shadowfax wrote:[emphasis mine]
I agree with you for the most part, as do I think most of us. I also agree that a mechanism for quantifying reproducibility in addition to just the comparative taste is important. However, I've totally lost you at the bolded comment. It seems like you waste a minimum coffee when you're running pre-weighted, single dosing vs. keeping a hopper full, where you (maybe) have to purge after making grind adjustments, after leaving the grinder to sit idle, etc. Did you switch the two around, or am I missing something key here?

Here are the sources of waste using single dosing using what I would define as good technique:

(1) If you assume that there is going to be coffee left over in the grind path from whatever was ground before, then you need to purge that. Simply running the grinder until no more grinds come out then using a chopstick or other implement to clear the chute(s) will still leave at least "some" of the former coffee in the grinder and the only way to expunge them will be to use some of the next coffee to push them through. This is certainly how I would do single dosing if I followed this practice, if I wanted to be sure that what I was tasting on my next shot was close to 100% of whatever coffee I intended to drink. So, if it were me, I'd purge the grind path with 2-5 grams of the next coffee and discard it before I loaded the preweighed amount into the hopper. Not purging what is left in the grinder means you will be drinking a melange of what was in there before and what you want to drink, now. Can people taste the difference? I don't know, but it would be worth the several grams of the newer coffee, to me, to eliminate this risk. Alternatively, you could change coffees using the first grind of the new coffee in a milk drink, if you make a lot of milk drinks daily.

Theoretically you could do this before each shot even if you didn't change coffees often, however testing by Abe and others on coffee ground up to 12 hours before and used in espresso doesn't support the notion that there is marked deterioration during this sort of time period, so in the scenario of not changing the coffees in the grinder often you could just purge it once a day, in the morning, and be fairly confident that no further purging was necessary.

(2) When you change coffees, no matter how you dose, the grind usually changes. Perhaps some people are so good at adjusting their grind settings on the fly that they can adjust for this, but I would suggest that most people are not good enough with this technique to avoid extra waste. The extra waste would come from either producing a higher proportion of sink shots or from drinking substandard shots that would require further grind adjustments in order to nail the grind adjustment necessitated by changing coffees frequently.

I think these things could be tested along the lines of what I suggested in my initial post, although as I also noted I doubt that whatever would be observed would change anyone's behavior so doing the study would be optional.

As to my own personal opinion, I'd suggest that for those who want to change coffees very often that it might make more sense to buy TWO of a very good but cheaper grinder (say a Max) instead of one more expensive one (a Robur or a Compak, for example) and by having two grinders one could always have at least two coffees available for immediate use, or even three coffees available on any given day by cycling the grinders among coffees every other day rather than cycling one grinder 3 times per day.

Just my opinion, and I doubt that this or any other similar thread will change the behavior of even 3% of those who read it.

ken

Moderator note: topic split from Single Dose Versus Hopper Grinding: Hypothesis about the exact difference. See the original thread for context.

Ken Fox wrote:(1) If you assume that there is going to be coffee left over in the grind path from whatever was ground before, then you need to purge that.

In my experience (with my usage pattern), this is not a good assumption.

I load a weighed dose into my grinder to start, and then bump the motor and brush the chute until no more coffee comes out. My doser is *slightly* modified to sweep very clean, so all of the coffee gets swept into the basket. When I re-weigh the dose after grinding and dosing it is usually spot on the same weight I started with, sometimes there's a tenth of a gram variance (or very rarely up to two tenths).

And so I see no need to purge anything at the start of the day or start of a session, because there isn't enough variance for me to taste any difference. If I start the grinder in the morning before loading it, sometimes I'll get a tiny wisp of grounds, but generally not enough to register on my 0.1g scale.

As far as single-dose grind adjustments goes, it's hard to imagine that you would waste less coffee making adjustments with a full hopper but I don't know your regimen. In my experience, it takes a minimum of a shot or two for any adjustment to be fully realized, and perhaps more if the adjustment direction changes(due to thread backlash). If the burr area was full of semi-crushed coffee then all of that would presumably need to be purged before you could start to judge the effect of your adjustment.

With single-dose grinding, there is no semi-crushed coffee inside the burr area so there is nothing to be purged after making an adjustment. It still takes a shot (or two or three) to fully realize an adjustment

To me, this is an amazingly contentious issue, especially given that there is virtually no objective evidence or physical analysis. And before someone objects to that, here is what I mean...

1. Comparing what you taste today to your memory of what you tasted yesterday is not objective:
(A blind taste test would be an objective taste comparison)

2. Looking at grounds on a paper or in your hand does not constitute physical analysis

A "poor man's screening" would be a rough physical analysis, while laser diffractometry would be more accurate. It is possible to make an optical comparison but that requires making careful measurements under magnification ("eyeballing" does not constitute an objective measurement).

The debate about bean cellulose structure is not interesting to me (at all), but I'll observe that even this is heavily contended when it seems to me that it does not directly address Jim's hypothesis for what is physically happening. Jim's description is about the dynamic of crushing/shearing in a less crowded chamber as opposed to a packed burr area.

Ken Fox wrote:Particle size distributions, although interesting, will not get at the heart of the matter which is whether one compromises taste and/or reproducibility with single shot dosing. If single shot dosing does compromise taste or reproducibility, then particle size distribution might (or might not) explain this result.....
<snip>
Who knows? You'd have to test it to know for sure, and it might turn out that some people would prefer one sort of result over another. The results of such a blinded study would be very interesting, even if they don't have the probability of getting very many people to change their technique.

In this case, I agree with Mr. Fox (of course with the exception of the waste comments).

To me, the bottom line is that in absence of any real objective evidence all we have is conjecture, speculation, and personal impressions. And while these are inherently debatable, it's all theory.

JonR10 wrote:In my experience (with my usage pattern), this is not a good assumption.

I load a weighed dose into my grinder to start, and then bump the motor and brush the chute until no more coffee comes out. My doser is *slightly* modified to sweep very clean, so all of the coffee gets swept into the basket. When I re-weigh the dose after grinding and dosing it is usually spot on the same weight I started with, sometimes there's a tenth of a gram variance (or very rarely up to two tenths).

My alternative explanation is that the amount of "retained grinds" is going to be relatively constant within any given grinder assuming the same "purging technique." In order to test this one would need to (as completely as possible) purge the grinder of whatever was in it before, then introduce some "labeled beans" of another type that you could observe the first bits of arriving into the doser from the grind chamber chute.

The beans could be labeled in a number of different ways. Radiolabelling would probably be the most elegant, however this would be hard to pull off unless one worked in a research institution. A foreign substance, such as grinder cleaning powder or instant rice, could be introduced and the amount of ground coffee expelled before the first bits of white grinder cleaner or rice shows up in the doser could be weighed. Such a foreign substance risks giving erroneous results, however, because it might behave differently inside the grind chamber than does coffee, being seen either earlier or later than its real position in the grind path between the two types of coffee; the foregoing concern is probably not important. Another possible approach would be to soak beans in some sort of substance that would change their color (without, hopefully, their behavior on being ground). Or, you could intentionally grossly under roast a coffee so that it is so much lighter than the other one that the change between the coffees would be very obvious on exiting the grind chamber. This might be the best compromise for a real test. Regardless of what one used, in all instances the grinder could be micro-pulsed and the contents of the doser weighed progressively until the foreign substance is first seen.

My opinion is that the larger grinders favored by many here, such as the Robur or the Compak K10, would have relatively larger amounts of retained grinds than would smaller grinders with lesser amounts of cutting surfaces.

JonR10 wrote:
And so I see no need to purge anything at the start of the day or start of a session, because there isn't enough variance for me to taste any difference. . .. . .

With single-dose grinding, there is no semi-crushed coffee inside the burr area so there is nothing to be purged after making an adjustment. It still takes a shot (or two or three) to fully realize an adjustment

This is highly conjectural for the reasons I gave in my first response in this post. No one can possibly know the answer to this without further testing, however it seems more logical to me that the grind path behaves more like a pipeline in that something is needed to push the retained grinds through, that air alone is not sufficient for these (hypothetical) retained grinds to make it from the grind path into the doser or exit chute in the case of a grinder without a doser.

ken

Ken Fox wrote:...it seems more logical to me that the grind path behaves more like a pipeline in that something is needed to push the retained grinds through, that air alone is not sufficient for these (hypothetical) retained grinds to make it from the grind path into the doser or exit chute in the case of a grinder without a doser.

Interesting. Thanks for clarifying this Ken, I now better understand your point.

And you're probably right that the actual amount of retained grounds that get into the next shot is unknown (if any). I suspect that, if any, it's a small amount (but this is conjecture).

For reference, I am using a dosered Robur.

JonR10 wrote:And you're probably right that the actual amount of retained grounds that get into the next shot is unknown (if any). I suspect that, if any, it's a small amount (but this is conjecture).

My experience as well. I am single-dosing non-espresso brews on a Mazzer Major (and Baratza Virtuoso at work), and have seen little grinds retention. After grinding, I brush out the Major's exit chute and pulse the grinder a couple of times. Virtually nothing more comes out, even the next day. It's possible that new coffee dislodges old grinds, but I just can't get excited enough about this to start radiolabelling my coffee.

Large conicals have larger grinding chambers, and it seems likely that grinds retention is a bigger issue for these grinders.

I think we have the makings of a very simple and somewhat elegant experiment that could be easily performed by anyone who owns an accurate scale, a grinder, and access to differentially roasted beans. This would most easily be performed by a home roaster, using beans that roast to a consistent color (e.g. preferably wet processed, and especially NOT something like Harrar Horse in which every bean in a roast batch has its own color).

Roast one batch into early first crack then stop roasting, retaining the light color. Roast another batch well into 2nd crack, although before the very dark and oily phase. Let the beans rest for a few days, if for no other reason than to mimic the physical characteristics of the beans as we typically grind them.

For the actual testing, start out by running enough of one of the coffees through so that the grinds are uniformly like the coffee you are using. Then, purge the grinder as much as you can by grinding empty and using a chopstick or other method to clear out the chute(s), then run the other type of coffee through with micropulses until you see the first bits of the 2nd coffee coming through. Weigh the coffee that came out before you noticed a color change. Repeat the test in the opposite order. Repeat the test at least several times varying the coffee used until you have enough relatively consistent results that can then be averaged to give an approximate average "retained grinds" for your given grinder. If you have more than one type of grinder you could use these beans to test all the grinder types you have and then you can post your "results" in this thread giving whatever details might be of interest to interested readers.

This strikes me as a worthwhile test that should at least be able to answer the question about whether this is or is not a real issue to be concerned about, the potential "melanging" of coffees by people who frequently switch coffees used in a single grinder, whether or not they single dose.

Once this is done and we have an idea of how much grinds are retained by the more popular grinders, then we could design another simple taste test where one intentionally mixes varying percentages of two widely dissimilar coffees to try to determine what is the minimum amount of an obviously different coffee that would need to be mixed in before an experienced home user could taste it. This latter study would in essence be a "titration study," and although tedious to perform could be done by any number of people themselves if they just wanted to get an idea of how sensitive their own palates are to this sort of "unintentional blending." The more similar are the two coffees used in succession would of course impact how much "melanging" there could be before it was noticeable in tasting, however it would be easier to run the tests with dissimilar coffees and then to mentally extrapolate to the sorts of coffee that one actually uses to determine whether or not this is an issue of real importance to a given home barista.

ken

Ken Fox wrote:No one can possibly know the answer to this without further testing, however it seems more logical to me that the grind path behaves more like a pipeline in that something is needed to push the retained grinds through, that air alone is not sufficient for these (hypothetical) retained grinds to make it from the grind path into the doser or exit chute in the case of a grinder without a doser.

So, when you have a hopper that is full (enough), it seems painfully obvious that the grinder acts more or less as a pipeline. It's probably more or less a FIFO queue with a modest degree of local mixing in the burr chamber.

However, arguing that it continues to act as a pipeline when it's empty is maybe not as intuitive as you suggest. If a water pipeline were consistently slanted down, for example, when water flow stopped, provided some aeration at the top allowed water to succumb to gravity, that would pull the water and empty the pipeline. It makes obvious sense to me (and I assume the Jo[h]ns as well) that pipelines have methods other than more of whatever is in the pipeline for purgation. In the case of the grinder, this is gravity and a 400-1600 rpm burr and burr carrier with vanes which expel coffee efficiently. If you're "bumping" the grinder and sweeping the chute thoroughly, I don't see there being anything on the order of 5 grams being held up in the chamber or burrs. It's more like a gram or two, and it's ultrapacked stuff on the faces of the vanes and a few nooks and crannies, as well as the 'ring' along the bottom of the upper burr where there is clearance between it and the sweeping vanes.

An easier confirmation than purging through a different color of bean (or Grindz/the like) is to do your normal inter-shot cleanup and then pull the upper carrier. On the Robur and Super Jolly, they're extremely clean unless there's some mixing of the packed stuff, and still pretty clean if there is. That's my experience, and FWIW I don't single dose, and I'm opposed to it, particularly on my own grinder. It's just my opinion that waste is probably the least of your potential woes with single-dosing, at least if you pick the right grinder--as I said, my Nino is a very messy grinder for single dosing, and also has a point of buildup at the 'flap' that is used to prevent clumping, which is rather challenging to clean out with a brush (though it's doable). I've read many claim that the Cimbali Max [hybrid]/Junior is ill-suited to single dosing compared to other grinders as well, owing to its hopper design (not speaking from personal experience at all--never seen one firsthand).

Still, I'm curious to see the result of this test for the Robur/K10, and any other grinders that people are single dosing on. FWIW, this thread might be worth a review--you'll find there results from a similar test to what you describe of big conicals, only with full-hopper dosing, where the retention numbers are nothing short of painful. I never formally tested single dosing in that manner, though, and I don't know of anyone else who did.

Disclaimer: I know nothing and I am stupid! My equipment is amateurish.

I have a single grinder at the moment (the B-Vario was unfortunately disabled!). It has relatively small 50mm burrs.

I am switching between two very different tasting coffees these days, a vacuum pack Italian espresso blend (very sweet and chocolaty with a light smell of candy) and a now 11 days old Ethiopian Yergacheffe from 49th parallel (floral, floral, floral, rose water, and light sweetness). I try the Yergacheffe every day because it requires considerable aging and It has not hit its peak yet. But it is getting better everyday. I was told by the Barista that sold it to me that it would require some aging. The first day was like drinking from a bottle of rose water perfume, it was so intense!!!!

Anyway, the reason I am writing this is that I have been single dosing both these coffees for the last 8 days now. I am very thoroughly cleaning and brushing between them to sample the Yergacheffe every day. I always start with the Italian blend and then try a shot of the Yergacheffe and then back to the blend for one or two shots. I did not weight the beans going in. I weight the grind in the basket. I purge off the first few grams and the last few grams of each coffee.

It is VERY EASY to taste the yergacheffe in the subsequent Italian espresso blend shot and I can taste the espresso blend in the yergacheffe following it. This is true sometime even two double shot worth after the yergacheffe.

Ken Fox wrote:This strikes me as a worthwhile test that should at least be able to answer the question about whether this is or is not a real issue to be concerned about, the potential "melanging" of coffees by people who frequently switch coffees used in a single grinder, whether or not they single dose.

There is no doubts what so ever that there is taste melange between shots. This is not a theory it is a fact!

I am not a master taster nor anything. But this combination of coffee makes it so damn easy to taste the influence of one to the other shot that even I feel confident posting this information.

Disclaimer: I am not taking a position on the current debate nor do I have a theory on anything at this time. I just wanted to pass on this information.

Ken Fox wrote:I think we have the makings of a very simple and somewhat elegant experiment that could be easily performed by anyone who owns an accurate scale, a grinder, and access to differentially roasted beans.

Ken, it certainly seems straightforward and I am sure several people would be willing to run the experiment. I know I certainly would. However, I suspect the results will be very grinder-dependent. More importantly, what question are we trying to answer?

For example, if we start out by saying that in coffee roasting, individual component beans percentages generally do not fall below 5% because of the threshold of taste perception, then we might want to perhaps know how many grams of Blend #2 must be processed to achieve a 95% threshold of Blend#2/Overall Mixture. This would determine the effective "waste" and with enough data from different grinders, we might provide guidance for those wishing to purchase a grinder for single dosing versus hopper grind.

That is just one thought; I am actually more interested in why you made the proposal for the experiment.

On another note, I must raise concerns that the oiliness of the darker roast may have an impact on the flow-through and retention, so that 1/2 and 2/1 measurements may not be the same.