Decent Espresso Machines (DE1) - Any thoughts?

[>decentespresso]

DE1CAFE question

John I've been sifting through old posts for info on the Cafe model and can't quite find what I want to know. So could possibly answer these quick questions regarding that model? Would be much appreciated.
• Roughly when?
• Rough price? (Or price range/ceiling)
• Will it be plumbed water only?

You haven't been able to find the answers because I've been tight-lipped about the DE1CAFE model, as I'd prefer to not blab on about "yet another espresso machine model we haven't shipped", instead leave that for future discussion, when the DE1/DE1+ are shipping, have been (hopefully positively) reviewed, and people are then asking "how can I get this for my café?"

But your questions are reasonable, so let me take a swing at them:
1) beta in January, shipping in April
2) Roughly AUD$5000
3) plumbed water is not required, you can choose to not plug in the external pump box, in which case you need to manually refill the water tank.

You can plumb the DE1CAFE (or the DE1PRO, which is shipping sooner) either in pressurized water, or simply dip a water tube into a 10-litre water container. That 2nd choice is my recommendation, actually, because then you can use nice filtered water to make your espressos.

Both plumbed and non-plumbed drip trays are included.

Our plans at the moment for steam are to have two 220V power plugs, one dedicated to steam, one for espresso. IF we didn't do this, the current espresso state would affect our power draw and thus change how much steam you're getting, which would be very annoying.

The three key features of the DE1CAFE are:
1) steam-during-brew and specialized steam wand
2) 15s to steam 200ml of milk from 10ºC to 60ºC. (this consumes 8amp at 220V)
3) it can be built into a table (though we're trying to design it so that isn't a requirement)

Have you seen this video? I put it on youtube recently but didn't post it here. It's still being worked on, as an intro to this model.

[LeroyC]

You haven't been able to find the answers because I've been tight-lipped about the DE1CAFE model, as I'd prefer to not blab on about "yet another espresso machine model we haven't shipped", instead leave that for future discussion, when the DE1/DE1+ are shipping, have been (hopefully positively) reviewed, and people are then asking "how can I get this for my café?"

But your questions are reasonable, so let me take a swing at them:
1) beta in January, shipping in April
2) Roughly AUD$5000
3) plumbed water is not required, you can choose to not plug in the external pump box, in which case you need to manually refill the water tank.

You can plumb the DE1CAFE (or the DE1PRO, which is shipping sooner) either in pressurized water, or simply dip a water tube into a 10-litre water container. That 2nd choice is my recommendation, actually, because then you can use nice filtered water to make your espressos.

Both plumbed and non-plumbed drip trays are included.

Our plans at the moment for steam are to have two 220V power plugs, one dedicated to steam, one for espresso. IF we didn't do this, the current espresso state would affect our power draw and thus change how much steam you're getting, which would be very annoying.

The three key features of the DE1CAFE are:
1) steam-during-brew and specialized steam wand
2) 15s to steam 200ml of milk from 10ºC to 60ºC. (this consumes 8amp at 220V)
3) it can be built into a table (though we're trying to design it so that isn't a requirement)

Have you seen this video? I put it on youtube recently but didn't post it here. It's still being worked on, as an intro to this model.

Thanks for once again being so prompt and generous with the info you share John. That answers everything very nicely.

[>decentespresso]

Graphic 182 Decent Espresso Shots

Thanks to "chemical engineer/coffee nerd" DIYCoffeeGuy for plotting 182 espresso shots we've made on my DE1+ espresso machine.

We do a huge amount of experimenting with grinds, grinders and roasts, so there's a lot of data all-over-the-place. The "pressure vs time" chart isn't especially enlightening to me (it shows that we're experimenting a lot and our pressure is all over the map). I'm not the only one using this machine, so lots of other people's experiments are plotted here, which accounts for some of the variability.

However, the flow vs time chart is a lot more interesting. Here are some takeaways from the data:
1) our pre-infusions are mostly around 4ml/second, but plenty around 2.5ml/second too
2) most shots have their pucks compressing around 6 seconds, and fully compressed by ten seconds (by looking at the slowdown in flow possible through the puck)
3) most of our shots are around 2.5 ml/s during the main extraction, though there is a healthy number of slower extractions around 1.8 ml/s
4) most of our shots end around 25s, some go to 35s, but very few beyond.

This sort of visualization is interesting to me for a few reasons:
1) it helps prove that I (the DE1+ tablet programmer) don't have all the answers and that others have other perspectives on the same data
2) data export from the DE1+, especially over a large sample size and in a cafe production use, could yield interesting and new insights. This sort of data is not usually collected outside of lab settings.
3) some of these visualizations that people do will make their way back into the DE1+ tablet software, other will be programmed by either me or anyone-who-wants-to using Javascript, once I integrate our espresso machines into the cloud, and make this data publicly crunchable.

You can visit DIY's Instagram post https://www.instagram.com/p/BXram2KByYk for more charts of the same data.

[>decentespresso]

Tension can lead to failure

Last week, Ray had a worry that tension from bending a tube might cause the water seal to fail sooner. While none of our connections is leaking now (even under tension), everyone agreed that it would be better to not have uneven tension on the water seal.

To research if any other manufacturer also worried about this, Jeff went diving through our morgue of dead & disassembled espresso machines. He found one Nespresso machine that apparently had this same worry because that machine had little plastic moulded de-tensioners to hold 90º and 180º bends in place.

Alex drew up a SolidWorks model inspired by what Jeff had found, and over the weekend he 3D printed our de-tensioners. There's a photo below, showing where they'll go in. They work nicely, and we'll be using them in all the machines we're building.

This may or may not be an issue, but with these detentioners costing about $0.10 each, it'd be crazy to not use them to make the worry go away.

[>decentespresso]

Insulating the heaters not so easy

A few weeks ago, I mentioned that one of our last engineering problems was to decide how to insulate our two heaters.

Our interns built a test rig, with a tiny Arduino computer, at both 100ºC and 170º, to test all the different ideas we had.

Originally, we liked silicone, but our first batch had a nasty chemical smell when heated. Two other manufacturer's samples also had a smell. Also, it's very hard to get air bubbles into the silicone, so the pure silicone insulating solution adds a lot of weight, and is not as insulating as it could be.

We thought we'd solved that problem when we found a no-smell silicone mix that bubbles when mixed, to a selectable amount between 10% and 30%. The problem is that it's extremely hard to get into a mould and in a regular shape. Immediately upon mixing it swells about 60% but then overnight, as it dries, it continues to swell unpredictably. You end up with a messy "pillow" in the morning.

We tried making a metal mould to contain the expanding foam, and that helped, but it left some areas under-covered as the foam expands in ways that differ each time (presumably because of uneven distribution of the foaming agent).

Ray realized that PC boards are made of fiberglass, an excellent insulator, and they're very cheap to have cut into custom 2D shapes. So, he designed a fiberglass box for our heater made from PCB, which is then soldered together.

It's a bit of an oddball solution, but it worked quite well.

We then tried foaming silicone inside the box, and the thermal insulation results were amazing--the best yet--but it was really hard to do and the results are very irregular and not pretty.

In the spreadsheet below, you'll see the results from all our materials, and you'll see that the air temperature (that's the important metric) is OK for all of them. We want to keep it under 40ºC inside the DE1, and all our solutions have done that. You'll notice that PCB+Foam is the best performing, though, by a small amount.

Today, we're doing our final test, which is placing a pre-cut insulating blanket inside the PCB box, wrapped around the heater. We're expecting that it'll give us a lot of the same insulating benefit of the silicone foam, but be much, much easier and faster to assemble (no pouring into a mould and waiting overnight).

Below are photos of the various attempts and tests we've done, for those interested in following along with our progress.

We're awaiting the parts for our "release candidate #3" which are due in about 7 days. That machine will likely use this PCB Box + Blanket heater insulating technique, which is what we'll use for the 300 machines we're building.

[gc]

Today, we're doing our final test, which is placing a pre-cut insulating blanket inside the PCB box, wrapped around the heater. We're expecting that it'll give us a lot of the same insulating benefit of the silicone foam, but be much, much easier and faster to assemble (no pouring into a mould and waiting overnight).

Below are photos of the various attempts and tests we've done, for those interested in following along with our progress.

We're awaiting the parts for our "release candidate #3" which are due in about 7 days. That machine will likely use this PCB Box + Blanker heater insulating technique, which is what we'll use for the 300 machines we're building.

So do let us know how that PCB Box + blanker(sic) works out since that's what you say I will be getting in the first 300. It is fascinating to watch this process, like watching a football game and tensing up each time you think the player might make, or be about to make, an error.(and then the relief when that doesn't occur!)

[>decentespresso]

So do let us know how that PCB Box + blanker(sic) works out since that's what you say I will be getting in the first 300. It is fascinating to watch this process, like watching a football game and tensing up each time you think the player might make, or be about to make, an error.(and then the relief when that doesn't occur!)

I hope you understand that (to pursue your football analogy) with something this complicated, it's impossible to not make any errors. The hope is for the errors that you do make to be either of negligible importance or repairable later on at low (or no) cost. You can still win the game even if you don't play perfectly.

[gc]

I hope you understand that (to pursue your football analogy) with something this complicated, it's impossible to not make any errors. The hope is for the errors that you do make to be either of negligible importance or repairable later on at low (or no) cost. You can still win the game even if you don't play perfectly.

Yes, I do understand the risks, in principle at least, and for me, good thermal separation within the machine has been something I've focused on and even commented in earlier posts. So I really hope you go the extra mile and get that right. The possibilities for rectification later is one of the things I also pay attention to in the drawings and photos you post. But its been obvious that the development process has become more refined and sophisticated [as it should of course!] and I have much more confidence now than I did near the beginning of your online threads. As always, thank you for sharing the process.

[>decentespresso]

Yes, I do understand the risks, in principle at least, and for me, good thermal separation within the machine has been something I've focused on and even commented in earlier posts. So I really hope you go the extra mile and get that right. The possibilities for rectification later is one of the things I also pay attention to in the drawings and photos you post. But its been obvious that the development process has become more refined and sophisticated [as it should of course!] and I have much more confidence now than I did near the beginning of your online threads. As always, thank you for sharing the process.

Besides (a) insulating the heaters and (b) thermal and moisture separation of electronics from the rest of the machine, the past few weeks have (as you probably suspected) been about thermal separation of various components.

We switched from a plastic group head cover to a metal one a few weeks ago, and we knew ahead of time that this would test our ability to keep heat from the group head from spreading. About two weeks ago we found the chassis was getting warmed by the group head, so we added an isolation barrier at the join. I didn't post anything about that, as I felt it was a fairly minor tweak.

We are reviewing today a small change to the group head, where a thin sheet of isolation has been added between the brass parts (which are preheated, and through which water flows) and the aluminium parts (which are there for strength, but which don't touch the water). The goal is to preheat only the parts that touch water, so as to (a) start up faster (b) not waste electricity (c) not have to worry about what to do with the wasted heat.

Below is a drawing of today's change to the group head. You can also see the two temperature probes in there, which measure (separately) the water mix temperature as it enters the group, and the water temperature on the other side of the shower screen.

[rayuki]

Really love the clean look of these machines but really really hate the name lol. Is it supposed to be pronounced DECAFE?

I mean it makes coffee, why call it that so confusing

[>decentespresso]

Really love the clean look of these machines but really really hate the name lol. Is it supposed to be pronounced DECAFE? I mean it makes coffee, why call it that so confusing

You mean you hate "Decent" or "DE1CAFE" model name?

I suppose it's a matter of taste and sense of humor. We call it the "decaf" here.

[rayuki]

You mean you hate "Decent" or "DE1CAFE" model name?

I suppose it's a matter of taste and sense of humor. We call it the "decaf" here.

Yeah the decaf part lol. I guess it's just annoying me since I hate decaf. I get it now since you call it the DECENT so makes sense. Was just annoying me with the decaf since I hate decaf anything!

[>decentespresso]

Yeah the decaf part lol. I guess it's just annoying me since I hate decaf. I get it now since you call it the DECENT so makes sense. Was just annoying me with the decaf since I hate decaf anything!

I realize "it's not funny if you have to explain it" but "DE1" was meant to sound either like The Matrix "He's De One !" or hip hop "Yo, it's duh one, wassup" as well as the more mundane "Decent Espresso Machine Version One"

So, DE1CAFE is "duh one for cafes, yo" though naturally, all this bad wordplay falls apart when we work on DE2 (decent espresso machine, v2). But, I expect we'll be able to come up equally bad wordplay for that one too.

FWIW, all our machines simply say "Decent Espresso" on the front (like Magimix and Cuisinart do) and the model name & number is on the back.

If you buy a machine from us, it's simply "A Decent Espresso Machine".

[>decentespresso]

Foot pedal espresso

For busy cafes, I've often received emails from baristas worrying that wet fingers won't work well on our tablet. It's true that if you get water on a touch screen tablet, you need to wipe it off for the touch function to work well. Wet fingers, on the other hand, don't seem to be a problem: water drops on the tablet do.

The medical, dental and beauty industries all have foot pedals for dealing with this problem, freeing up your hands to do something else.

I especially like the idea of a foot pedal for steam, because I sometimes destabilize the milk jug as I let go with my left hand to turn the steam off.

You could even put away our Android tablet and make espresso drinks only using the pedals.

I've done a quick look at available foot pedals and here are some features I'm looking for:
- can cope with a wet floor (raised off the floor)
- can cope with water poured directly on the footswitch (IP64 or IP65)
- colour coded so you can easily tell the difference between steam/hot-water/espresso buttons
- comes in 1, 2, and 3 pedal versions, so that people can choose what espresso functions to move to a foot pedal
- the pedal design makes it unlikely to push the wrong one
- rugged

The picture below shows 3 variations of one pedal design that meets those requirements. There is also a compact two-pedal variant from this company that is more attractive but also more confusing to use.

My mine "gripe" with this design (and pretty much all industrial foot pedals) is that the colour scheme is chosen for functionality, not aesthetics. The light blue colour of the "chassis" is a bit "hospital" for me. I'm asking if we can choose a different coloured plastic for our versions.

These foot pedals would plug into the bottom of our DE1CAFE machines, and provide a mechanical alternative to using the tablet. You can keep the tablet around to see the shot quality or put it away completely.

[trentski]

Yeah the decaf part lol. I guess it's just annoying me since I hate decaf. I get it now since you call it the DECENT so makes sense. Was just annoying me with the decaf since I hate decaf anything!

Off topic but, why the hate for decaf?