Decent Espresso Machines (DE1) - Any thoughts?

[>decentespresso]

How is it possible to have a 4ml/sec flow for 10 seconds? Wouldn't an 18 gram coffee puck be over saturated at 4ml x 10 sec? Am I wrong to think that 40ml of brew water be too much for the coffee puck to hold it in? Is something else in play like the excess water draining out somewhere during pre infusion?

Well spotted!

You're right, it isn't possible, which is why there is a "4 bar start" choice (in the hold section) feature. When the pressure reaches 4 bar, the preinfusion stage exits, regardless of how many seconds of preinfusion you set.

In the shot above, you'll see a vertical gray line drawn around 6.5 seconds, and this indicates the end of preinfusion, at around 26ml total water put into the puck. Preinfusion ended there because 4 bar of pressure was detected. In that shot, preinfusion was cut short 3.5 seconds early.

The reason for that feature is so that:
1) you can set preinfusion flow rate to whatever you like
2) but preinfusion exits if a desired pressure occurs, otherwise the preinfusion lasts the number of seconds you indicated
3) this also allows Slayer-style preinfusion, so that you could (for instance) do 1 ml/s for 25, require 4 bar of pressure, which would cause the puck to compress before switching back to flow control mode. Without this puck compression step, you could end up not having any pressure at all. Note that if you want to do that, you still can, you just set the "start pressure" to 0.

[>decentespresso]

News: final proof of sheet metal, Ultem, brass pieces.

We're about to receive the final sheet metal "factory proofs" for our chassis, and all the Ultem and brass pieces too. We need to check these all for final accuracy before we give approval for mass manufacture.

[MrJack]

As of today we’re able to make flow controlled espresso shots.

I don't know of any other espresso machine able to do this, so this could be a world premiere.

The screen shots below are the results of our first real (coffee, not Scace) fully flow controlled shot, which just occurred a few minutes ago.

[ATTACH=CONFIG]16145[/ATTACH] [ATTACH=CONFIG]16146[/ATTACH] [ATTACH=CONFIG]16147[/ATTACH]

Had a bit of a look at the data from this shot last night (extracted from the images using WebPlotDigitizer). Made a few different plots (including some with derived / aggregate variables). The result was rather interesting!

Would be very keen to see some more (in particular one with a fixed pump speed).

[JMcCee]

I've been asked many times "how do you fill the water tank on the decent espresso machine?" So I made a movie today showing how.

Thanks for posting. Does the drip tray lock in or is it just sitting there and subject to movement or knocking?

[>decentespresso]

Thanks for posting. Does the drip tray lock in or is it just sitting there and subject to movement or knocking?

It passes by very quickly in the video, but at 0:01 you can see that there are large "dimples" in the bottom front of the drip tray.

These fall into cut outs in the leg rails and thus lock it into place.

Here's a screen picture and I've put an arrow to them.

[>decentespresso]

flow profile editor evolution

The Flow Profile shot editor is undergoing rapid change at the moment, because I took my first attempt at it to London, where I'm here with Scott Rao, and he immediately poked holes in some of my assumptions.

Below is today's iteration, trying to address Rao's points.

Even though this is a flow controlled shot, there is
(a) an optional exit during preinfusion if a given pressure is reached, and
(b) an option for a guaranteed minimum pressure after preinfusion, in case you want flow control, but you want to make sure the puck was compressed.

Scott and I are with James Hoffmann tomorrow, who I'm sure will poke more holes in our assumptions.

Hopefully, with all this expert advice on what we're doing, we'll be able to find a happy compromise between power and simplicity. It's a work in progress at the moment. My goal is to have something relatively well thought out for next week's World of Coffee conference in Budapest.

[JMcCee]

It passes by very quickly in the video, but at 0:01 you can see that there are large "dimples" in the bottom front of the drip tray.

These fall into cut outs in the leg rails and thus lock it into place.

Here's a screen picture and I've put an arrow to them.

[ATTACH=CONFIG]16187[/ATTACH]

I like it. Thanks for clarifying.

[>decentespresso]

Had a bit of a look at the data from this shot last night (extracted from the images using WebPlotDigitizer). Made a few different plots (including some with derived / aggregate variables). The result was rather interesting! Would be very keen to see some more (in particular one with a fixed pump speed).

Here are two shots, zoomed and not, with different grind settings, both set to provide a 1.6ml/s flow rate. It looks like I had different preinfusion timings on the, which you can see by the vertical grey line indicating the end of preinfusion.

On one, the grind is fine enough to reach a 8.7 bar pressure, and the pressure stays constant. In the other, the grind is too coarse to create much pressure, and so pressure slowly ramps up over the shot to end around 3.5 bar. What's interesting about this shot is that it still tasted good and had plenty of crema.





I'm working today on a "historical shots" feature, automatically saving every shot to a history/ directory, with all temp/pressure/flow and settings. These will be text files, and if there's any shot I do that looks interesting, I can post the raw data behind it. I often have trouble recreating a shot that I really liked, as I fiddled with settings afterwards, and I'm hoping that helps with that problem.

Regarding derived data, the dotted line on the bottom is the flow rate change, but only the positive component of it. Loss in flow is not currently charted, as it didn't seem useful enough to warrant the added visual complexity.

[gc]

I'm working today on a "historical shots" feature, automatically saving every shot to a history/ directory, with all temp/pressure/flow and settings. These will be text files, and if there's any shot I do that looks interesting, I can post the raw data behind it. I often have trouble recreating a shot that I really liked, as I fiddled with settings afterwards, and I'm hoping that helps with that problem.

I'm going to really appreciate this feature as I am sure I'm going to have the same trouble - as I already do when I'm trying to use a notepad for the same job.

[>decentespresso]

I'm going to really appreciate this feature as I am sure I'm going to have the same trouble - as I already do when I'm trying to use a notepad for the same job.

Since it's not much data (about 4kb) for a shot, I decided to
(a) save all the historical pressure/flow/temperature data, but also
(b) all the settings that were in place for every shot
(c) able to chart historical shots, and leave that historical shot charted in place as you make the new shot, so you can compare
(d) able to load any historical shot, with all settings intact.

Essentially, every historical shot becomes a "profile" that can be brought back.

At the moment, I've programmed (a), (b) and (d). I'll get (c) done after my European tour, in a few weeks.

[>decentespresso]

Four engineering interns started this week

Because they're all engineering students, we can put them to work on fairly sophisticated tasks. We have them designing stations to automate testing of our components, and we're also having them create CAD drawings of standard parts that we buy off the shelf (that therefore have no drawings).

Man Yi turned around the drawings below in an hour, which is really impressive, so it looks like we're going to give her more things to draw.

[Shortyesquire]

Four engineering interns started this week

Because they're all engineering students, we can put them to work on fairly sophisticated tasks. We have them designing stations to automate testing of our components, and we're also having them create CAD drawings of standard parts that we buy off the shelf (that therefore have no drawings).

Man Yi turned around the drawings below in an hour, which is really impressive, so it looks like we're going to give her more things to draw.

[ATTACH=CONFIG]16204[/ATTACH]

It's a good thing you're in Hong Kong.

Millennial interns in Australia would have spent the first hour asking when they could start their company funded mba and then join the exec team before becoming ceo.

They would spend the second hour comparing paleo bread and civet coffee from the local cafes.

[gc]

Since it's not much data (about 4kb) for a shot, I decided to
(a) save all the historical pressure/flow/temperature data, but also
(b) all the settings that were in place for every shot
(c) able to chart historical shots, and leave that historical shot charted in place as you make the new shot, so you can compare
(d) able to load any historical shot, with all settings intact.

Essentially, every historical shot becomes a "profile" that can be brought back.

At the moment, I've programmed (a), (b) and (d). I'll get (c) done after my European tour, in a few weeks

Will there be an easy way of renaming/grouping/tagging the logged settings file that's created with each shot setting? eg "a really great shot#1xxxdate". I realise this can be done with any file manager once save location is known, but within the app would be nice

[Erimus]

Will there be an easy way of renaming/grouping/tagging the logged settings file that's created with each shot setting? eg "a really great shot#1xxxdate". I realise this can be done with any file manager once save location is known, but within the app would be nice

It'll be logged under God Shot :-)

[Magic_Matt]

It's a good thing you're in Hong Kong.

Millennial interns in Australia would have spent the first hour asking when they could start their company funded mba and then join the exec team before becoming ceo.

They would spend the second hour comparing paleo bread and civet coffee from the local cafes.

Internships are generally illegal under Australian law... 🤔