Protocol for measuring percentage of coffee fines?

Grinders are one of the keys to exceptional espresso. Discuss them here.

#1: Post by boren »

I'd like to compare the level of fines between my grinders and am wondering how to normalize grind level between them. Would the following protocol be accurate/useful enough?

1. Dial in a shot of espresso so that each grinder produces the same shot time with the same amount of coffee (e.g. 30 sec, 10 gram in, 20 gram out).
2. Grind the same amount of coffee at this grind level and transfer into a 250 micron sieve. Shake for a minute and measure the weight.
3. Attach a battery-powered blower and a custom cap to the sieve (see comment below) in order to apply air pressure to blow out more fines. Measure the weight.

I tried this with a friend's Kingrinder K4, grind level set to 46 clicks. Here are the results:

Slowly cranking (about 80 RPM)
  • Initial ground coffee weight: 10.13 gram (includes RDT)
  • After 1 minute of shaking: 9.20 gram
  • After blower: 8.33 gram
Using an electric screwdriver (a few hundred RPMs)
  • Initial ground coffee weight: 10.08 gram (includes RDT)
  • After 1 minute of shaking: 7.86 gram
  • After blower: 6.51 gram
Do these results make sense? Would they be comparable to another grinder adjusted to produce the aforementioned 30 sec shot?

boren (original poster)

#2: Post by boren (original poster) »

To those wondering, below is the contraption I DIYed for step #2. I basically drilled a hole into a Nikon F-mount back lens cap to fit the nozzle of a blower and inserted the cap into the sieve.

boren (original poster)

#3: Post by boren (original poster) »

Any opinions?

Supporter ♡

#4: Post by JonathanMichael »

Shalom Oren,

Sounds like an interesting approach. However, in my experience, these sorts of homemade attempts to understand particle distribution don't really ever amount to much in the end.

I'm sure you have done taste comparisons between your grinders?



#5: Post by rmongiovi »

I'm relatively convinced that fines alone aren't really the issue although they certainly may be a symptom. You may think of fines as an issue since they extract extremely quickly but In reality there just isn't that much coffee there (assuming the amount of fines is within reason, of course).

I think the real issue is the shape of the main particle size distribution "hump". You can remove fines with a sieve but that doesn't change the shape of the rest of the particle distribution. It just truncates it on the low end. If you want your extraction to be controllable then you need as high an narrow a main particle size hump as possible. That is, the output of the grinder should be unimodal and as close to all one size as possible. Then when you adjust the grinder you can actually achieve the extraction you're looking for without too much being over or under extracted.

You're always going to have fines because coffee beans break the way they want to. It makes sense to me to concentrate on the main mass of coffee instead. Improve the quality of that part of the grind and you'll get much more benefit for your labor.

boren (original poster)

#6: Post by boren (original poster) »

@Jonathan - taste comparisons are a challenge, because cupping isn't very similar to the coffee I'm used to drink. In the few attempts I had with triangle tasting using this method I could detect the odd cup from clearly different beans, but not (gasp) when comparing the same beans that I roasted to the same level but using different methodology (specifically, no preheat + internal cooling vs preheated + external cooling). If I can't tell the difference between beans that are supposed to be easily distinguishable, then I'm afraid detecting nuances between different but similar grinders may be even more challenging. This is one of the reasons why I'm curious about particle distributions and level of fines.

@rmongiovi - I experimented with measuring particles distribution beyond fines using macro images and Jonathan Gagné's software, but the results where not very consistent. I think sifting out sub-250 micron particles may help get more useful ("cleaner") information about the 250+ micron range. I'll definitely try that.


#7: Post by rmongiovi »

The reason I entered that post is that whenever I ask about a burr set that produces fewer fines I'm told, "Use a sieve to filter out the fines and then taste the coffee. You'll probably find it flat and unappealing." This is the "fines are really good" attitude.

The thing is, this is a straw man argument. The thing I want is a grinder that produces, to the greatest extent possible, coffee grounds that are all the same size. If all the grounds are the same size then I can adjust the grinder to produce the extraction I want while minimizing how much I have to compromise in order to deal with grounds that are both above and below that desired size. Rather than trying to average out over and under-extracted grounds I can concentrate on correctly extracting the unimodal grounds.

Removing fines with a sieve does nothing to fix the remaining coffee grounds. If they're all different sizes you'll still have to compromise when you adjust your grinder to average out the over and under-extraction. You've removed the fines but not the difficulty.

Fines seem to be an unavoidable side effect of grinding things that are brittle. You get dust. But since the total mass of coffee in that dust is insignificant it seems to me that it's the rest of the coffee grounds that matter most.

The counter argument to what I'm asking is that potentially having a range of size in the grounds being produced helps to improve the taste of imperfect coffee. I can't judge the truth of that argument since I've never really had a grinder/burr set that seems to produce what I would call unimodal output.