim wrote:The two distributions themselves (rather than the samples, which is the normal case for statistical testing) have an uncertainty band, and these will usually overlap. This odd problem of uncertain distributions affects all three of the data formats: count, area and volume.

of course it does. what i meant was that if the instrument is getting its data as counts and converting it into volume based data, you are multiplying the error, especially in the case of non-spherical particles.

i am not sure if it is exactly the same for laser diffraction methods (probably not, since measuring principles and calculation is quite different), but in dynamic light scattering you have an error multiplication when going from counts to volume by the factor of 10^3 and going from counts to number by the factor of 10^6.
because of that i would view volume and number distributions from those methods always critically...

#102: Post by **another_jim** » April 13th, 2011, 6:15 am

Bane wrote:but in dynamic light scattering you have an error multiplication when going from counts to volume by the factor of 10^3 and going from counts to number by the factor of 10^6.

Nice try, but sadly it doesn't work.

The test/retest variance defines an error band. So instead of the grind distribution being a line, it's a ribbon. The ribbon is so wide that the other grind distributions we were comparing fall inside it, and hence the comparison did not tell us anything.

Now, if you transform that ribbon and all the lines inside it, each in exactly the same way, say going from volume to counts, the transformed lines would still be inside the transformed ribbon, and the data still tells us nothing. The count ribbon may end up being a million time narrower than the volume ribbon, but the other distributions, turned into counts, would still be inside that much smaller interval.

For instance, a 1/2 lies between 1 and 1/4; transform the three numbers the same way, say divide each by minus one billion; or take their log, then divide those by a billion. The transformed 1/2 still lies between the transformed 1 and 1/4. Same for all the numbers in these distributions.

#103: Post by **Bane** » April 13th, 2011, 9:36 am

another_jim wrote:Nice try, but sadly it doesn't work.

umm... ok, i somehow phrased that wrong. i wasn't suggesting these facts it as a kind of solution for the variance problem, but stating it as a general problem of that measurement method and type data handling.
of course, if you can't see any difference in the original data sets, no kind of transformation will make that different.

as to that test/retest variance thing: that just might be due to the method's restrictments in terms of non-spherical particles (and maybe too low sensitivity in general)

as i said one might compare automated visual measurements which might give more consistent results.
(if that variance is due to measurement variance and not variance in the grinding sample itself)

regards
Georg

#104: Post by **Bane** » April 15th, 2011, 3:34 am

to go to more practical matters:
has anyone here ever tried that Mazzer Major bulk grinder burrset :

FMA00048D/QQQ Mazzer Drogheria (outer diameter 83mm, inner diameter 44mm, whereas major espresso burrs have an inner diameter of 49mm)

i am asking since Majors seem to be quite common in the US (you don't see them that often in Germany) and getting that burrset for a Major might be the cheapest option to get a nice bulk grinder. (if that burrset is indeed constructed for monomodal size distributions)

regards
Georg

#105: Post by **another_jim** » April 15th, 2011, 4:00 am

It's a good idea. One possible problem that should be checked first is that the buyers of preground coffee in Italy are mostly using it in mocha pots; so the burrs may be engineered for a finer grind setting than Ditting or Bunns.

#106: Post by **Bane** » April 15th, 2011, 4:15 am

yep, that might be a problem. despite intensive search i haven't found a decent picture of that burr set.
maybe a interested Major-owner will give it a try and post the results...

#107: Post by **Bob_McBob** » July 13th, 2011, 7:18 pm

Coming back to this thread after a 9 month hiatus. Over the last year I've been getting more and more interested in pourover and other regular brewing methods, to the point that I am actually drinking less espresso than a year ago. I am pretty interested in buying a bulk grinder for regular brewed coffee.

My original thought was to pick up a cheap Bunn G-series and either buy a new set of burrs, or replace them with the Ditting machined burrs. I monitor Craiglist and Kijiji, and most of what shows up is the larger G2 and G3, and they are often rather old and not particularly cheap. Unfortunately, the prohibitively expensive shipping cost from the States (as well as lack of vendor interest) precludes buying any of the used Bunns on eBay.

I found a good price on a new G1 (~$700), and I am strongly considering that option. Unfortunately, it looks like the burr swap is off the table, though. The Ditting burrs cost about $450 now vs. $200 for the Bunn burrs. Has the price gone up considerably since the thread was started?

The other option I've considered is spending $1000+ on a refurbished Ditting KF804 from Sammy Piccolo's store. I am unclear on whether I would typically have to replace the burrs with a refurbished Ditting, and how much of an improvement it would be over a G1 with brand new burrs.

I am having a tough time deciding which option I should go for. If I had $2000 to blow on this, I would probably just get a brand new Ditting. I gather the Bunn is a well-respected and solid grinder, but I've read posts claiming it isn't much better than than the Vario I already use. There seems to be a general preference for the Ditting's grind quality in high-end cafes. I would appreciate the finer grind adjustment of the Ditting. I am also not sure which grinder is better in terms of retention, which is rather important because I would always be single dosing.

#108: Post by **Warrior372** » July 13th, 2011, 8:43 pm

Peruse craigslist for a the next few weeks. I occasionally see Dittings / Mahlkonigs for around $500.

#109: Post by **Bob_McBob** » July 13th, 2011, 9:29 pm

I am in Canada, so the used selection is rather limited. I've seen exactly two used Dittings on Craigslist and Kijiji in the last year. The first one (a KR804) I was going to buy, but the owner decided to keep it at the last minute. The second one looked like it was about 30 years old, so I passed. Used Bunns come up somewhat regularly, but usually LPGs and the like or G3s. The few G1s I've seen are generally expensive and with owners unwilling to ship them even within the province.

#110: Post by **JohnB.** » July 13th, 2011, 10:28 pm

Who quoted you $450 for the Ditting burrs? Did you contact Ditting directly for pricing?