Recently there's been some, um, spirited debate on the (de)merits of per-shot dosing your grinder, but little discussion of technique. Astute readers already know that I am not a fan of per-shot dosing, but I recognize its attractions. How convenient to premeasure your dose, dump the beans in the hopper, and grind into the filter basket. This avoids waste, stale coffee grounds, and you can change coffees at any time.

Unfortunately, there is a dark side to dosing the grinder on a per shot basis. The grind setting must be adjusted, not just a little bit, but a huge amount. JonR reports a grind setting that is 9-10 ticks finer on his Robur. I only need 7-8 ticks finer on mine, but that's still more than the entire espresso range with a bean load, regardless of type of coffee, roast degree, or age. When the grind setting changes so profoundly, it's a foregone conclusion that the particle size distribution is significantly altered. I'm convinced this is tasteable in the cup, although it's certainly possible that blind taste tests* will prove me wrong.

The finer grind setting is not the only thing that will affect the particle size distribution. As the bean load goes from 15-20g (for a double shot) down to 0g during the grinding process, the first few grams will be ground more finely and the last few grams more coarsely. This implies that the range of particle sizes will be greater for single shot dosing. Assuming that grinder manufacturers know a thing or two about grinder design, and commercial grinders were never designed for a per-shot dosing usage pattern, this is probably not a Good Thing.

One technique that may help: start the grinder running, and gradually trickle the beans into the hopper, at about the same rate they are ground. At least the bean load will remain constant throughout the grinding process. You will need to set the grind adjustment even finer (just a couple of ticks on my Robur), and of course be prepared for increased popcorning while grinding the first part of the shot.

Again, I'm not advocating single shot dosing and grinding. But this approach seems more logical than dumping in one shot worth of beans and hitting the grind switch.

* which I am, fortunately, unable to perform in my kitchen

In the absence of taste evidence, or even the far more elementary evidence from a laser sizer that the grinds are in fact different, or even the more elementary data on bean feed rate versus the amount of beans in the grinder, the most convenient technique is best.

Let me give you another example of reasoning at the intelligence level of these grinder feeding wars:

• Everybody knows that the more uniform the grind size, the better the extraction. Now it is geometrically clear that the more varied the grind size, the coarser the grind required for the same flow, since particles with varied sizes form a tighter aggregate than particles of the same size Contrariwise, the more uniform the particle size, the finer the grind must be to maintain the flow. Therefore, whatever method produces the finest possible grind setting will produce the best shot.
• This means the best grind method is to feed one bean at a time and let it grind completely. Everything else just creates suboptimally distributed espresso shots

Jim, I was kinda hoping that a new thread would avoid more "grinder feeding wars", and perhaps even elicit novel approaches to per-shot dosing. Silly me.

Regardless of how much sarcasm you want to throw at it, this is a topic of some significance. Many of us subscribe to the grinder-first mantra, and have adopted commercial grinders for home use. But we are using them in ways that they were never designed for. I began to think more seriously about the implications of this during the TGP, especially after Greg Scace's comments about hoppers:

The level of modding at this point seems a bit over the top... I can tell you that both the Kony and the Robur like having some beans in the hopper... Are we testing grinders, or are we testing modded grinders, or are we just modding grinders?

I agree that this is a topic of some significance.

I at first posted a lengthy reply here about single dosing in general and think it's not appropriate here, so have moved it to the original thread.

But Jim, there is taste evidence - and evidence to the naked eye. You simply choose to ignore it

Go on! Do a quick kitchen test!

To take up John's specific question and suggestion; from a solely theoretical standpoint it assumes that each bean will hit the same point of the burrs in the same way and popcorn in the same way, so that consistency is improved. I doubt this is going to happen, though I imagine the method might reduce interference from beans ricocheting off one another as they popcorn around... if you really let the beans trickle in practically one by one.

This of course assumes that consistency is what it is all about. Jim suggests that consistency is not the whole game and from what was found when placing the grinds from conical burr grinders under the microscope I believe he may have a point there.... but within reason. Maybe there is a particle size consistency window- go beyond it and things get worse....

The single-dosing technique that I adopted was to close off the chamber as close as possible to the burrs once I had placed the beans in there, so as to limit the space for popcorning as much as possible. Beforehand I was doing what Jon does. I honestly don't think changing that practice made any difference. I just did it because at first I freaked out about how big the differences between single-dosing and hoppered use were and couldn't believe what I was finding, so tried to reduce the popcorning as much as possible.

Cheers

Mike

Interesting topic my friend, but the timing may be too soon to elicit a civil discussion.

Maybe your 7-8 ticks is more accurate....I'm guessing you looked while I just went from memory. In any case, it's a pretty big adjustment when we're talking espresso adjustments that are generally on the order of microns in burr separation. I find that just a couple of ticks is usually far enough to cover the largest spreads between different beans/blends, etc.

And while I also think the particle size spread changes for such a "large" adjustment, it is yet to be determined what that difference looks like overall and IF (or HOW) it truly affects the taste in the cup.


All that said, you may be onto something by addressing the changing column weight. I'm also guessing that it's different for flat vs. conical because of the volume of beans in the grind path. Some folks have reported that a Robur can hold as much as 30g in the grind path and chute, so "trickling" a single shot may not actually affect the grind at all whereas a 65mm flat burr like a Super Jolly may show a difference (I honestly have no clue if it would be true).


My thinking is that if we believe the particle size distribution is changing as the bean column is consumed then it's possible that for single dosing that change occurs in much the same way every time. This might explain how my shots can be so consistent and I can make small grind adjustment changes with very predictable results. Another possible explanation is that the distribution does not change much, if at all, through the course of grinding a single shot because there is never enough column to have a load on the big conical burrs.

To answer the question that was asked:
My general practice is to load the dose onto the closed gate, start the grinder, and then open the gate (swiftly) and close it again to trap the beans in the burr area so they can't fly out.

JonR10 wrote:Interesting topic my friend, but the timing may be too soon to elicit a civil discussion.

I hope you're wrong. Hey, we're all on the same side here, right?

Of course this is still largely speculation - including Jim's claim that "the most convenient technique is best". (No, it's simply the most convenient.)

As you point out, we have one piece of hard evidence: gross grind adjustments are required as the bean load is reduced. This does not tell us how the particle size distribution changes, but it would be an unbelieveable coincidence if it did not change at all. Different grinders generate measurably different particle size distributions, even when adjusted to produce similar pours. If you make gross changes to the burr spacing on a grinder while simultaneously reducing the bean load, why would you expect the grind distributions to be identical?

Some good observations in the above posts:
* The bean load change in per-shot dosing may be insignificant. Covering the beans in the grinder throat to prevent popcorning may be the best approach. This may depend upon the grinder.
* Consistency is probably not the issue. You can produce consistent enough pours with per-shot dosing or with a bean load in the hopper. The big question is taste.

Some interesting questions that remain to be answered:
* Is it possible to produce a similar particle size distribution with per-shot dosing?
* If per-shot dosing generates a different range of particle sizes, how does it impact the pour?
* Can per-shot dosing be manipulated to produce a similar (or perhaps even better) taste profile than maintaining a bean load?
* If not, how much bean load is enough?
* Do different grinders respond in different ways to per-shot dosing? Flat vs. conical burr, throat and burr chamber geometry, augers to pull in the beans... all of these factors may come into play.

Per-shot dosing is of more than purely theoretical interest for me. I've moved away from it for espresso, but I use per-shot dosing for all of my non-espresso brewing. Why? Coffee consumption habits. I rarely brew anything other than espresso at home, and I prefer to reduce coffee staling by keeping my vac pot/press pot beans in airtight containers or the freezer rather than the hopper. So we all make compromises... well, except maybe Chris Tacy.

I wish Mazzer would lighten up on the safety issue and sell me an augur that I could adapt to my Major. My strategy is to keep about 1/4 lb of beans in the "short Mini" hopper and keep replenishing the level every other day from a nearby airtight jar. I run through about a pound of beans per week and although the characteristics change slightly during the week it's fairly consistent. I find that if I let the beans get too low there are significant grind adjustments required. By keeping the levels fairly constant I minimize grind adjustments to the point where it's just an indent or two or three in either direction.

That said... it takes greater adjustments when I change bean types. For those who regularly change bean types in and out the hopper arrangement is sub-optimal. I've seen some arrangements with flexible plastic tubes that show promise but apart from the augur I have yet to see any mod's that improve the hopper arrangement.

I apologize for my sarcasm. But my experience does not confirm any of these assertions.

Moreover, since I regularly (i.e on a daily basis) switch doses ranging from 12.5 to 17 grams, grind for ristrettos to lungos, and go with hopper or without hoppers, all in response to using a wide variety of coffees on a daily basis; I do not need to be told what the best technique is for per-shot grinding.

I suspect that these dramatic observations on grinder changes come from people who make only the minutest grinder adjustments, no dose adjustments at all, and drink the same coffee all week. So I'm beginning to think that the people I'm arguing with live in a sensory universe that takes place at a completely different scale of magnitude.

Quite frankly, this all feels like I'm being told how to adjust the stereo's volume by the denizens of the House of Usher.

In any case, if you want to know my technique for per shot grinding, here it is:

• I have three to six coffees in containers.
• Each has a slip with the grind adjustment and dose.
• There is a small 0.1 gram scale on the doser fork.
• The basket is on the scale.
• I set the grinder.
• I feed the roughly right amount of coffee plus a bit extra into the throat of the grinder, add a tamper, and grind.
• Then I clear the chute and pulse the grinder three more times (sadly, the big conicals require this -- if I'm in a hurry, I just sacrifice grind 3 to 4 grams).
• I turn on the scale, and dose until the weight is right.
• The remaining grinds are flushed onto the doser lid and discarded.
• I then nutate to distribute the coffee, rotate the basket under the tamper to make the puck surface parallel to the bottom, and tamp lightly.
• The machine is flushed while I do this.
• Then comes the shot

Please let me know what mistakes I'm making

That's pretty much how I did it, Jim.

I reckon it's pretty close to what most people single-dosing do.

Sometimes I found it more fiddly to have the scales on the doser fork, so they were next to the grinder and I'd move the the basket there and back till the dose was right. But that's about it.

But I'm not sure that's what John was looking at - in good humour and with customary precision. As I understood it he was perhaps enquiring about whether it could make a difference whether one switches the grinder on before or after dosing the beans....

I'm not personally sure whether it would make a difference, but I'm happy to follow the thinking....

And don't recall anyone trying to tell you what to do....

RapidCoffee wrote:* Is it possible to produce a similar particle size distribution with per-shot dosing?
* If per-shot dosing generates a different range of particle sizes, how does it impact the pour?
* Can per-shot dosing be manipulated to produce a similar (or perhaps even better) taste profile than maintaining a bean load?

Good points John. I am particularly interested in the taste differences (from similar pours) using per-shot vs. bean load in the grinder. It would interest me also to correlate the grind size distribution with taste from the same grinder with per-shot vs. loaded hopper grinding.


What I'm getting at is that maybe the change in particle size distribution still leaves the flow and extraction in a relatively close state of balance for both cases. In other words, what if the taste is very similar even though the grind size distribution shifted?


It seems possible that IF the fines regulate flow and the main size peak determines taste then it may be possible to get similar taste balance with multiple distributions.