AssafL wrote:David -

I wonder if you elucidate on some of the "flow profiling" capabilities of the system.

1. Units in the app: The phone app with that stepped flow profile. What are the units that are shown as 1-9? Obviously those are not ml/min (as those run anywhere from 60-120ml/min for PI and drop later on). Is this pump speed? or lever position?

2. Flow profile behavior: For flow profiling - as the headspace fills up - continuing the same flow becomes harder (if at all possible) for the pump; there is no longer any air and the pump has to overcome the resistance of the puck. How does the Mina handle this? Can you hear the pump trying to keep up the flow?

3. If there are other understandings or comments about the way this flow-profiling works I'd be grateful for those as well.

FWIW - I am asking because I am working on my own flow profiler and it is a tricky endeavor (due to the "new law of thirds"). As an example, say I am at a water debit of 200ml/min (entirely possible with the GS/3 and gear pump) during PI - once PI is over and pressure start increasing the machine will no longer be able to sustain 200ml/min and flow rate will drop to, say, 80-100ml/min.
I just wonder what the pros like DC (and Decent) do about this... My current approach flips to pressure profile.

So, at the present time, the units in the app correspond to real ml/s out of the grouphead with no portafilter added. They go from 2 to 12, so you could convert those to ml/min if you'd like (120-720ml/min).

2. The flow you set from the DFR (the restrictor) is the maximum flow you allow the machine to output. So, if the puck only lets out 1 ml/s and you are pushing 8 ml/s, you'll just see the pressure go up to 9 bar (or whatever you set it at from the pump). With rotary pumps, the pump doesn't try to keep up, since the flow it can deliver is a lot greater than the flow needed by the machine. Even on the highest setting, you still restrict the flow from the pump.

This is similar with the Slayer needle valve, and there is a lot more material online about the Slayer than on the Mina. The only difference is that this is electronically programmed, while on the Slayer it's fully manual.

The Mina is only profiling flow. Pressure gets profiled as a consequence. For instance, if you have a flow rate of 1.5ml/s in the cup (measured by a scale) and you only let 2ml/s of water into the grouphead (from the DFR setting), then chances are you won't be at 9 bar, because the system is losing a lot of water compared to what you put in.

Or, if you start the shot at 2ml/s, it'll take some time until the headspace gets filled with water, and you will be at a low, close to zero, pressure until that time. You are just wetting the puck without pressure added. When you see the manometer get up to pressure, then you can start pushing more water and you will get a normal extraction.

A preinfused puck will run faster than a dry puck, so you can use a really fine grind.

Seems like we need to talk.

Thank you for the comprehensive answer. Cool. So it is a rotary pump and a variable (5 step) gicleur.

In effect it isn't flow profiling in the sense of measuring and controlling the flow as much as profiling the water debit. (The same is with Slayer, BTW, except 2 step rather than 5). The lever sets the compliance (maximum flow) available for the puck to drink.

Should enable for variable doses (singles grind fine, reduce flow to 1 and wait for first drops ; triple grind coarse and level 2 etc). In that respect the lever is not just decorative - it should behave almost like a lever machine.

I wonder if the 5 step gicleaur has sufficient resolution to taper the pressure as flow increases in the last 1/3 of the pull (like a lever would do do to spring expansion)? It doesn't add much to the cup but it is yet another difference between levers and pumps. (I am trying to parameterize that last 1/3 and coming up empty handed as to value; still do it but not sure why...).

DFR should indeed be highly effective towards getting great espresso IMHO. Why are they such a rarity in the marketplace?

Yup. A rotary pump and a variable gicleur. But it actually has 20 steps (2-12ml/s in 0.5ml/s increments). You can set 5 steps for your extraction. But each individual step can have any of the 20 possible values.

So it is not flow profiling, if by flow you mean the flow of coffee in the cup. It is exactly as you said, you set the max flow available for the puck.

It can act as a lever machine if you mimic the flow profile (or corresponding pressure profile, since a lever manipulates pressure).

From what I can tell by using it, you only need 1 or 2 steps for the last part of the extraction. So if I'm at 10ml/s mid-extraction and, let's say 9 bar and 25g in the cup, I move to step 4 which is, let's say 4ml/s (pressure drops to 8 bar). them in a couple of seconds I move to step 5, which is 2ml/s (pressure drops to 5 bar). Which, from what I saw, cuts back on the harshness / astringency and allows more sweetness to come through in the shot (if any in the beans).

Almost all my shots are with a 5-7s ramp-up to pressure, usually steps 1 and 2. Step 3 is for the extraction, and 4 + 5 for the ramp down.

Yesterday, I opened a new bag of coffee and my first shot was with a non-preinfused profile. It came out in 72 seconds. It was drinkable, but a full chocolate / cocoa flavour. Interestingly, no bitterness (a natural processed Ethiopia Gigessa). Just for the experiment sake of it, since I was with a friend, I kept the same grind and dose and I used a 2ml/s first step, until the brew pressure reached 9bar (about 10 seconds). Then I switched to 10ml/s and I had a really good flow in the cup, and then ramped down at the end.

With the same grind and dose I now had a 41 second shot that was a tad unextracted. All because I slowly wet the puck and (probably) fixed the fines in place. When I let more water in, the puck offered less resistance and I got a faster shot.

FWIW - on Jake's ramble thread we think we know why with PI pulls flow faster and pull better.

The reason is if the puck is still partially dry, that part of the puck (fissures, cracks and void) would collapse since they contain air, which unlike water, does compress. So the density in the bottom of the puck increases reducing the flow until it absorbs water and slowly expands back.

Jake actually found a nice scientific treatment (of filtering under pressure) that describes these processes. As well as fingering flows which probably explain why wider pucks perform better.

Your pull strategies sound really good (you thought them through). You did not state this but I'd wait for first drops to appear (like you would in a lever). May require you to grind even finer.

From your descriptions I really like the DC approach. Seems like DC really thought this through. It also seems like instead of showcasing just how smart their approach is (like Slayer did), they oversimplified it in the marketing (and ended up sounding like hype).

Thank you for sharing!

Radu, it are you using the Kinu M47 for your pulls?

I used a Compak E5 for about 2 months, a Dalla Corte DC One for about a month and at the time I am using the M47. I see no difference between the M47 and the E5, but the DC One is a different bird.

Thank you for the well reasoned, well written review. Profiling is tricky for many here on HB so any review from people who actually profile is refreshing.

Do you happen to know who makes the proportional valve used in the Mina (the "DFR")? Burkert or someone else?

Also - if you opened the machine and have internal pictures those are always welcome.

Very nice review and pictures!

PS What lens did you use to take those shots ?

I am using the Mina for about 1.5 years. It is a lovely peace of technology. The patented DFR valve is invented by DC and not available over the counter:


here aare a few profiles to give you an idea what is possible
https://www.dallacorte.com/en/magazine- ... n-121.html
https://www.dallacorte.com/en/magazine- ... n-135.html