Anyone with a Comandante and Kruve?

Zephyp · 6 Jan

Does anyone have a Comandante manual grinder and Kruve sifters? Preferably the C40 grinder, but if the burrs are identical, other models are also welcome.

Any sifting numbers for settings between 20-30 are most useful, but lower and higher also help.

I ask because I get quite different sifting results with my 10 year old C40 than MWJB did with an X25 he had earlier. Where he got 8% through the 400 micron sieve on setting 21, I get 10-11% on setting 30. To reach 8%, I’d probably have to move up to 32-33, which is a big difference in grind distribution from 21.

I haven’t had much luck finding numbers from anyone else online. The C40 is reputed for low fines, so the goal of this endeavour is to see if my grinder produces more than it should, or if the particle distribution has changed after years in use.

FFadedFrontiers · 6 Jan

Unfortunately I don’t have a Kruve to test but spotted this which provides numbers for clicks 2-20: https://mimcoffeelab.com/en/articles/investigating-the-size-distribution-of-coffee-particles-with-the-comandante-c40-grinder/

Also different x25 models have different burrs apparently, the forest version comes with the tigershark burrs but the others come with the same ones as c40.

MMWJB · 6 Jan

FadedFrontiers Those charts are very difficult to understand, as are most non-cumulative plots. It’s not clear to me how you get the same proportion of grinds under 400Kruve at both 8 & 12 clicks without some sort of glitch.

The Lightells plots that I have seen also have a strange binning/interval structure and give a different D50 dpending on whether you start at the largest, or smallest end.

Zephyp · 6 Jan

FadedFrontiers Oh yeah, I think I’ve seen that before.

@MWJB Yeah, there’s something weird going on there. Do I understand correct that all columns start with 100% into the first sieve (1000+), then the number on each row represents how many percent stopped at that sieve? The percentage that went through the 400 sieve at 10 clicks were 16% and 5% at 20?

If so, it falls more in line with your results than mine.

Socratic Coffee posted plots including the C40 some years ago, but I never found out which setting they used.

MMWJB · 6 Jan

Zephyp The percentage that went through the 400 sieve at 10 clicks were 11% and 5% at 20?

I’m reading it as 4% went through the 300 and 5% went through the 400 at 20 clicks - subsequently 9%total under 400.

Zephyp · 6 Jan

MWJB That makes more sense. I meant 16%, not 11, but that doesn’t work since the numbers can go up from one sieve to another. It’s a weird and confusing way to show the data.

MMWJB · 6 Jan

Zephyp Note that setting 9 has a total of 109%

Zephyp · 6 Jan

Yeah. Could be some input errors in there.

I found a guide in Norwegian with some information (I’m from Norway).

https://www.tek.no/nyheter/guide/i/8meWed/derfor-boer-du-kverne-kaffe-for-haand

Their results:

Setting 16: 9% < 400
Setting 18: 8% < 400
Setting 20: 6% < 400
Setting 22: 5% < 400

That’s the third source which has a similar number around 20.

If all of this this correct and there is something off with my grinder, it could explain some results I’m getting. Low extraction, high percentage <400 and long brew times. I don’t know if wear, misuse or something can cause something like this. The percentage below 400 is one data point, but what about the rest of the distribution above 400?

There are a lot of uncontrolled variables in these numbers, but it is interesting. If my brother didn’t live a six hour drive away I could compare to his C40 that’s only several months old.

MMWJB · 6 Jan

Zephyp You definitely don’t have Redclix installed?

Zephyp · 6 Jan

MWJB 100% sure I don’t. I found a video on Youtube showing both threads. The one to the right is mine. I got the original, black turning knob, which wouldn’t fit the Red Clix threads.

I found some results from when I first bought the Kruve in 2017 and they are simular to what I get today. 13% on 28 clicks. Maybe the model I got is different from the rest I’ve found data on.

FFadedFrontiers · 6 Jan

Anything visibly wrong with your burrs or alignment either? Commandante apparently offer a lifetime guarantee so if you get in touch they may be able to help you out.

Also another not very useful set of numbers since it doesn’t state grind setting (somewhere in the filter range) but there’s this from Coffee Chronicler’s review:

MMWJB · 6 Jan

The Akantus bearing holders/spacers are just an interference fit, I guess they could shift over time?

Zephyp · 6 Jan

FadedFrontiers To my eye they look fine, but I’ll find some high res photos online to compare with. I have thought about contacting them, but wanted to try figuring out if there’s actually something off with it first, and gather as much info as I can. If I have to send it to Germany and someone look at it, the cost might be too high. I can try e-mailing them with what I got and maybe they can explain it.

For all I know, it’s perfectly fine.

MWJB There’s a tiny bit of play on the axle and with the burr installed it wobbles a little, but very little. I can try getting some photos or a video. From zero (where the handle doesn’t fall down, held by the burrs), the burrs touch a little on settings 1 and 2, but from 3 and up there’s no sound or feel of contact.

I can’t see anything being off with bearings, bearings housing or any other part really. IIRC, it had zero visual wobble years ago, but I don’t know if that can be the reason for the grind distribution conundrum.

MMWJB · 6 Jan

Zephyp What is the largest sieve size that you have? (To see what kind of % you can capture between them

Zephyp · 7 Jan

MWJB The ones I got are 400 and 800. I’ll try them both today.

Zephyp · 7 Jan

What’s the best way to use multiple sieves?

I tried just putting the 800 on top of the 400, using the same two minute method, but got a much lower number below 400 than when I only use 400. I did disassemble and play with the grinder yesterday, so something might have changed, but I see now that it says non-interlocking on your sheet.

MMWJB · 7 Jan

Zephyp Non interlocking on my sheet just means that the >1200 doesn’t have the >1400 subtracted, so it’s the amounts you see if you only used one or the other. The 400 proportion assumes one or the other of the larger sieves is in place.

I like to use the largest sieve available above the 400, so you can capture a meaningful % between the sieves, smallest would be 1100, more typically 1200, or 1400. As large as 1600 for the coarsest grinds. (The amount captured will decrease as you go coarser on grind setting).

You have 28% between 400 & 800 with 10% below 400 - This, without other direct comparisons with C40, seems reasonable compared to readings from my Feldgrind V1 (27%) and an OE Apex (29%).

Zephyp · 7 Jan

Shaking three minutes rather than two with both sieves installed on setting 26 gave me:

59% > 800
27% between 800 and 400
13% < 400

That is more similar to the other data.

Would be interesting to try some brews which fits the 800 > < 400 numbers and remove some of the fines. Do you have those numbers on your current Rombouts setting with the Ode? Or an estimation. I only found < 1200 and 1400 in your sheet.

MMWJB · 7 Jan

Zephyp You can’t remove fines for drip/pour over.

I almost never use the 800 Kruve. However, for the Ode I got 18% at 700 and 28% at 900. From other sifting results, I’d expect the peak from the Comandante to be at a higher Kruve range (e.g. a coarser overall grind), so maybe you’ll see less % at 700-900 at a comparable setting?

Zephyp · 7 Jan

I just remembered about the Germans. I’ve been browsing Kaffee Netz before regarding something about the C40. Of course there are a lot of talk about microns and Kruve results.

https://www.kaffee-netz.de/threads/comandante-c40-mk3-schon-erfahrungen.95950/

I haven’t gone through all of it, but from a few samples, others got results ranging from sub 10% to over 15% for clicks between 20-25. I can’t verify the techniques used, but at least I found others with similar numbers to mine.

Maybe I should just keep testing and see if I can find a setting that works. If I don’t, I can test with my Lido 3 or Wilfa Uniform to see if I can get something different. I don’t know if comparing grind distribution numbers between grinders works as well as I’d hoped, when even different grinders of the same model get different numbers. It is after all only one or two data points from a grind distribution that spans a larger area.