Titan Grinder Project - Page 15

[Nick]

But in the particle analysis and SEM analysis of comparably adjusted grinders, there was no significant delta between burr sets. We thought we would see an increase or decrease in the fines (or other particle sizes) that would correlate to the burr set. At that point you would be able to look at a histogram of the particulate and know that the sample you are looking at is from a flat or conical burr. There would be a defining trend to the graph for each burr type, but there was not.

I'm being careful not to characterize/generalize too much... conical vs. planar or anything.

That said, on the contrary, the particle analysis information was very illuminating. "Significant" is a very relative term, and in the volatile environment of espresso-grind particle sizes, it doesn't take much to make a big difference.

I am, however, interested (as I said before) in further study.

[Ken Fox]

I'm being careful not to characterize/generalize too much... conical vs. planar or anything.

That said, on the contrary, the particle analysis information was very illuminating. "Significant" is a very relative term, and in the volatile environment of espresso-grind particle sizes, it doesn't take much to make a big difference.

I am, however, interested (as I said before) in further study.

That depends on what it is that causes the need for grinder adjustment. Folks assume that it's the coffee that changes, but we see here that conicals appear to behave differently than flat-burr grinders in their perceived need for adjustment. The coffee, however, stays constant in this (same type presumeably), which could lead one to the conclusion that the grinder may be moving around as well. Grinder housings are aluminum, which has a pretty high thermal coefficient of expansion. Depending on the design, it's entirely possible that burrs move relative to each other quite a bit when temperature changes. This would be due to motors being mounted in one part of the housing (near the bottom), while the upper burrs are a spaced quite a distance away, usually threaded into the top. Thermal expansion in grinders using a conical burrset is likely to be less of a problem than for flat-burr grinders for the reason that Nick presents. I can think of other sources of grinder instability that would be less of a problem for conicals compared to flat-burr. Motor shaft end float would be one, as would deflection under stress from cutting the coffee.

-Greg

I think we all find this stuff interesting, however in the end the most important thing is the observation itself; that as a home barista who wants commercial equipment in the home, that with a conical grinder one is not sacrificing anything in taste (which might even be "better,") PLUS one gets more repeatability, less sink shots, and less need to fiddle with grinder settings in order to make good espresso with a conical vs. a planar grinder.

ken

[cannonfodder]

That said, on the contrary, the particle analysis information was very illuminating. "Significant" is a very relative term, and in the volatile environment of espresso-grind particle sizes, it doesn't take much to make a big difference.

I am, however, interested (as I said before) in further study.

I need to go back and read the thread again and look at my notes. I thought we came to the conclusion that there was not enough difference between samples to account for the changes, but I could be wrong.

[RapidCoffee]

I need to go back and read the thread again and look at my notes. I thought we came to the conclusion that there was not enough difference between samples to account for the changes, but I could be wrong.

You're right, it's probably premature to draw any conclusions from these preliminary particle size distributions. But if you look at the first set of graphs (wet analysis: particle volume), there's a definite trend. The flat burr Mazzer SJ has the highest, narrowest large particle peak and the fewest fines, followed by the Robur. The Macap MXK has the lowest, widest large particle peak and the greatest amount of fines. Does this size distribution correlate to forgiveness of grind settings and/or taste in the cup? In terms of taste:

The conical burr MXK pours are sharper, edgier, with more cleanly defined individual flavors, and more of a bitter finish. The flat burr Super Jolly pours are softer, rounder, more blended, sweeter, and more chocolatey. The Robur lies somewhere in between. I've been sampling each pour as straight espresso, then with sugar, and then as a cappuccino. These tastes come through quite clearly in milk.

In terms of forgiveness: I found the conicals to be more forgiving of grind setting than the SJ. But I honestly can't remember if one was notably more forgiving than the other.

Obviously more work is needed. I'd love to see grinder manufacturers supplying particle size distribution plots with their grinder specs. The typical home barista may have trouble gaining access to a laser diffractometer!

[cannonfodder]

Remember when we use to drink coffee versus shoot it with lasers and electrons?

[RapidCoffee]

Remember when we use to drink coffee versus shoot it with lasers and electrons?

You mean, back when it was fun instead of work? Sure do...

[AndyS]

For our use - folks using commercial gear in a light duty environment, temperature issues are pretty much moot.

Grinder housings are aluminum, which has a pretty high thermal coefficient of expansion. Depending on the design, it's entirely possible that burrs move relative to each other quite a bit when temperature changes. This would be due to motors being mounted in one part of the housing (near the bottom), while the upper burrs are a spaced quite a distance away, usually threaded into the top. Thermal expansion in grinders using a conical burrset is likely to be less of a problem than for flat-burr grinders for the reason that Nick presents.

Greg, I'm not sure how these two statements fit together, but I'm intrigued with your idea that moderate temperature changes might cause the grind setting to change more in some grinders than in others. Because I like to put numbers to things when possible, here's a little chart that takes a SWAG (scientific wild-ass guess) at comparing how the temperature of the grinding chamber is affected by grinding one 13 gram dose in a Mazzer Mini and in a 3-phase Robur.

This chart is partially based on the ideas that I struggled with here and here, which involve a lot of speculation. But I suspect it is somewhere in the ballpark. It summarizes variables like how much energy is consumed, how much is converted into heat, how much is absorbed by the grind chamber, the mass of the grind chamber, and the material characteristics of the grind chamber.

The gist of it is in the last column: grinding a single small dose in the Mini potentially heats up the grinding chamber four or five degrees F more than it would in the Robur.

Please feel free to critique (or savage ) this analysis.

Based on temperature data taken in the chutes of the Mazzer Kony vs. SJ grinders, there was no compelling reason to believe that temperature plays a major role in grind quality, at least not for these grinders under typical home conditions.

One more piece of speculation here, and then I'm done: coffee beans are subjected to high temperatures (400-500F) for long periods of time (many minutes) during the roasting process. Unless the beans are heated to much higher temperatures during the much shorter grinding process, this argues against heat being a major factor. Grind particle properties (size, shape) are more likely to play a significant role in determining pour characteristics, especially when evaluating grinder forgiveness (not taste).

John, I'm not saying that temperature plays a major role in grind quality. But it is possible that, as Greg says, temperature plays a role in grind adjustment. And it is also possible that grind chamber temperature plays a role in drying out the partially ground dose that sits in the grinder between shots. Both of these mechanisms might help explain why the bigger grinders seem to need less frequent adjustments than the smaller grinders.

I think you'll agree that comparing the cutting surface length of conical grinders vs. planar grinders may explain differences in grind quality, but does not explain differences in grind repeatability.

Also, you make a good point that any inadvertent "heat treatment" in the grinder pales compared to what the coffee has already experienced in the roasting process. But in roasting, the beans are whole. There is a HUGE difference between what happens with whole beans (miniature pressure capsules) and with ground beans (finely subdivided with cellular structure partially destroyed). I bet if we compared the result of roasting whole beans vs finely ground green beans using the same input temperature profile, the differences would be profound. It is similar to the shelf life comparison of whole roasted beans (10-14 days) vs ground roasted beans (10-60 minutes?)

[AndyS]

The people that I know of who have made this observation online would very seldom grind three shots in succession (Andy's threshold), and this is unlikely to be their normal usage pattern, although it might happen on occasion. The people who I recall making this observation (of the need to change grind adjustments less frequently with conicals) include Dave, myself, Jim Schulman, Jason Casale, and to some extent Jeff.

FWIW, when I got the Versalab grinder (with hybrid burrset) I frequently compared it to the Mazzer Mini. I didn't notice a difference in adjustment frequency between the two.

When I got the Robur, I was so taken with it that I stopped using the Mini entirely. I pretty much use the Versalab for presspot only. So I have no real data. But speaking very subjectively, the Robur does seem awfully forgiving.

It has always seemed comical to me that Schomer calls the hybrid burrset "conical," when that's only half true.

The comical conical.

[Matthew Brinski]

And it is also possible that grind chamber temperature plays a role in drying out the partially ground dose that sits in the grinder between shots.

I've never considered that (which makes me feel stupid - wait, I AM stupid), but that is a really good point. That idea seems to have just as much, if not more, of a potential affect than the changing ambient humidity.

[Matthew Brinski]

To the best of my knowledge, nobody is claiming heat issues have been categorically excluded from grinder evaluation. If you have criticisms of the data that was presented, why don't you run your own experiments and generate new measurements for the rest of us to chew on? Further armchair quarterbacking is optional. :P

I'm sorry. I forgot that I am only supposed to contribute dialogue that is in line with popular opinion.


.

[gscace]

John, I'm not saying that temperature plays a major role in grind quality. But it is possible that, as Greg says, temperature plays a role in grind adjustment. And it is also possible that grind chamber temperature plays a role in drying out the partially ground dose that sits in the grinder between shots. Both of these mechanisms might help explain why the bigger grinders seem to need less frequent adjustments than the smaller grinders.

I think you'll agree that comparing the cutting surface length of conical grinders vs. planar grinders may explain differences in grind quality, but does not explain differences in grind repeatability.

Also, you make a good point that any inadvertent "heat treatment" in the grinder pales compared to what the coffee has already experienced in the roasting process. But in roasting, the beans are whole. There is a HUGE difference between what happens with whole beans (miniature pressure capsules) and with ground beans (finely subdivided with cellular structure partially destroyed). I bet if we compared the result of roasting whole beans vs finely ground green beans using the same input temperature profile, the differences would be profound. It is similar to the shelf life comparison of whole roasted beans (10-14 days) vs ground roasted beans (10-60 minutes?)

In the first instance I was talking about heat produced by burrsets. In the second instance I began to think about changes in ambient temperature possibley having an affect. Your quick little table is quite interesting. I'll have to look at it some more.


At the WBC machine trials, i had intended to install thermocouple probes just above the burrsets, just downstream of the burrsets, on the housings near the burrset adjustment threads, and one in the air nearby (ambient). I wanted to see how much things changed relative to ambient conditions when the grinders were used more or less heavily. Unfortunately we were so busy that I never got to do it. It would have been interesting to me because I think that thermal expansion of materials prolly plays a big part in stability of grinder adjustment, or lack of it.

-Greg

[AndyS]

Your quick little table is quite interesting. I'll have to look at it some more.

Uh-oh. I'm in trouble now.

[Matthew Brinski]

Andy,

I just got home and finally had a chance to look at your chart. I think I pretty much understand how everything correlates except the (sp ht). Is that the specific heat gain of the grind chamber? where does the 0.09 come from? I'm an idiot.

[AndyS]

I just got home and finally had a chance to look at your chart. I think I pretty much understand how everything correlates except the (sp ht). Is that the specific heat gain of the grind chamber? where does the 0.09 come from?

0.09 is the approx. specific heat of brass. So it takes 0.09 btu to raise the temp of 1 lb of brass 1 degree F.

[gscace]

0.09 is the approx. specific heat of brass. So it takes 0.09 btu to raise the temp of 1 lb of brass 1 degree F.

It's prolly not the brass that is so interesting. It's the aluminum housings, which expand / contract much more with changes in temperature.

-Greg

[AndyS]

It's prolly not the brass that is so interesting. It's the aluminum housings, which expand / contract much more with changes in temperature.

I just put in a bid on an eBay Handbook of Chemistry and Physics so I can look up specific heats and coefficients of expansion and all that stuff.

[Nick]

Andy, your recent posts in this thread make some sense, except for one thing: In a busy coffeebar setting Roburs get really hot too. REALLY hot!

But when they do heat up, they still don't require as much adjusting as a flat-burr grinder. For your theory to hold water, Roburs should, once they heat up, suffer from the same sort of symptoms... but they don't.

I still go back to a simple fact: Roburs grind coarser than Super Jollys. Look at the grinds. Feel them. It's right there.

[Nick]

It's prolly not the brass that is so interesting. It's the aluminum housings, which expand / contract much more with changes in temperature.

But if it were indeed the aluminum burr carrier that was expanding and contracting, heat should cause the grinder to grind finer, as it heats up... which means you'd need to adjust it to grind coarser.

On a Robur, in my experience, once it heats up, you actually need to adjust it to grind finer to maintain flow rates. The obvious reason, which may or may not be true, is that the coffee is a little dry-er coming out of a hot burr set.

I really have a hard time believing that metal expansion has anything to do with anything.

[AndyS]

I still go back to a simple fact: Roburs grind coarser than Super Jollys. Look at the grinds. Feel them. It's right there.

Nick, if you adjust a Robur and an SJ to give approximately the same flow rate with the same coffee, it'll be easy to see that the Robur grind contains more coarse particles? That's really surprising to me, even given the results of John's laser particle-sizing experiment. His data seemed to indicate that the differences were real, but that they'd be very difficult to see with the naked eye (but not, perhaps, with the naked portafilter).

[gscace]

But if it were indeed the aluminum burr carrier that was expanding and contracting, heat should cause the grinder to grind finer, as it heats up... which means you'd need to adjust it to grind coarser.

On a Robur, in my experience, once it heats up, you actually need to adjust it to grind finer to maintain flow rates. The obvious reason, which may or may not be true, is that the coffee is a little dry-er coming out of a hot burr set.

I really have a hard time believing that metal expansion has anything to do with anything.

My thinking is that the housing grows more than the motor and motor shaft with increasing temperature, so the upper burr carrier moves away from the lower burr. Your observation supports that assumption.

-Greg