Is having a wide particle size distribution associated with severe acidity/sourness in the espresso shot? I have a big sourness problem and I want to rule out to possibility that my grinder is to blame. I can easily get the grinds fine enough to choke the machine, or get a 25sec shot, or 40, or whatever, so it's not an issue of "just grind finer". The question is, assuming the average grind size is fine enough to get a 40sec shot, would a large size distribution make a sour shot, or just a bad shot?

I would think that a wide size distribution would skew the taste towards bitterness, since while the larger particles may be underextracted and produce sourness (right?), the relatively huge surface area of the smaller particles means a lot more of the coffee by weight is overextracted, producing bitterness (right?). The final taste profile would be maybe a bit more acidic than it should be but a lot more bitter than it should be, resulting in a "bad" shot not a "sour" shot. If correct, my sourness problem is not because of my grinder and I should move on to other troubleshooting.

Why might I have a wide particle size distribution? I haven't analyzed the grinds, but I have a Compak K6 grinder that has been modified for single-dosing by attaching a 3D-printed auger to the top burr to force beans downward and prevent popcorning. The auger sometimes audibly rubs against the grinder throat which may tilt the top burr out of alignment. Also, when I very lightly press down on a side of the top of the grinder while it's running I can make the burrs chirp, which makes me think the burr carrier is really easy to tilt out of alignment and may not be perfectly level even when I'm not touching it. Severe misalignment would cause a wider particle size distribution.

Thanks for your help!

It's complicated.

Acids and sugars extract fast, so regardless of grind size or extraction, you get all the acids and sugars in the coffee. It's what you don't get that's critical. If the grind is too coarse, the heavier compounds fail to extract. These compounds add sweetness in the form of caramels, mouthfeel in the form of oils and long chain compounds, umami flavors from the unreduced Maillard compounds of the earlier roast (these are more apparent in lighter roasts) and complex bitter tastes like cocoa, spice, and smoke (these are more apparent in darker roasts).

Now, you can always get a balanced shot by grinding finer and dosing lower, regardless of how wide the particle distribution is. So the first thing you need to do is grind finer, and let the shots run longer. That will take care of the unbalanced acidity, guaranteed (if you are using a very acidic light roast like a Kenya, it may just be the coffee, and the super fine grind will just taste bland rather than balanced).

The allure of more unimodal grinders is clarity. Very high extractions will always balance a shot, but at the cost of blurriness in the flavors and aromatics. If you have a wide dispersion grinder, fully extracting the larger particles requires adding taste blur from the super extracted finer particles (the idea that over extracted is super bitter is a myth -- that's over-cooked or over-roasted coffee). If you use a unimodal grinder, the sweet spot will be clearer tasting; but finding it will be a lot harder.

I use a fuzzy conical grinder, because I am constantly switching between coffees and want a big sweet spot I can hit fast. If I were to use one coffee at a time; I'd use a more unimodal flat, and dial it in to the precise sweet spot. That choice depends on your way of doing coffee.

(Caveat: blind testing the added clarity of big flat burrs is very hard, precisely because they are so hard to dial in perfectly. Making comparable, blind, side by side shots is almost impossible. The account I've given is the current expert consensus, and accords with my own episodic, but not blind tested experience)

Here's the mental model I've been playing with. It largely agrees with Jim (however I didn't know about the myth of over-extraction and bitterness, but I'll keep it in for now). Caution - math ahead.

Distribution typically follows a bell curve, so let's look at a wider and narrower grind distribution. Let's assume here that for a 30s shot, the blue box represents particle sizes that will be optimally extracted. This leaves yellow large particles that will be under extracted and red particles that will be over extracted. The green line represents a boundary where if particles are smaller than it, they will cause increased puck resistance and pressure. So too much and too little is detrimental to pulling an espresso shot.

An optimal shot would fall squarely in the blue, with the average in the middle. The grinder with a wider distribution would have more falling outside the optimal band (here at 8% each under and over extracted) than the grinder with a narrower distribution (here at <0.5% each). We'd probably describe the wider distribution shot as rounder and the narrower distribution shot as having more clarity. Also note that puck resistance is at 50% for both, so we are able to hit 9 bar for 30s.


But what if we grind a little too coarsely? Here we can see that the grinder with a wider distribution has more large particles that will be under extracted (26% vs 10%) and would be more sour. It's also interesting to see that the puck resistance for the grinder with narrower distribution falls much faster and would show as a lower possible pressure to reach a 30s shot (this agrees with Jim's comment that grinders with narrow distributions are harder to dial in).


If we go the other way and grind too finely, we can see that the grinder with the wider distribution has more small particles that will be over extracted (26% vs 10%, and interestingly also more particles that will be under extracted). Puck resistance grows much more quickly for the grinder with tighter distribution (again being more sensitive and harder to dial in).


This model isn't perfect. For example particles on the far left of the green line will contribute more pressure than particles close to the left of the green line, but I think we're directionally accurate here. Also Jim's comment on bitterness is interesting and would need some thought. And how do bimodal distributions compare to unimodal? I don't quite know how to get Excel to draw that

That's a well illustrated, good explanation.

My comment on overextraction is based on extraction being asymptotic, with gross extraction levelling out at around 25%. No matter how fine you grind, or how long you extract, you'll level out at the same amount. So the problem isn't overextraction, but the degradation of some of the more delicate flavors, i.e. taste noise, by over-brewing

There is a caveat -- instant coffee is created from a brew that is around 75% extracted This is done by percolating the coffee at 20 bar and around 140C for around 8 hours. The result is that the cellulose is hydrolized and breaks down into soluble compounds that give instant its characteristically bitter taste. You get some of the same embittering effect by having office coffee sit on a hotplate for five hours, or by using boiling water when brewing.

The bimodal distribution is in fines, particles that are about 1/100th to 1/10th of the diameter of the rest of the ground coffee and which never amount to more than 3% of the weight. People without this quantitative knowledge think getting rid of fines is going to do wonderful things. They are mostly wrong. In espresso, no fines means no puck resistance and a mess. In regular brewing, the amount of fines is too small to make a taste difference. The exception is that for many styles of pour over, they make the coffee bed too resistive. Large flat burrs set fairly coarse, and long, slow pours, may work better for these styles than stirring a more resistive coffee bed with a higher percentage of fine particles.

People would be better off distinguishing sharply between the larger effect of the width of the coarse grind distribution from the smaller effect of how bimodal the distribution is.

I don't know scientific reasons why, but I did buy a couple of 1zpresso models of handgrinders. I found the JX model really highlighted the acidity of coffee, while their JE model was more creamy and balanced.

another_jim wrote:Now, you can always get a balanced shot by grinding finer and dosing lower, regardless of how wide the particle distribution is. So the first thing you need to do is grind finer, and let the shots run longer.

(my emphasis added)

So to boil this down to practical application, what parameter(s) would OP vary, and in which order, to achieve the more balanced shot? E.g., first grind finer at same dose to go longer (longer shot time) for the same gram yield? Or dose down and grind finer to produce the same gram yield and shot time?

Tyme wrote:Is having a wide particle size distribution associated with severe acidity/sourness in the espresso shot? I have a big sourness problem and I want to rule out to possibility that my grinder is to blame.

Obvious question: what coffee are you using? Before you blame the grinder, make sure that sourness is consistent across a variety of beans. Light roasts are in vogue these days, but your entry-level espresso machine (Gaggia Classic) lacks the ability to control temperature, pressure, preinfusion... in short, everything needed to tame the acidity of light roasts.

If you have already experimented with several different traditional espresso blends and consistently get sour results, ignore the above.

iploya wrote:(my emphasis added)

So to boil this down to practical application, what parameter(s) would OP vary, and in which order, to achieve the more balanced shot? E.g., first grind finer at same dose to go longer (longer shot time) for the same gram yield? Or dose down and grind finer to produce the same gram yield and shot time?

To get a certain rate of flow, the puck has to have a certain resistance. If you grind finer, the unit resistance of the coffee goes up, so the puck needs to be thinner to compensate, and you have to dose less. As you become experienced in shot making, this kind of calculation will become second nature. (Please, instantly forget the one variable at a time idiocy -- correlated variables need to be changed together)

BuzzedLightyear wrote:I don't know scientific reasons why, but I did buy a couple of 1zpresso models of handgrinders. I found the JX model really highlighted the acidity of coffee, while their JE model was more creamy and balanced.

Have you tried finer grind shots with the JX burr?

I have, it causes over extraction