The rest of the paragraph in the Illy book has a bit more to say on the subject:

....... If the height-to-diameter ratio of about 0.2 is lowered, very fine grounds are required for coffee to percolate within a standard time; this choice would degrade the reproducibility of percolation, because of the creation of localized channels by water. A larger height-to-diameter ratio would give the cake the shape of a column; in this case, optimal extraction would demand an excessively high pressure (as in liquid-phase chromatographic columns). Conversely, for normal pressure values the grind would be so coarse as not to offer a sufficient number of fractured cells, resulting in a low extraction. Consequently, only minor variations from the optimal form of the established filter can be recommended.

Endo wrote:I do miss my old ridgeless basket (not available on the Vivaldi).

Ridgeless double-basket? I have both a ridged and ridgeless for the Vivaldi, so they should be available somehow.

TruthBrew wrote:Ridgeless double-basket? I have both a ridged and ridgeless for the Vivaldi, so they should be available somehow.

I have seen someone remove the ridge (with a bit of time and good metal working skills), but I have never seen one for sale. If you have a source, please let me know.

Where did you find a ridgeless 53mm double basket? Inquiring minds...

Frost wrote:The rest of the paragraph in the Illy book has a bit more to say on the subject:

Yes, I read that part. The paragraph wording addresses the basket shape, not the diameter, and does not suggest that any testing was actually performed. There's also a paragraph on grinds expansion (with a suggestion of 5-6mm headspace), and a paragraph on the holes in the bottom of the basket. But there is nothing that compares the performance of different diameter filter baskets.

Frost wrote:The rest of the paragraph in the Illy book has a bit more to say on the subject:

Man, that Illy book is about as interesting as reading a Chemistry textbook (but without the reference notes).

His "magic number" is 20% (puck height 20% of puck width). Anyone know what the ratio numbers are for a double (14g) 58mm and for 53mm? Also, what's the ratio for a more typical 17g overdosed shot of both diameters?

Also, according to Illy, should I be using a pressure higher than 9 bar for a 53mm? I'm experimenting with different pressures on my Vivaldi right now, so that part is very interesting to me.

Yes, the Illy 'specs' a large headspace and a thin puck. The relation to basket diameter would imply that the smaller is more ideal for smaller doses. No mention of what is considered a small or large deviation either. One must assume an empirical method was used to arrive at the specs.

I'll try and dig up the study.
It's taking more time than I thought (though the search has resulted in some good things - like my finding the Illycaffe research abstract that talks about the perspex portafilter results; and the Nestle abstract talking about how both water pH and coffee type impact dynamic compaction of coffee during extraction).

Frost wrote:Yes, the Illy 'specs' a large headspace and a thin puck. The relation to basket diameter would imply that the smaller is more ideal for smaller doses. No mention of what is considered a small or large deviation either. One must assume an empirical method was used to arrive at the specs.

Or a WAG. As much respect as I have for Petracco, he needs to support his observations with evidence, just like everyone else.

But this is a good point. According to Illy/Petracco, the 53mm Spaziale basket really should be shorter, not taller. This is not the case:

From left to right: 57mm Marzocco double, 58mm Quick Mill double, 53mm Spaziale double, 58mm Rancilio (?) triple

The 53mm Spaziale basket is the same height as a 58mm triple basket. The 58mm Quick Mill double approximates the height:width ratio of 0.2 more closely than the other baskets, using the ridge as a rough guide to height.

The Petracco excerpt states that a taller puck requires a higher pump pressure. It would be easy enough to run La Spaz, DC and San Marco groups at 10 to 11 bar, if one feels the shots at 9 bar are underextracted or ground too coarsely.

However, this is hard to understand in context. He says taller percolation columns require higher pressure (and this is borne out in instant coffee plants), but he also states that the flow is maximal at 9 bar and actually slows down at higher pressures (and this has ben confirmed by people experimenting). The two statements together contradict each other, or imply that grind styles have to change if the puck depth changes a lot. Finally, most shots in Italy are singles, and single baskets have truncated cone cross sections, not a cylindrical one, and this either contradicts his statement about cylinders being optimal, or means that the Italians pull bad single shots.

The Illy book is wonderful; but it also has a lot of holes, non-sequiturs and contradictions, like this one, probably due to trade secrets.

Personally, I prefer the 58mm goups to the smaller ones I've tried; but I don't think the information in Illy invalidates the DC and LS machines. Also, my preference is based on very limited experience with small groups and a lot of experience with full sized ones; so even I consider it quite suspect.

another_jim wrote:The Petracco excerpt states that a taller puck requires a higher pump pressure.
...
However, this is hard to understand in context. He says taller percolation columns require higher pressure (and this is borne out in instant coffee plants), but he also states that the flow is maximal at 9 bar and actually slows down at higher pressures (and this has ben confirmed by people experimenting). The two statements together contradict each other, or imply that grind styles have to change if the puck depth changes a lot.

Jim,
I don't have the text, but might it just mean that for any specific puck height, a graph of flow against pressure is humped? That suggests that 9bar is optimal for the most common puck heights and a higher pressure will deliver highest flow for a taller puck.

Greg

Yeah! That does makes sense of all the seemingly conflicting statements. But I'm not sure if their data goes that far, at least they don't say so.

I spoke extensively with Dalla Corte last year at the SCAA conference. I questioned the 54mm basket and they claimed to have done research not only on proper diameter/height ratio but also what diameter was best at that ratio. Of course that still begs the same question as illy's specs do - where is the research? Or better yet, what qualifies as research to them?

ameza wrote:I spoke extensively with Dalla Corte last year at the SCAA conference. I questioned the 54mm basket and they claimed to have done research not only on proper diameter/height ratio but also what diameter was best at that ratio. Of course that still begs the same question as illy's specs do - where is the research?

Yes, where is the "research?" And of course, research or none, we can hardly expect them to say 54mm is inferior to 58mm.

To me, the discussion over "proper" basket diameter makes some sense when applied to Italian espresso, but no sense elsewhere. Let me explain:

In Italy espresso dose appears to cluster (by tradition and practice) at around 13-14 g. Thus, when one specifies a basket diameter, one also specifies a puck depth. Since the puck depth has a major effect on extraction characteristics, Italian diameter recommendations correlate fairly well with what the baristas want the espresso to taste like.

Outside of Italy, especially in areas where baristas are trying to reinvent the espresso beverage, dose varies widely. Typical dose recommendations proposed by "third wave" roasters span a range of at least 13-22g. This means that for a fixed basket diameter, puck depth will vary tremendously (and so will the extraction characteristics).

For instance, say you liked the way a particular coffee tasted as a 14g dose in a 54mm basket. If you wanted to switch to a 58mm basket, you'd have to up the dose to 16g in order to have the same puck depth. And you'd would need a 68mm basket(!) in order to use a 22g dose with the same puck depth.

Summarizing, I would say that in Italy, the debate over puck diameter is really a debate over taste preferences based on puck depth. Outside of Italy, dose varies so much that arguing over an "optimum" basket diameter makes no sense.

I'm not sure:

I think the discussion by Petracco is based more on flow theory than experience. He basically recommends the shallowest puck possible consistent with no channeling. He also states that a deeper puck requires higher pressure -- presumably to maintain the same pressure gradient (e.g. pressure drop per millimeter travelled inside the puck). The second point is contradicted, or at least obviated if most of the pressure drop is at the bottom of the puck where the fines aggregate.

In Italy, most shots are singles, by a large margin, and the group diameter is pretty much irrelevant (i.e. the effective diameter of a single basket is smaller than the smallest commercial group)

Finally, a cafe will want its singles and doubles, thwacked out from the same grinder, to roughly taste and flow the same. If there is no reason for singles to taste different, I have a hard time seeing how doubles will.

Perhaps the commendatori of Italian espresso machine companies are just as stubborn and opinionated about espresso as we are, and group diameter has more to do with that.

Not to over-simplify things, but is it safe to say a smaller 53mm is best for people who prefer the taste somewhere between a double basket and a triple basket (but using the same dose)?

Another thing I seem to be discovering is that the 53mm basket is more sensitive to variations in grind. Does this make any sense or am I imagining things?

another_jim wrote:I think the discussion by Petracco is based more on flow theory than experience.

I think you and I agree that the more one scrutinizes the Petracco article, the more omissions and contradictions become evident.

another_jim wrote:In Italy, most shots are singles, by a large margin, and the group diameter is pretty much irrelevant (i.e. the effective diameter of a single basket is smaller than the smallest commercial group)

It was probably a mistake for me to bring "Italy" into the discussion.

So here's a rephrase: The discussion of "optimum" diameter for a double basket is futile as long as people are pulling shots using anywhere from 13 to 22 grams.

I think we left the framework of the initial question long ago.

"Does basket diameter matter?"

Yes. Basket diameter absolutely without a doubt changes the characteristics of the espresso drink.

Just like any other variable in espresso extraction, the optimum diameter is dependent on ALL the other variables. The multi-variate answer is, however, beyond me.

Unless we make some of those variables static there isn't going to be a simple answer to what is optimum. I think it is fair to say that a roaster is generally testing on a particular type of equipment. So it would seem to me that our coffees are how we want them on the equipment we test. Like-wise when we develop a custom blend for a client, we are optimizing it for their equipment and their objectives.

I agree the tastes on a 53mm and 58mm are different. I am just not sure if one is "better". It reminds me of the "conical" versus "planar" burr discussions. I don't think it will ever be settled. It will just have to be left to personal preference.

AndyS wrote:So here's a rephrase: The discussion of "optimum" diameter for a double basket is futile as long as people are pulling shots using anywhere from 13 to 22 grams.

I'm going to have to redo the damn dosing study, using several grinders, and distinguishing between head space, puck depth and dose. Not looking forward to it.