Cafelat Robot - Dosing vs Grind

A haven dedicated to manual espresso machine aficionados.
VoidedTea
Posts: 191
Joined: 4 years ago

#1: Post by VoidedTea »

A week ago, in a coffee forum, I asked about dealing with citrusy blends that become more acidic when cooling down. One suggestion was to reduce the dose and make the grind finer with the Robot. I've tried this method but the results were quite watery and still too acidic to my liking. So my initial reaction was to stick to the beans that work well with my technique and I shifted my focus from beans to pressure control (hence the flow profiling post that followed). Today, I decided out of curiosity to go in the other direction and try increasing my dose and grind with some citrusy blends. I picked up a blend that I thought was interesting but could never give me any excitement in taste (again, too acidic in the end, when cooling down). I went with 21 grams and the grind which was at the top of my usual espresso range. I also applied my visual flow control technique that I described in the other post, which relies on the crema look and not on the gauge. No preheating either. Surprisingly, the result was one of the best shots I could remember with the Robot. I deliberately left the cup to last longer until it was completely cool and there was no acidity, or bitterness, whatsoever. It was very tasty and sweet, with very pleasant aftertaste still present in my mouth after 20 minute of finishing the cup. I will try the same overdosing approach with even more difficult and citrusy roast later today or tomorrow, but intuitively I expect very similar results. So again, I am getting more and more convinced that for whatever reasons, Robot's internal mechanical and/or physical properties work best with higher dose and coarser grind, even with more difficult and naturally acidic roasts (as indicated on their labels).

jtrops
Posts: 500
Joined: 9 years ago

#2: Post by jtrops »

It sounds like you finally got the right setup for the grind.

From your original description you were under extracting. That being said I could understand grinding finer, but I wouldn't grind finer, and reduce the dose in one step (that advice seems a little suspect to me). In fact, from what you said I wouldn't have considered a lower dose at all. Change one thing at a time to see the effect. The solution for under extraction will vary but it will include: grinding finer, dosing higher, or more heat during extraction (or a combination of the three). On the Robot you don't have as much control over the extraction temperature, so the grind and dose will be your tools of choice.

One day I hope to join the Robot club!

lessthanjoey
Posts: 362
Joined: 4 years ago

#3: Post by lessthanjoey »

VoidedTea wrote:A week ago, in a coffee forum, I asked about dealing with citrusy blends that become more acidic when cooling down. One suggestion was to reduce the dose and make the grind finer with the Robot. I've tried this method but the results were quite watery and still too acidic to my liking. So my initial reaction was to stick to the beans that work well with my technique and I shifted my focus from beans to pressure control (hence the flow profiling post that followed). Today, I decided out of curiosity to go in the other direction and try increasing my dose and grind with some citrusy blends. I picked up a blend that I thought was interesting but could never give me any excitement in taste (again, too acidic in the end, when cooling down). I went with 21 grams and the grind which was at the top of my usual espresso range. I also applied my visual flow control technique that I described in the other post, which relies on the crema look and not on the gauge. No preheating either. Surprisingly, the result was one of the best shots I could remember with the Robot. I deliberately left the cup to last longer until it was completely cool and there was no acidity, or bitterness, whatsoever. It was very tasty and sweet, with very pleasant aftertaste still present in my mouth after 20 minute of finishing the cup. I will try the same overdosing approach with even more difficult and citrusy roast later today or tomorrow, but intuitively I expect very similar results. So again, I am getting more and more convinced that for whatever reasons, Robot's internal mechanical and/or physical properties work best with higher dose and coarser grind, even with more difficult and naturally acidic roasts (as indicated on their labels).
Controlling the flow and preinfusion (perhaps holding around 2-3 bar for 10-20s or so first can make a huge difference in light roasts.

cap2
Posts: 93
Joined: 3 years ago

#4: Post by cap2 »

VoidedTea wrote:A week ago, in a coffee forum, I asked about dealing with citrusy blends that become more acidic when cooling down. One suggestion was to reduce the dose and make the grind finer with the Robot. I've tried this method but the results were quite watery and still too acidic to my liking. So my initial reaction was to stick to the beans that work well with my technique and I shifted my focus from beans to pressure control (hence the flow profiling post that followed). Today, I decided out of curiosity to go in the other direction and try increasing my dose and grind with some citrusy blends. I picked up a blend that I thought was interesting but could never give me any excitement in taste (again, too acidic in the end, when cooling down). I went with 21 grams and the grind which was at the top of my usual espresso range. I also applied my visual flow control technique that I described in the other post, which relies on the crema look and not on the gauge. No preheating either. Surprisingly, the result was one of the best shots I could remember with the Robot. I deliberately left the cup to last longer until it was completely cool and there was no acidity, or bitterness, whatsoever. It was very tasty and sweet, with very pleasant aftertaste still present in my mouth after 20 minute of finishing the cup. I will try the same overdosing approach with even more difficult and citrusy roast later today or tomorrow, but intuitively I expect very similar results. So again, I am getting more and more convinced that for whatever reasons, Robot's internal mechanical and/or physical properties work best with higher dose and coarser grind, even with more difficult and naturally acidic roasts (as indicated on their labels).
Since reading this the other day, I have been I have been experimenting with your approach. Very promising results so far.

VoidedTea (original poster)
Posts: 191
Joined: 4 years ago

#5: Post by VoidedTea (original poster) »

Agreed, keep it as simple as possible, finding beans that work well with my Grinder/Robot combination, not the opposite. I know the taste that I like and if I can get it with the Robot, I don't mind sticking to it. I value simplicity of the process more than the opportunity of discovering something new in coffee taste. And one word on preheating - recently I started preheating the piston by dipping it into my cup filled with hot water while I am waiting for the kettle to boil. But the only reason I started doing it is not to change the taste of the coffee, but to clean the piston of coffee residue from previous shot. I noticed that using just my fingers is not enough to clean the silicone seal, so I decided to use hot water instead. I hope by doing this I may potentially prolong the life of the seal because leftover coffee grinds may wear it down faster. Maybe not, but I just feel better doing it regularly.

cap2
Posts: 93
Joined: 3 years ago

#6: Post by cap2 »

I was struggling to get a good pull with some light roast supermarket beans (Starbucks Veranda). I thought of this post: I found grinding 21 grams so fine that I it took 12-13 bar to get 25 grams out in 25 seconds gave amazingly good results. Not so good I would use these beans if I had a choice, but still very enjoyable.


KInda painful to push on those robot arms hard enough to get 12 bar. Longer, rounder arms would be a big improvement IMHO. One does not always or even usually want such a high pressure, but it need not be painful to get it when you do.

Jonk
Posts: 2219
Joined: 4 years ago

#7: Post by Jonk »

cap2 wrote:light roast supermarket beans (Starbucks Veranda).
They really should not be allowed to call those light :wink:

Anyway, why don't you lower the dose instead if you want to use a finer grind?
My theory is that grind and ratio are the two main factors to influence flavor here. Increasing the dose is sometimes required to be able to use a certain grind setting, or in your case perhaps lowering it to end up with a less strenous pressure.

https://www.cafelatstore.com/collection ... obot-hands help if you want to reach 8-10 bar fairly comfortably. I think it's supposed to be a little painful to go much above that to prevent users from harming their Robot (especially for those without a gauge).

About the grind setting, since I've started measuring and taking notes I've noticed a trend where as I adjust finer the extraction increases to a certain point, perhaps near where VoidedTea pulls his updosed shots - explained by Christopher Hendon last year - only to drop when the grinds are starting to clog the flow and causing channeling.

But going finer still and doing things to alleviate channeling like using a paper filter in the bottom, using an extended pre-infusion or bloom (and that means pre-heating as well).. extraction will increase again and even surpass the coarse setting.

mikelipino
Posts: 258
Joined: 3 years ago

#8: Post by mikelipino »

I'd agree with the above about lowering dose for a finer grind on lighter roasts. This is especially true if you suspect the distribution on your grinder is fairly wide, as the large number of ultra fines increase the pressure needed extract in a reasonable time, but then cause channeling with longer times and higher pressures. Lowering the dose makes the puck thinner allowing water to more easily pass through, both with less pressure and less time, and because of that less potential for channeling.

A tighter distribution grinder would allow for a finer grind with larger dose, because while the average particle size would be around the same (affecting overall extraction), you would have fewer ultra fines to clog the filter. That's going to be my test once I get that nice grinder :D

cap2
Posts: 93
Joined: 3 years ago

#9: Post by cap2 »

Jonk wrote: I think it's supposed to be a little painful to go much above that to prevent users from harming their Robot (especially for those without a gauge).
.
I think the uncomfortable hands are a result of the design aesthetics, not an ergonomic choice. Paul himself runs some very high pressures in some of his videos when maxing out dose sizes. He says it won't hurt the machine.


I need the high pressure 1:1 pull to get good espresso out of those beans. Maybe there are better solutions.

VoidedTea (original poster)
Posts: 191
Joined: 4 years ago

#10: Post by VoidedTea (original poster) »

Hi all,

Just wanted to reply to my own original post based on some of my more recent experimentations with the Robot. Which will pretty much contradict my original statement and resonate with the advocates of finer grind. What I recently discovered by trial and error is how to successfully overcome channeling and chocking problems with finer grind which I experienced at the early days of working with the Robot.

First, while working with coarser grinds, I learned the longer preinfusion and more tactile flow control can significantly mitigate channeling problem as well as the need for high 7-8 bar pressure. I called it "go with the flow" technique and shared my discoveries a few weeks ago on this forum. My next discovery, which put the final nail in the coffin of "coarser grind" approach, was an additional technique to longer preinfusion I happened to learn recently. There is probably a term for this technique, but I don't know for sure. I came across it either here or on YouTube when someone was talking about DE1 and mentioned that one profile allows for a rapid change in pressure during preinfusion to help saturate the puck. So I said to myself, "that sounds like an easy thing to try on the Robot". So I did. During preinfusion phase, instead of holding low pressure steady, I gently "pump" my hands at about one second intervals. I do fully understand that one cannot raise Robot arms during pull, so this is not what I am doing. Just a very gentle, barely noticeable, "pump", almost like a heart pulse in your arms. The results were very pleasantly surprising. The flow was even better and easier to achieve. So I started experimenting with finer grinds, eventually going all the way to the lowest setting on my grinder - zero. Unlike before, when using such a fine grind would definitely choke my Robot, I had absolutely no problem pulling a perfect shot with my new technique. I did lower the dose from 21g to 14g at the finest settings, of course.

But the most important part is that I liked the taste of finer extraction so much better. Not only the shots with 14g tasted stronger than with 21g, pleasantly stronger, no acidity or bitterness, they also revealed a whole pallet of new notes, which I could not pickup before. It also helps save on coffee for those who care. So there you go, I have completely converted myself from the "coarse grind" newbie, to a "fine grind" aficionado :D. Just shows that learning is a journey, not a destination.

Post Reply