Indeed this is something I've experienced on Mazzers before, although I don't think it's a big issue as the burrs are a) pushed apart by springs b) pushed apart by beans when coffee is ground. This "experiment" shows that there's slack in the thread and I would say that's pretty much it. I do understand your comment about similar sounds being generated when pushing on each of the 4 quadrants, but I'd say this method is far from perfect when trying to measure or even judge misalignment (you'd have to be able to compare the sounds somehow and also make sure the same amount of force is applied in the exact same spot in each quadrant).

On the ZR I've had 0.02mm of slack in the adjustment mechanism, but I'm currently working on bring this down to below 0.01mm. I can measure the burr gap on the fly as the grinder is being used, so I see how static and dynamic scenarios differ from each other.

T.

While adjusting my Mazzer Robur today, I have come up with another question/thought (again).

The one notch of Mazzer adjustment seems to correspond to 0.015mm (according to 2 sources at least). The one notch on HG-1 are calculated to be 0.0133mm. However, I felt that the one notch on the Mazzer gave a much larger flow fluctuation (5-7sec) than the one notch on HG-1(3-4 sec). Could this be used for alignment indication (or it might be due to the inherent slack on the Mazzer adjustment thread)?

Also, to put it into perspective (just realize this), when we have an alignment tolerance of 0.02mm and that we're adjusting the burr gap by 0.015mm, the 0.02mm seems to be much larger in that context and aiming for 0.02mm alignment didn't seem unreasonable - the gap change due to adjustment is less than your misalignment!

I guess this is also what Nick mentioned before but the reason didn't sink in completely for me. When you have a 'perfectly' aligned grinder, you get the full effect of the burr gap change of 0.015mm, for each notch you move. If you have a less 'aligned' grinder, the effect of adjustment is negligible compared to the inherent misalignment (you'd have to move more than one notch to see the adjustmet effect). That's why a less aligned grinder is more forgiving and vice versa.

This might sound like a heresy - if there's no noticeable negative effect in terms of taste/grind quality (something that needs proving) from minor misalignment, it might be even desirable to have/introduce some misalignment in your grinder!

samuellaw178 wrote:The one notch of Mazzer adjustment seems to correspond to 0.015mm (according to 2 sources at least). The one notch on HG-1 are calculated to be 0.0133mm. However, I felt that the one notch on the Mazzer gave a much larger flow fluctuation (5-7sec) than the one notch on HG-1(3-4 sec). Could this be used for alignment indication (or it might be due to the inherent slack on the Mazzer adjustment thread)?

It could be that the Robur is more misaligned than the HG-1, which in effect gives you more boulders / large particles even with small grind increments.

samuellaw178 wrote:I guess this is also what Nick mentioned before but the reason didn't sink in completely for me. When you have a 'perfectly' aligned grinder, you get the full effect of the burr gap change of 0.015mm, for each notch you move. If you have a less 'aligned' grinder, the effect of adjustment is negligible compared to the inherent misalignment (you'd have to move more than one notch to see the adjustmet effect). That's why a less aligned grinder is more forgiving and vice versa.

If we are talking about angular misalignment like here:

http://www.dacell.com/images/sub10/acc_04_01.gif

and your A - B = 0.015mm, but the misalignment is caused by the burrs not sitting perfectly flat in relation to each other (so not entirely like it's shown in the link above, as there the shafts / supports are also misaligned), you will maintain that misalignment through the entire adjustment range (assuming the thread section of the burr chamber is machined at 90deg to the shaft's axis of rotation). Burr gap change per notch won't matter here, you will always have the same delta between A and B.

T.

dsc wrote:It could be that the Robur is more misaligned than the HG-1, which in effect gives you more boulders / large particles even with small grind increments.

If we are talking about angular misalignment like here....
Burr gap change per notch won't matter here, you will always have the same delta between A and B.

T.

Thanks! The angular misalignment makes sense (I was too focused on the parallel misalignment alone).

Unfortunately, there's no way to check the alignment on the Robur as it's not easily accessible like the HG-1, which I can look through the burrs gap and feel the parallelism.

Another minor difference I've noticed vs the Hg-1, the flow rate seems to increase quicker on the Robur as the extraction progresses. Any idea why? Granted, we're comparing 83mm burrs vs 71mm burrs and they may very well give a slightly different particle distribution.

samuellaw178 wrote:The one notch of Mazzer adjustment seems to correspond to 0.015mm (according to 2 sources at least). The one notch on HG-1 are calculated to be 0.0133mm. However, I felt that the one notch on the Mazzer gave a much larger flow fluctuation (5-7sec) than the one notch on HG-1(3-4 sec). Could this be used for alignment indication (or it might be due to the inherent slack on the Mazzer adjustment thread)?

If you have two identical grinders, the one which is furthest away from the zero point is the best aligned.

In your case you have two identical burrsets, but they spin at different RPM's. Since you know the thread pitch, you could calculate how far from the zero point they are grinding. The one with the bigger gap is probably better aligned; although it's not apples to apples as they are running at much different RPMs.

In terms of your initial question, the better aligned your grinder is, the narrower the window for hitting a good flow rate. So little changes make bigger flow differences. So that may indicate your Mazzer is better aligned, but once again it's not apples to apples (way different rpms on each grinder). With the same grinder, distance from zero (burr gap) is the best indicator.

samuellaw178 wrote:Also, to put it into perspective (just realize this), when we have an alignment tolerance of 0.02mm and that we're adjusting the burr gap by 0.015mm, the 0.02mm seems to be much larger in that context and aiming for 0.02mm alignment didn't seem unreasonable - the gap change due to adjustment is less than your misalignment!

Think about it in terms of percent. If you shooting for a .2mm average grind size, and you have a .02mm axial misalignment. That's 10% your target grind size.

Many grinders are .05-.1mm out, which is 25-50%, just to give you an idea. This is why I'm banging the drum on alignment so much.

samuellaw178 wrote:This might sound like a heresy - if there's no noticeable negative effect in terms of taste/grind quality (something that needs proving) from minor misalignment, it might be even desirable to have/introduce some misalignment in your grinder!

Pro and a con to doing that:

Misaligned grinder = wider grind distribution = bigger the sweet spot in grind setting for getting a good flow = less even extraction = worse taste shot.*
Aligned grinder = narrower grind distribution = smaller the sweet spot for getting a good flow = more even extraction = better tasting shot.*
* Assuming you pull a "good" shot (you still need to do everything else right )
Which would you prefer?

You'll also find with a better aligned grinder, you'll also have more latitude for missing your target brew ratio, and it still tasting good.
Say you have two identical grinders. One grinder is about .01mm axially, and the other .05mm.

Hypothetically speaking, with the crappy aligned grinder, you may be pulling:
18g in = and maybe only 34-38g out will taste okay, and 36g is clearly the best. Or you get too much sour or bitter.
With a good grinder, you maybe able to get a good tasting shot from 32-40g. And at 36g it's going to be way better than the previous best (even the 32 & 40 probably still taste better than your previous best shot). So you have more latitude on hitting your target brew weight, and getting way better tasting shots.
Once again, what would you rather have?

People are still praising grinders for having wide grind distributions, because they notice how forgiving the grinder is in flow rate. They don't understand the trade off being made. So is introducing mis-alignment heresy? Depends on which camp your in. Personally, I'd take the aligned grinder with a tight distribution, every single time.

nickw wrote: People are still praising grinders for having bad alignment, because they notice how forgiving the grinder is in flow rate. They don't understand the trade off being made. So is introducing mis-alignment heresy? Depends on which camp your in. Personally, I'd take the aligned grinder, every single time.

Interesting. So what accounts for big conicals typically needing less adjustment than big flats? At least that's their reputation. That conicals are more "forgiving" and require narrower grind changes to dial in.

I'm assuming it's related to burr geometry, and how this affects particle distribution and the shape of each particle. Or is there something else in the mix, including alignment or lack thereof?

Sorry, I probably should of said "wide grind distribution" rather than alignment (I'll edit the above post). While alignment will make the biggest difference (it doesn't matter how the good the burrset is if it's mis-aligned) the burrset and other factors matter too (such as burr type/geometry/rpm/sharpness/texture/temperature/etc). Final grind distribution is a combination of everything.

That being said, it seems like current conics give a wider grind distribution than flats. I base that on a few things:
- The grind distributions we've seen (from Mahlkonig)
- People's real word experience (they have more forgiving of flow rate) and what that infers
- The EY's people are hitting with conics vs flats. Don't really see conics pushing 20%, where flats going 22-25%*
- The shots I've tasted. They always have more bitters/sours at the same time (not as even tasting). Alignments are unknown, but there is a common trend/taste

Truth be told, I don't think we yet know how good conics can be. Tom/Dsc was looking into that. Are the current distributions a result of burr cut/design, their conical nature, or both? The big companies may know, but if they do, they aren't sharing. I'd stil love to see/use/taste Tom's SR71.

* EY has it's problems too. I see it as a guide, not an absolute. It says how much was extracted, but not what was extracted, or what it tastes like.

nickw wrote:It seems like current conics give a wider grind distribution than flats. I base that on a few things:
- The grind distributions we've seen (from Mahlkonig)
- People's real word experience (they have more forgiving of flow rate) and what that infers
- The EY's people are hitting with conics vs flats. Don't really see conics pushing 20%, where flats going 22-25%*
- The shots I've tasted. They always have more bitters/sours at the same time (not as even tasting). Alignments are unknown, but there is a common trend/taste

Interesting Nick. It may be the comparison was done on a different basis, but my experiences have been I liked conical flavor better. You might have remembered I had a Mythos (hoppered dosing) versus Hg-1 (single dosing) side by side for a short period. My preference for pull is 1:2 brew ratio and medium-light-ish roast. At those parameters, I don't find the Mythos to perform any better than the HG-1 and I prefer the shot from HG-1 by a hair (not sure what you meant by the bitter/sour as I haven't had an EK shot often enough).

Just wondering, when we're comparing EY, can we also include TDS as that's an equally important vital? I believe you can extract up to 25% on the conical but the taste is going to be terrible (and low tds). So EY alone is not exactly an accurate indicator other than telling where the grinder performs best. Also, I recall reading somewhere that conical has more microfines that migrates into the shot, thus conferring a silkier mouthfeel (compared to a flat). That is most likely not detectable by the refractometer.



Addition: I am sure many of you have read this (fairly old article from David Schomer about microparticles migration). I find the article quite informative and probably quite ahead of its time. I didn't understand much of its significance until recently. http://espressovivace.com/index.php/mic ... g-systems/

Hi Sam, so many things I can respond too. And I have this bad problem with trying to be thorough.

samuellaw178 wrote:Interesting Nick. It may be the comparison was done on a different basis, but my experiences have been I liked conical flavor better.

In regards to "Conical (burr) flavor", I think we should be careful when we say "conical burr flavor" or "flat burr flavor". Otherwise we may end up mis-associating results with burr type rather than the resulting grind distribution.

Tom/Dsc may surprise us one day (and I'm not being sarcastic here) and come up with a conic which is the most even grind of them all. Even with flats, they're all different too... so I think we should be careful with putting things into boxes like that.


In regards to your preference, I think you bring up a good point, and that is: we all have different taste preferences. Which we haven't really talked about yet.

A narrow vs wide distribution will each taste different. Even if both brews are extracted to the same EY.

I much prefer a more even distribution, which all things being equal will be a more even extraction. But others may not (valid reasons as to why), which is why I'm not upset if others like something different than me. Hopefully everyone is finding their own nirvana
Either way the preference discussion and how that plays out, is another discussion itself.

samuellaw178 wrote:You might have remembered I had a Mythos (hoppered dosing) versus Hg-1 (single dosing) side by side for a short period. My preference for pull is 1:2 brew ratio and medium-light-ish roast. At those parameters, I don't find the Mythos to perform any better than the HG-1 and I prefer the shot from HG-1 by a hair (not sure what you meant by the bitter/sour as I haven't had an EK shot often enough).

I have no doubt you found that.

The Mythos seems to get mixed reviews, even from industry professionals who I know and trust. Some love it, others quite disliked it. Some said it gave delicious, even, and higher EY extractions (22-23%), while others said it was uneven and they couldn't break 20% EY. With the exact same make/model grinder and such different results (based on the unit tested) I think alignment is the most likely culprit. I wouldn't be surprised if they have large batch variation and you got poor example of one.

samuellaw178 wrote:Just wondering, when we're comparing EY, can we also include TDS as that's an equally important vital? I believe you can extract up to 25% on the conical but the taste is going to be terrible (and low tds). So EY alone is not exactly an accurate indicator other than telling where the grinder performs best.

TDS while connected to EY isn't the same thing.
TDS (total dissolved solids) is the percent of solids dissolved in solution.
EY (Extraction yield) is the percent of solubles extracted from the coffee.
Note: Max EY is about 30% (by weight) - the rest is insoluble in water (cell wall and fibre).

EY (while good, and the best we have right now) has problems too. EY is objective in it says how much was extracted from the coffee. But it doesn't say what was extracted and in what proportion, or what it tastes like. You can have two 20% brews which taste very different. Higher doesn't always equal better, it just means more was extracted.

The difficulty with extraction isn't getting high EY. Boil coffee in a pot for awhile, you'll extract everything which is water soluble (about 30%). The difficulty is in being selective with the extraction. The difficulty is in trying to extract what we want, while leaving what we don't.

The more even we can be with the extraction, the higher EY we can extract before things start tasting bad. For example:
- With a robur the best tasting shot might be 20% EY (run it longer - say to 24% and it will becomes very bitter).
- With an EK43 the best tasting shot might be 24%. Stoping the same shot early at 20% EY would taste very different than the same 20% EY off the robur.

So to repeat: EY only says how much was extracted, not what it tastes like.

samuellaw178 wrote: Also, I recall reading somewhere that conical has more microfines that migrates into the shot, thus conferring a silkier mouthfeel (compared to a flat). That is most likely not detectable by the refractometer.

I think that's making an incorrect correlation again in regards to conics vs flats.

Either way, when using the refractometer (and filtering the espresso before hand) you end up filtering out all the particles in suspension. The filters I use (same filter size as VST) are .00045mm (.45um) filters. Thats just under half a micromter/micron!
ExtractMojo software and filters

samuellaw178 wrote:Addition: I am sure many of you have read this (fairly old article from David Schomer about microparticles migration). I find the article quite informative and probably quite ahead of its time. I didn't understand much of its significance until recently. http://espressovivace.com/index.php/mic ... g-systems/

David has done a lot for speciality coffee, and I'm thankful for all he's done. Although I disagree with him on some of his conclusions.

But hey, looking back in 10 years, I'll probably disagree with myself on some of my current conclusions.