Is single dose [without hopper] grinding inconsistent? - Page 2

[Ken Fox]

The same observation applies to any study, including your work comparing vibe vs. rotary pumps. I could point out that your results apply only to your Cimbali Junior espresso machines, whatever grinders you used for the study, and that particular batch of beans. Ditto for your more recent article on freezing greens. But those studies are still valuable, providing (at least) a starting point for future investigation. That's all I'm suggesting here.

John, I think this analogy doesn't really work. Particle size distribution, although very interesting, is an observation, not an end point. I would define the real end points here as shot quality and consistency, plus usability, the latter being almost entirely subjective. The interaction between particle distribution and the real end points are to my knowledge, as yet unknown.

The fact that particle size distribution on a SJ might be effected in XYZ fashion when the grinder is used hopperless, has no known relationship to what happens to the particle size distribution on a Kony or Robur when used hopperless, unless this has previously been defined, which I doubt. The risk is that real differences among the grinders may be obscured, because some of them function better hopperless than others, which might not relate to how they work with a bean load on top. This could easily give the appearance that there is no difference between the grinders when in fact, when operated with a bean load, there would be. And as before, the particle size distribution has an undefined relationship to the real endpoints.

In contrast, the studies I've done on Cimbalis and on Freezing were controlled experiments with balanced variables. Like any such study they are open to question and would need to be repeated by other investigators if the results are desired to be confirmed. I think it is a bit more likely that all freezers do more or less the same thing to coffee, whereas less likely that hopperless operation of grinders has the same impact on all grinders. But I could be wrong

All the Titan grinders are more than capable of producing excellence in the cup, so to some extent we're talking personal preferences rather than better or worse. Good grind quality and ergonomics are to be expected from grinders in the $1K price range, and I think the Titans deliver.

I'll be pleased if we reach consensus on a few general issues, such as fundamental differences (if any) between flat/planar burr grinders and conical burr grinders. (And perhaps dispel some of the recent hype surrounding conicals. ) The TGP testers (Dave, Jim, Ken, Dan, and I) have all noticed the high forgiveness factor associated with conical grinders, and that may be one of the most compelling conclusions of the study. Tastewise, there's less consensus - neither Dave nor I are rushing out to buy a conical just yet.

I appreciate the plug however I'm not on the TGP panel, and have only been pressed into service around the edges by Dan, and that long after this project was conceived and started.

ken

[HB]

When John and Dave kicked off the first leg of the Titan Grinder Project, I was surprised to see the grinders prominently displayed hopperless:





I cut down my Mazzer Mini hopper to 4" so it would clear the cabinets (this was before the "short version" Mazzer hoppers were available). At one time I would toss in a shot's worth and run the grinder until it was empty, but later I decided the extra steps when preparing multiple espresso weren't worth it. Filling the mini hopper with enough coffee for 4-6 espresso wastes more than precise one at a time measures, but it's acceptable to me.

To get a better idea of how many home baristas employ the "one shot, grind until empty" technique, I've added a poll "Do you dose and then run the grinder until empty for each espresso?" My apologies to Ken for "overloading" his thread.

[RapidCoffee]

John, I think this analogy doesn't really work. Particle size distribution, although very interesting, is an observation, not an end point. I would define the real end points here as shot quality and consistency, plus usability, the latter being almost entirely subjective. The interaction between particle distribution and the real end points are to my knowledge, as yet unknown.

My statement wasn't an analogy, it was a general comment on the nature of scientific observation. If you test how well one bean freezes, nobody can stop you from extrapolating your results to other beans, but you haven't tested them all. If you compare different pumps on one espresso machine with one type of boiler, heat exchanger, gicleur, group head, etc., you haven't tested them all.

Likewise, if I measure particle size distributions for weighted and unweighted beans on just one grinder, obviously I've only tested one grinder. But if the profiles are similar, I'd argue that this supports Jim's contention of little or no difference between pours. Since I had already planned on a day of particle sizing, and the question arose, I assumed a couple more samples wouldn't hurt. Naturally you are free to disagree with any conclusions I might draw, and if you feel strongly enough about it, to conduct blind taste studies of your own. (Fortunately, with only one espresso machine on my countertop, I am in no position to conduct such studies... )

But as for the remainder, you're absolutely right. Measurements of particle size distribution are of purely theoretical interest unless they can be correlated with results in the cup. That was the goal from the beginning. Measurements such as particle size distributions have the virtues of quantifiability and reproducibility. If meaning could be attached to these measurements, perhaps manufacturers might include particle size distribution plots with other grinder specs such as power, burr type, burr size, and rpm.

[RapidCoffee]

When John and Dave kicked off the first leg of the Titan Grinder Project, I was surprised to see the grinders prominently displayed hopperless...

Jim removed the hoppers too. At least we were consistent.

To get a better idea of how many home baristas employ the "one shot, grind until empty" technique, I've added a poll "Do you dose and then run the grinder until empty for each espresso?"

This isn't a binary decision for me. I typically scoop enough beans into my minihopper for one session, and refill as necessary. Sometimes the hopper runs dry, but certainly not with every espresso. If I'm cupping, I'll scoop beans on a per-shot basis.

[Jacob]

FWIW: I only had my Robur a few weeks before selling my espresso machine - I'm in between machines at the moment!
My casual observation at the time were that the shots made without an extra load of beans in hopper had less consistent pours, but the shots that were made with a clean empty chute to start with were the ones that produced the most problematic pours. I blamed clumping, not only the amount of clumping, but also the 'quality' of the clumps. Note; I insisted on not using WDT or the like and if any clumps survived the doser they were ignored during basket-prep. I would love a second opinion on this!

Teme is not one of the grinder testers, since shipping the grinders to Copenhagen is somewhat awkward.

My guess is that Teme is way too busy at the moment (welcome home and congratulation to the both of you), but Teme lives only a few miles away and he is always welcome to use my Robur if he want to play around some more.

[gscace]

Hi:

I wanted to clarify something. My observation with the Kony for sure is that once popcorning starts, the required grinder adjustment changes. I can dial it back in if I want to just dump a shot's worth of coffee into the grinder, but my practice is not to do so. I keep a small amount of beans in the hopper - maybe 1-2 days worth because I prefer not to see beans flying around inside the hopper. The small force (weight of the beans) is enough to keep popcorning buried within the pile of beans and out of sight. I have absolutely no idea whether visible popcorning has any affect on taste other than it seems to require a change in grinder adjustment (for the Kony anyway. I never even bothered to check it with the Robur). I completely defer to Jim on this, since he now has the most credible experience comparing these things of anyone I know. So if he says it doesn't matter, I believe it.

My initial comment was that folks were receiving grinders and then removing all sorts of bits from them in what appeared to be a frenzy of modding these things in a way that folks were used to. I felt that the grinders should be tested as received. In my case, both my Kony and my Robur are completely box stock. I like the deflection plate and I've never removed the autofill switch. In the morning I flush out the grind path by grinding a second's worth of coffee and chucking that into the knock box.

-Greg

[cannonfodder]

After Greg's comment about all the little mods way back when, I restored both of my machines to their original condition (put the portafilter fork back on the Kony and the bean hoppers back on both). I normally left about a quarter pound (couple of days worth or less) of beans in the hoppers of both the Kony and Super Jolly for the remainder of the test.

That is more in tune with my normal habits. I normally keep a couple of day's worth of beans in the hopper of my Cimbali Jr/Mazzer Mini. I stated earlier that the flow changed when there was no weight above the beans in the burrs of the Kony, vs. having a couple of shots worth of beans above them. I do not recall much of a flavor shift, just a faster pour. If you adjust the machine to work with a single dose, I do not believe there would any difference in the end product, but I did not test that while I had the grinder.

My meager subjective taste tests pale in comparison to the work Jim has put into this project. It has sure morphed from the 'kick the tires' review it started out as. I am even looking at a TDS/PH meter. I never thought I would have one of those to test espresso.

[another_jim]

I'm testing the Max against the Robur today. The Max is unsuitable (at least without irreversible hacks) for single dosing, so I'm running it with beans on top. As an experiment, I'm alternating single dose and beans on top with the Robur. I thought the grind setting change might be a function of the grinds being fluffier and volume dosing differently with the slower single dose method, but it turns out that the beans on top does require a slightly looser grind for the same pour even when weighing the doses. The difference is about 3 to 4 seconds. So far, no difference in taste.

I'm scratching my head about the grind setting change. If popcorning creates more fines, one would expect a coarser setting to compensate. Single dose grinding is slower, since not as many beans enter the burrs at the same time. It could be that this means less fines are created. However, John particle sized a pair of samples ground using both methods, and didn't find any smoking guns in either direction.

This grinder stuff is becoming a lot like the espresso machine stuff -- there are real quality differences between grinders, but the explanations for them aren't holding up too well.

[RapidCoffee]

I stated earlier that the flow changed when there was no weight above the beans in the burrs of the Kony, vs. having a couple of shots worth of beans above them. I do not recall much of a flavor shift, just a faster pour. If you adjust the machine to work with a single dose, I do not believe there would any difference in the end product, but I did not test that while I had the grinder.

As an experiment, I'm alternating single dose and beans on top with the Robur. I thought the grind setting change might be a function of the grinds being fluffier and volume dosing differently with the slower single dose method, but it turns out that the beans on top does require a slightly looser grind for the same pour even when weighing the doses. The difference is about 3 to 4 seconds. So far, no difference in taste.

I'm scratching my head about the grind setting change. If popcorning creates more fines, one would expect a coarser setting to compensate. Single dose grinding is slower, since not as many beans enter the burrs at the same time. It could be that this means less fines are created.

More confirmation of the above:

I pulled four 16.5g dosed doubles this morning:


Using a minihopper and alternating pours, I ground enough unweighted beans for a single shot twice, and had at least 20g of beans on top twice. I did not adjust the grind between pours. All four shots turned out nicely:


Shot timings were measurably different. The weighted beans took 28-30 seconds to pour 50ml, and the unweighted beans took 20-22 seconds. Shot taste was good, but there's no doubt that shorter timings impact flavor.

This does not address the issue of grind adjustment. If you grind finer for unweighted shots, is the end result the same? With only one espresso machine, I'm not equipped to answer this question at home. But I do intend to run a particle sizing on weighted vs. unweighted beans, adjusted to the same pour times.

Ken and Greg deserve a vote of appreciation for bringing this important practical matter to our attention. Another interesting sideline of the TGP.

[another_jim]

Shot timings were measurably different. The weighted beans took 28-30 seconds to pour 50ml, and the unweighted beans took 20-22 seconds. Shot taste was good, but there's no doubt that shorter timings impact flavor.

I got less, about 5 seconds difference, but still enough to require a setting change when switching between the two methods. However, I put my single dose beans into a cylinder and weight it with a piston, so the early part of my grinding may approximate the weighted process.

[another_jim]

I'll be very curious to see what the particle analysis shows on this. My only idea to explanation the finer grind required when beans popcorn is fewer fines. I known fewer fines leads also leads to longer dwell times for an overall equal pour, so that is an alternative way to check it.

Perhaps the popcorning does create fractures in the beans -- and like creases in paper, you can tear along them and get smooth edges.

That's the good explanation for single dosing; the bad one is that the fines are closer to dust, like from a whirly blade grinder, and can't control the flow as well.

[Jepy]

Another factor is that there are grinders that simply cannot be used without a bean load above the burrs; two prime examples from my own experience are the Cimbali Junior/Cadet and Max models.

I have a Junior and DRM/Cimbali that are pretty much the same hopper design. A while back I made this piston to get rid of the hoppers. It's always worked very well, without the waste that used to get stuck in all the little places in these grinders.

[RapidCoffee]

I got less, about 5 seconds difference, but still enough to require a setting change when switching between the two methods. However, I put my single dose beans into a cylinder and weight it with a piston, so the early part of my grinding may approximate the weighted process.

I was also surprised at the magnitude of the difference, although Dave had mentioned "blond gushers" in past posts. I know I'll be more careful about letting my (mini)hopper run dry in the future...

I'll be very curious to see what the particle analysis shows on this. My only idea to explanation the finer grind required when beans popcorn is fewer fines. I known fewer fines leads also leads to longer dwell times for an overall equal pour, so that is an alternative way to check it.

My intuition tells me that popcorning is likely to generate more fines, together with a wider spread of particle sizes. More fines does not explain the faster flow rates, but perhaps the uneven larger particles do not pack as well, or are more permeable, allowing water to flow through the puck more rapidly.

On a related note: after examining preliminary SEM images, I postulated that conical grinders, due to the longer grinding path, may generate smoother, more regularly shaped particles, together with more fines. These particles might pack better, producing more even flows and more forgiving pours.

Unfortunately all of this is post hoc analysis. As you note, fewer fines from popcorning would neatly explain the faster pours. Interesting grounds for speculation, and a clear call for future study.

[hbuchtel]

A question for you guys using dosers and dosing/grinding per shot-

Do you clear out the grind path (between the burrs and the doser) and use those grinds as well? Or do you just use what falls into the doser?

It would seem that the last grinds to exit the burrs would be the ones most affected by the presence or lack of weight above them...

Henry (enjoying the ongoing discussion!)

[another_jim]

A question for you guys using dosers and dosing/grinding per shot-

Do you clear out the grind path (between the burrs and the doser) and use those grinds as well? Or do you just use what falls into the doser?

It would seem that the last grinds to exit the burrs would be the ones most affected by the presence or lack of weight above them...

Henry (enjoying the ongoing discussion!)

I always clean the chute, pulse the grinder, and clean the chute again. This leaves the grind chamber empty of anything except the compacted coffee. If you have beans in the hopper, you can still clean the chute if you want, but you can't pulse the grinder.

This may sound like extra work. But if you don't do it, you'll need to pulse the grinder, and sweep out the doser before you grind, to get the stale coffee out from the previous round. No matter how you handle the dosing, you need to handle the coffee left in the chute and grind chamber in one way or another.

[cannonfodder]

I always clean the chute, pulse the grinder, and clean the chute again. This leaves the grind chamber empty of anything except the compacted coffee. If you have beans in the hopper, you can still clean the chute if you want, but you can't pulse the grinder.

As do I. I clean the chute after the grind when doing a single dose. When the hopper is full, I clean it before my next shot and run a two second pulse to clear out the old stuff.

[cannonfodder]

I was also surprised at the magnitude of the difference, although Dave had mentioned "blond gushers" in past posts. I know I'll be more careful about letting my (mini)hopper run dry in the future...

I have not noticed the same thing on a flat burr grinder. I have a Cimbali Jr. and Mazzer Mini and do not recall having a similar issue although I may have and just never paid it any attention. The Kony was the first time I had noticed it. I will put a few doses of beans in the Jr, adjust it in and see if the timing changes as the beans disappear over the day.

[Psyd]

If you have beans in the hopper, you can still clean the chute if you want, but you can't pulse the grinder.

Ehm, you could close the gate in the hopper when you're done with the dose and let that run through, but that would be a huge waste of coffee. The Major allows enough coffee for two 18 to 20 gram doses under that gate.

I usually pack two, and sometimes three, doses in depending on how many folks I'm making coffee for at any given moment. If there were a disparity between the quality of the grind with a dose on top, or two doses on top, and none on top, wouldn't the pulls be significantly different? I'm noticing that one, two, and three tend to be remarkably the same in appearance and taste, regardless of the amount of beans on top of them when they're ground. OTOH, two doses bearing down on the first may not be enough to make a difference.

[another_jim]

As do I. I clean the chute after the grind when doing a single dose. When the hopper is full, I clean it before my next shot and run a two second pulse to clear out the old stuff.

That's the whole story in a nutshell -- "Single portion grinding with a doser - 101"

One reason I go hopperless is that using a hopper and sacrificial grinding entails around 4 to 5 grams waste per shot. Since I home roast, any time saving on grinding will be more than eaten up by the extra roasting time.

[RapidCoffee]

Do you clear out the grind path (between the burrs and the doser) and use those grinds as well? Or do you just use what falls into the doser?

Chute is always swept clean; grounds are used if needed for that session. My Super Jolly doserless mod makes chute access trivial:


zero grounds left behind in the chute

That could be why I saw such a big effect on flow rate...