Soliciting Insight As To Pros/Cons Of Variable Speed Grinding - Page 4

Grinders are one of the keys to exceptional espresso. Discuss them here.
Charlene (original poster)
Posts: 494
Joined: 7 years ago

#31: Post by Charlene (original poster) »

walt_in_hawaii wrote:I would agree. The heat conduction from the particulate to the room temp burrs would vary as what, the square of the radius of particle size? Since we are talking near powder sized, my money would be on conduction of heat into the particles from your room temp burrs as responsible for the measured temps shot by your IR gun, not as a result of friction generated by the grinding process.
Agreed. One can think of it like two bodies in space near enough to each other that gravitational pull between them becomes a factor. The smaller density object will close the distance between the two faster than the denser one.

User avatar
AssafL
Posts: 2588
Joined: 14 years ago

#32: Post by AssafL »

I can throw in triboelectric effect to spice up the bench race. For that enough study has been done (for the powder coating industry) to show that speed does matter. Slower flow makes for more charges per gram of powder.

That being said, one key observation made above is that normalization of the grinds may suffer at low/high rates. On the VL, higher speed means more defined donuts that warrants more care in distribution. For the Monolith Flat Cebseb mentioned above (Since I do not own one I take his word for it):
cebseb wrote:The one thing that did stay true was that a higher RPM on the Monolith Flat required distribution (shaking inside a cup in my case) in order to prevent channeling.
In our case one has to be careful not to confuse improper normalization due to speed with grind size/grind quality.
Scraping away (slowly) at the tyranny of biases and dogma.

Charlene (original poster)
Posts: 494
Joined: 7 years ago

#33: Post by Charlene (original poster) »

AssafL wrote:I can throw in triboelectric effect to spice up the bench race. For that enough study has been done (for the powder coating industry) to show that speed does matter. Slower flow makes for more charges per gram of powder.

That being said, one key observation made above is that normalization of the grinds may suffer at low/high rates. On the VL, higher speed means more defined donuts that warrants more care in distribution. For the Monolith Flat Cebseb mentioned above (Since I do not own one I take his word for it):

In our case one has to be careful not to confuse improper normalization due to speed with grind size/grind quality.
I haven't been reading complaints about static charge that RWT does not ameliorate. The grinder I am using is grounded, thus provides a discharge path for electrostatically charged grinds, thus effectively neutralizing cling.

I think of it as a factor but not a significant one.

The issue, to me, is: which rpm range, with a given variable speed grinder, delivers the best taste in the cup?

There may just be preferred ranges for different types of beans/roast levels.

walt_in_hawaii
Posts: 665
Joined: 9 years ago

#34: Post by walt_in_hawaii »

oooh, what a big messy door you just opened :)
that's probably true, but one cannot sit on the side wondering if its the best grind approach; we all have to do it regardless of whether or not there is an accurate protocol for maximizing flavor when presented with x bean roasted to y done-ness. I think I will shoot for around 425rpm as an all-around compromise for now and see where that gets me. I'm not all that convinced I will have access to a dc motor with which I will even have that kind of control in the future, though. If you do it, please post your findings.

aloha,
walt

User avatar
homeburrero
Team HB
Posts: 4863
Joined: 13 years ago

#35: Post by homeburrero »

Earlier in this thread I said I'd do a quick test on my new Monolith flat, look at flow rates and grind rates versus burr rpm. I did that a couple days back and am finally posting what I saw.

The main point was that although the EG1 (and, it appears, Frank's Titus hybrid) produce faster flowing shots as the grinder speed is increased, the Monolith is the opposite. Dominick (dominico) and others reported this a while back in this thread: Grind speed RPM changes--what effect are you noticing?

My tests corroborated that for the Monolith. I did a preliminary test and got a 37 gram shot in 30 secs at 250 rpm, a 12.7 gram shot at 400, and nearly choked at 550 rpm. So I coarsened my grind and did two tests at 300-600 rpm in 100 rpm increments. Got these results:

vvvvvvvv Monolith Flat vvvvvvvvvv

^^^^^^^ Monolith Flat ^^^^^^^^

My Monolith did produce slower shots as the RPM is increased. (A simple Pearson's r test for linear correlation gave me an r value of -.892, which for two tailed and 6 degrees of freedom is a p value of .003.) It appears that there may be a flattened curve, in this case in the 450 rpm neighborhood, where changes in burr speed have less effect on the flow rate. I would need to do more samples if I wanted to try to confirm that. (450 rpm is where I had settled in to using my grinder, but I'm still getting used to it.)

While I was at it I also weighed the grams out of the grinder in the first 100 revolutions along the lines of what Jim suggested. There appears to be a positive correlation between grind speed and the coffee ground per revolution in this range, but my test wasn't good enough to find any possible sweet spot. I think I'd need more coffee and go with longer grind times (maybe 500 revolutions) and more samples in 50 rpm increments to get a better idea whether there might be a sweet spot somewhere in there.

Note: The Monolith's fastest speed is 800 rpm, and Denis recommends less than 600 rpm for light roasts.

FWIW, here's the full details of my method:
Spoiler: show
Grinder: Kafatek Monolith flat., 75mm Titanium Nitride coated burrs (Spring 2018 build). Not yet fully seasoned (has had 3lb or so of grinding on new burrs.)
Beans: Allegro (medium roast - Costa Rica Dota) ~month old, room temp, lightly misted beans.
Espresso machine: Rocket Giotto Evo V2: HX, E61 group, Rotary vane pump set at 9 bar max, standard 0.6mm gicleur. VST 18g basket in bottomless portafilter.
Scale: Acaia Lunar (0.1g resolution)
Location and Date: Albuquerque, NM, 5000ft ASL, May 21 2018
Method: After very light RDT, as near as possible to 18.0 g of beans dropped into empty non-running grinder, set and locked at 7.25 on the grind adjust scale. Grinder was switched on for ~100 revolutions (timed manually, 20, 15, 12, and 10 seconds between pressing grinder button on and off.) Output was collected in tared portafilter+basket+funnel then weighed and recorded (g/100rev). Then the remainder was ground til empty, with a brief high speed purge and spout tap per usual monolith single-dose technique. Final output was weighed and recorded, WDT, leveled with blade, groomed with OCD style tool, tamped til fully compressed, then pulled on the espresso machine (flush-and-go, lever up 30 seconds) and the beverage output was weighed. Sample order was 300-1 ... 600-1, 300-2 ... 600-2. There were approximately 10 minutes between the grinding of each sample.
Pat
nínádiishʼnahgo gohwééh náshdlį́į́h
★ Helpful

User avatar
another_jim
Team HB
Posts: 13871
Joined: 19 years ago

#36: Post by another_jim »

Thanks for doing these tests. At last somee real data to think about and discuss

The faster grinds have more throughput per revolution and more puck resistance (i.e. less flow). If this is due to bean to bean grinding, this will be the way to tasty shots. On the other hand, if the faster grind produces more fines, the quality will be a wash, or maybe slightly worse.

I was under the impression that the feed rate (throughput per revolution as you call it) would peak somewhere. I guess I was wrong; although it may happen up at the usual flat burr speed of ca 1400 rpm. If you have a bunch of junk coffee, you could measure the throughput er revolution up there.
Jim Schulman

Post Reply