The one with everything on the left looks neater.
Originally Posted by decentespresso
I'd have thought that the one that gives a smoother workflow would be preferable to neatness.
Originally Posted by StaticBlue
Two new espresso baskets
Today we sent off to our basket manufacturer, two new products:
- After a year of R&D with Scott Rao and 5 iterations, we're finally happy with our "pour over basket" for espresso machines. This converts your espresso machine into a pour-over shower head, with each hole spaced and calibrated so that the entire coffee bed is evenly turbulent. Naturally, this works best of our espresso machine, where you can control flow and pressure, but this basket can also be used on traditional machines, which follow the 58mm standard.
- We have made dozens of these Puck Simulator baskets for our own use over the past 4 years. We use them in testing, burning in machines, and calibration. There's a single hole at 0.3mm, which gives a similar flow rate to espresso, at 9 bar of pressure.
Both these items will be shipping in a few months, but are available for pre-order now. You can find them on our NEW page, off our home page. https://decentespresso.com/new
They said it couldn't be done....
.... and they were right?
The first casts of our stainless steel drip tray cover have come out. The other casting vendors said our design couldn't be done. But one company said their lost-wax casting process could do it.
The upper photo is the unpolished part. You can ignore the finish (it would polish up shiny) but the non-straight wires, are a big problem. They're going to try again, with a cooler metal cast, but they aren't optimistic.
So... the solution here is to reinforce the design to prevent the warpage.
Years ago, we tried a grid of wires on the top, but I didn't like how it looked from above.
Now, I'm suggesting going back to the grid idea, but putting the 90º rotated wires below the main level, so that they'll be much less visible. These intersections will cause coffee to catch and be a bit hard to clean, but I think it'll not be too bad (should rinse off, usually).
Thankfully, our customers have enthusiastically responded to my idea and have sent in about 5 weeks worth of unused drip tray covers. That gives us a bit of time to solve this problem, and still keep shipping machines.
More news on this pressing topic, as the situation evolves.
panorama of various machines
Interesting panorama of various machines, so you can compare:
- the Linea Mini
- the Slayer 1 group
- several E61 machines
- the Decent Espresso DE1PRO
and see how they all look, in comparison to each other.
Thanks to https://www.instagram.com/talkcoffee/ for the video footage.
Other approaches to the drip tray cover
Since our recent setback in stainless steel molding our drip tray cover design, we're trying two other designs, while we try to debug the original approach.
Both new approach to this part retain the main advantage of using rounded wires to avoid water retention.
The main difference here is that we're building up the drip tray cover by welding wire together, and this is is a well known technology.
We're having the prototypes made now, should have them in hand within 2 weeks.
Ben has drawn up a few different alternative drip tray design ideas. The first two could be done with the existing molding technology we are trying, while the last one could be stamped or molded.
John, I wonder if the top right design may retain more liquid than your present and alternative designs? It wasn't till Luca commented on how good your drip tray was I realised the dissatisfaction I had with the one on my machine. Small feature, but one you notice every day and nice to see the effort you are putting into it.
Just chiming in! Regarding the drip trays I really like the design you landed on earlier, so I would strongly prefer the bottom design of the three showed as a good compromise.
If going with the welded option, the Alex Duetto design is not that bad as a starting point IMO. Could probably be niciefied a bit to mach the DE1
I always thought that the welded trays looked too much like a cheap oven tray.
Originally Posted by swevictor
Personal preference I suppose
Whatever gets used, any water retention far out ways aesthetic value. Make it water free, then make that pretty. The zigzag design above looks to be quite functional.
Drip tray bar = a cup centering guide
After debating various ideas for a revised drip tray cover, we came up with the attached design.
This design has a nice (almost invisible) feature:
- the leftmost bar is directly under the center of the group head
- thus giving you an easy guide to follow, to center your cups
- we slightly altered the spacing on the outside, to cause this effect, but I think the spacing change is hard to see as a "defect", and instead looks like a "happy accident" that the bar lines up in this way
- there might still be a slight deformation in the bars after molding, even with 3 bars, but the company feels that their correction tool can repair those.
- the casting company told us that it's hard to get liquid stainless steel to flow into crossbars, and that 3 or for 4 bars was ideal. More bars would be problematic to cast.
So... we're wax casting this revised design now.
I really like this idea. I ended up using a felt tipped pen to mark a centre for my cups!
Originally Posted by decentespresso
My vote and dollars will also go with the first person to make a Bluetooth. 01gm scale that is no more than 1cm high and water resistant/proof.
My plan is to go even further with thin-ness, making the scale disappear entirely by putting the load cells under the drip tray corners themselves.
Originally Posted by gc
Ha! Ninja'd me.
Decent Coffee Cart with Wood Effect Paneling
This is IKEA's larger BROR table ($150) at 110cm x 55cm (with the $99 BROR wheels on it), with one white DE1XL on it. I photographed a section of the wood I'd varnished, and had foam panels printed up of the wood top, but for use as side panels. The effect is quite nice, and makes for a good coffee cart for the home. Two DE1XL espresso machines would fit on this table, if I wanted to make this a high-volume setup.
Originally Posted by decentespresso
Like all the best ideas, it is so simple when someone actually thinks of it. I will happily test it for you.
12th September 2019
Casting a wide net
While we wait for our new design for a cast drip tray covers to be prepared, we're also asking for samples of alternative approaches.
Prototypes of 3 alternatives arrived in my inbox today.
Uhm.... not quite acceptable, I'd say.
The search continues!
12th September 2019
I dunno, if you made it bigger it might be OK as a farm gate.
Originally Posted by decentespresso
12th September 2019
It makes a great little hibachi grill for when you go camping.
Originally Posted by Lyrebird
13th September 2019
If you replace the bottom rod with thin bar you may be able drill holes at either end to fit the carrier rods.
Originally Posted by decentespresso
May not be cost effective but would look better.
13th September 2019
First look: group head shows pressure and flow in real time
This will be available on Decent machines starting in January 2020, and will also be available as a low cost upgrade for all Decent machines built after January 2019.
This is our first working attempt at this feature. There are still more improvements to come, which I'll be posting here as they happen.
13th September 2019
Coffee Cart Underbelly
People seem to doubt that 5L water bottles fit in this small 80x55cm coffee cart. In this video you can see both 5L bottles, and an 8L Volvic water bottle as well. And there's a layer for coffee mug storage.
The Joys of Thermal Cycling
Espresso machines turning on and off makes them heat up, and cool down. This "thermal cycling" causes metal to expand and contract. On traditional boiler based machines, this often causes welds to crack with time.
There is a long history of discussion of this topic, and many many manufacturers recommend leaving their boiler based espresso machines on 24h a day:
With the Decent Espresso machine, we use a clip-connection system which can withstand thermal cycling, as the seal is made by a rubber o-ring under pressure, instead of a weld. As the machine cools, the seal loosens and allows movement.
However, there are some parts of the Decent that are susceptible to thermal cycling. This is especially a problem where different materials are in close contact, because each material will expand/contract differently to temperature.
A few customers, in the past 2 years, have had the temperature bead "squeezed out" of our steam waters. This is caused by thermal cycling causing the epoxy that holds the temperature bead, to slowly push outward, and thus upward.
We spoke to our temperature probe manufacturer about this, and showed them customer photos. It turns out that they had already spotted this potential flaw, and fixed it in the design that we started to use in January 2019. That explains why only Decent v1.0 customers had this problem.
They've now put a deep channel in, and threaded it, so that the epoxy sticks to the threads, and cannot push itself up as it expands/contracts.
We were pleasantly surprised today, to discover this. As the change was inside the epoxied part, we never knew it. We had in fact sawn temperature probes in half, to discover the source of this problem, and were going to propose much the same modification!
Naturally, this problem falls in the realm of "manufacturer's defect" and we'll replace this part and fix the machine--for life and for free, any machines that exhibit this problem.
We're seeing more of our machines in commercial, high volume use these days. This means that the ways the Decent might break are coming to light, and we're able to address each one as it appears. I think that in two years we'll be really rock solid, but in the meantime we'll work closely with anyone who does have a problem, and address it for free and hopefully quickly.
You should also ask them for the glass transition temperature* (Tg) of the epoxy used and look at the effect of changing to one with a higher Tg.
* Epoxy goes through transition from a glassy state to a rubbery state with temperature, accompanied by a change in coefficient of thermal expansion. The epoxy I use in my bikes achieves a TG of 130 oC with an 80 oC cure; higher Tg epoxies exist but they require more agressive cure schedules.
Last edited by Lyrebird; 4 Weeks Ago at 10:13 AM.
Is it OK to encase the end of the thermocouple in epoxy like that? I would have thought it would affect the readings. Since it is not in direct contact with the water, it might be better to have the inside lower half of that fitting full of thermal transfer compound for the thermocouple to sit in, and then have a sort of bung or cap only made of epoxy to keep the probe in situ.
You can get thermally conductive epoxies, which I imagine, is what is being used...
A Decent Manual
We’ve organized all the accumulated documentation pages (and videos) into a Decent Manual.
We’ll continue to increase the size of scope of the manual, both with text and videos, as this is a “living” set of documents.
A key difference between a normal manual and the Decent Manual, is that each exists on a web page where you can ask questions and discuss, with other users as well as the page’s author.
This is the main reason why the manual is currently only available to owners.
I’ve always been impressed by the PHP Manual, ie:
which breaks each function into a page, gives a concise official manual page, and then evolves over time in the conversation below.
I’m trying to do the same with Decent, though I also create a lot of videos, as I find them often clearer than any text could be.
Attached is an example of a page with a video and a short conversation.
Decent Scale News
Cleared my schedule for a few days to punch through on the Decent Scale. Sorry for the delay, guys.
I got my bluetooth implementation somewhat working (TARE/LED off/on, weight) and then totally rewrote the weighing code in the tablet app to support a different kind of scale. Now, my tablet app is very broken, and I have a few days to get it working again.
However, I now have placed two new feature requests with the firmware programmer we've hired:
- flow rate measuring built into the scale : this would be wonderful, because it won't depend on bluetooth messages arriving reliably at 10x/second. It would be much more accurate to have the scale calculate flow and weight, and send both via BLE. No bluetooth scale currently does this, I believe.
- an ultra-low-power mode, with the scale off, except for BLE. This would address the Atomax Skale issue people have raised, where even with LEDs off, the Atomax Skale batteries go dead in a few days. When the DE1 were asleep, the Decent Scale would also be really, really asleep.
Separately, I've also identified that bluetooth control over the Decent Scale's timer, isn't working, so that's for him to fix.
We're also going to paint on a O and  symbol on the plastic cover, as it's too hard to see at the moment. That'll also give a bit more tactile feedback.
I know theres a lot more going on here but im stuck on that milk jug washer traydoodat
I need that!!
Great work on everything else too!
That pitcher rinser has changed my life, because now I can clean everything up in one place, and my girlfriend doesn't fret about my clogging the plumbing up with coffee grounds.
Originally Posted by Sullo
After I knock the puck out out of the portafilter, I push the portafilter onto the pitcher rinser for a second, and then give it a quick dry with a brown cloth.
I also use my pitcher rinser to give me a bit of water on my brown cloth to wipe down the steam wand after milk steaming.
Getting the fixture right from the flowjet to the pitcher rinser has been a bit tricky. I currently have a ugly hack that connects two long parts together, with a worrisome bend at the connection point. Plus it uses a lot of height and appears to be easily breakable:
Screen Shot 2019-09-22 at 11.58.46 AM.jpg
However, on monday I have a package coming from Decent HK, with a single part that (theoretically) does the job. It's in metric, not imperial measurements, so the supplier "cannot ensure" but my HK guys tell me that it does work. We'll see.
If it does work, I'll include it for free with the flowjet we sell that matches our pitcher rinser.
Do shower screen spacers cause better espresso?
John Weiss' extensive review of the DE1+ in Home Barista was quite positive, but did call out a difference in style between his classic La Spaziale machine, and the Decent:
"DE1 shots feature cleaner taste, greater clarity, and improved flavor separation, whereas body/mouthfeel is noticeably enhanced on the Spaziale. With lighter roasts, where a long preinfusion is practically essential, the DE1 produced obviously superior extractions."
Since our espresso machines are "open for tinkering" and the (almost 1000 now) Decent customer base has a lot of tinkerers, for the past 2 years there has been a furious amount of theorizing and experimentation, to figure out *just why* would mouth-feel be different?
Yes, many people like the "clarity" of Decent espresso shots, but others want to recreate the classic thick mouthfeel. It *should* be possible with the Decent, if one could isolate the important variables.
Damian was one of the first to create a variety of different spacers which pushed the shower screen down only the coffee puck. He experienced greater shot consistency, increased drink quality, and greater mouthfeel.
From there, Sheldon machine his own spacers from stainless steel, and experienced similar results.
Now, Sheldon has produced his own run of stainless steel spacers for v1.0 and v1.1 Decent customers, and is selling them at US$100 at whatever thickness the person wants (they’re made to order).
The goal here is to determine if thicker mouth-feel is directly causable by having a shower screen pressed against the coffee puck. Initial results indicate "yes", and now the community is moving to a much larger test, with dozens of people participating.
We want to learn if the spacer change is "better" in "all cases" or just "in some cases". Is there a clarity vs mouth-feel tradeoff?
Over the next few weeks, we're going to learn a lot...
Last edited by decentespresso; 3 Weeks Ago at 05:03 PM.
A different kind of relationship
Dimitri, the owner of Oven Heaven https://www.instagram.com/ovenheavencoffeeroasters/ cafe/bakery in Bordeaux, France, sent me a package.
I had screwed up, and accidentally sent him an extra bluetooth scale. This had happened because we ran out of stock just as he ordered. He had to wait for his scale, and then I messed up the invoicing and accidentally sent him a 2nd one.
So, Dimitri then offers to send it back to me, and to include some free coffee beans. At his cost.
That's not a normal commercial relationship. How does this sort of thing happen?
When Dimitri received his white DE1XL for his café, he was a little but disappointed that we didn't offer a "natural wood" version of our double-spouted portafilter. We sell very few of these (most people use the bottomless) and I have a ton of stock of the black double-spouted portafilter, so I hadn't planned to order 1000 natural wood double-spouted portafilters, and have them sit around my warehouse for a few years.
Nonetheless, I thought he was right to be disappointed. We'd missed an important aesthetic detail.
So, I tried to sort something out.
I contacted the wood turner that had made the handles, to see if they could make a short run for us. It turned out that they'd made 10% extra, 6 months ago, assuming a 10% defect rate, and they had 32 handles "just sitting there", waiting for our next order. I offered to buy those 32 handles, and sent one (for free) with a double spouted head, to Dimitri.
I'm writing the tale above, just to give you an idea of what our daily customer interaction is like.
Because of how I and the rest fo the Decent staff interact with people, we find that we can call on their help. They want us to succeed.
That's why, when I recently emailed 230 customers asking if (at their cost) they'd mail in their unused drip tray covers to help keep our factory running, 96 of them actually did. That's a 41.7% response rate. Amazing.
And I think this is also why dozens of customers are now working together to understand the relationship between basket head space, and espresso mouthfeel. We're in this together.
How many companies do you know, have such great customers?
Coffee cart DIY kit
My Decent team is working on converting my hand-made IKEA BROR cart into a "coffee cart do-it-yourself" kit, which we'd sell. It'd only require a screwdriver to assemble, and would take about 2 hours of your time.
Here's the idea.
You would buy (on your own):
- an IKEA BROR 85x55cm table
- 3 water bottles (5 liters, or 10 gallons)
- Niche grinder (or other)
You would buy from us:
- pre-cut and varnished IKEA BROR table top
- Rinser/flow jet matched pair
- Other coffee accessories (optional)
You'll only need a screwdriver to follow the videos I'm making, and I estimate that your total build time will be 2 hours. We're working with a local CNC shop to pre-cut and varnish the table top for you.
We're working on different two cart models:
- 1 DE1XL model: uses 1 IKEA BROR 85x55cm table
- 2 DE1XL model: uses 1 IKEA BROR 110x55cm table
I will provide the Photoshop files for you to order locally, your own printed foam boards to wrap around your cart. It costs me $60 to print locally.
Coffee Carts for Rent?
Starting in March, we're doubling the size of our factory. We will then have more espresso machines being produced per month, and can start to contemplate other ways to them out to people.
We're thinking of building complete coffee carts ourselves having them available for rent on a daily/monthly basis in a few cities where Decent has its own people. For now, that'd be New York, Seoul, and Hong Kong. Next, possibly San Francisco, Seattle, Montreal. We'll start slow, and build out as we see what demand there is, and how it goes for people.
My goal is to provide a no-financial-risk coffee cart for people. We take on all the responsibility of making, maintaining, repairing and insuring they're fit-for-service.
A different kind of relationship
... and that is exactly why I am so happy with the "Decent interaction part" of my DE1 V1.1 (apart from the great coffee it purveys).
You guys have provided such great service. You have even sent me (gratis) a couple of spare parts that you discovered had given other users a problem - even though my parts were still working well. How rare is that in these "profit at all costs" times?
I have always supported companies that do fine research and then back it with great service. That is why I am still trying to decide whether to upgrade to a DE "shoot and froth at the same time" model or "merely" to update my grouphead when V1.3 comes out. If your service was not so great I would just minimise my cost and update the group.
I look forward to the result of your mouthfeel testing. FWIW, my Electra "eagle" manual lever probably had slightly better mouthfeel than the DE1. However, the taste of the DE1's coffee is so much better (now that I have mostly tamed the beastie) that I was not too worried about it. As the only "neat" espresso drinker in this household these days I am slowly converting more to lattes so it is a moot point here.
Oh, and great work on the cart - it is looking really promising.
PS: John, my brother is now 3D printing 20% carbon fibre / 80% nylon on a fairly inexpensive rig. If you are not already playing with one, it is worth checking out. Magic for making the odd "one off" part in a hostile environment.
Comparing USB cables
We're really in the nitty gritty details here! We're trying to make the "perfect" USB cable for our espresso machine. This latest version uses a silicone rubber insulator, so that it's more flexible, than its stiffer predecessor. We're trying to avoid a small problem with the previous stiff cable, where it pushed the tablet slightly, as the tablet sat on the machine. Details details!
It's also 1cm longer, for a bit more tolerance for those people who position their tablet at a slightly location than I do.
I just received this cable, and so far, it looks like it's "the one". In a few days I'll pull the trigger and by 3 kilometers of this wire material (that's the "minimum order quantity" !) as the sample I have has a slightly thinner metal wire than its predecessor. We preferred the thicker wire as it had the highest conductivity tests of all USB cables we tried.
ShinTV: going to the SSP burr factory
In this episode, Shin visits Hansung Lee, the CEO/engineer/mastermind behind SSP burrs, gets a tour of the factory, the burrs, and sees how Hansung uses his DE1 to test burrs and grinders out.
These burrs have emerged in the past two years as "one of the top 3 in the world" and many feel they are simply the best.
Interestingly, Hansung's only complaint (the USB cable) happens to be the subject of my previous post today.
Decent Owner Reed Taylor is working on DAYBREAK, an accessory for Decent Espresso Machines.
DAYBREAK plugs into our standard communications port, replacing the bluetooth module that's placed there by us. The bluetooth module then plugs into his board.
Here is what DAYBREAK does for you:
- it is a full featured Linux computer, that can fully control the Decent Espresso Machine
- It can run the tablet app I've written, and present it as a full-featured web app. This allows you to run what is now the "Android tablet app" on any computer (iPad, Windows, Mac) from your web browser.
- There is a USB port on Daybreak, so you can plug your own computer in. A number of robotics projects are underway with the Decent, and wired-control is much preferred over Bluetooth for reliability.
- My tablet app becomes fully functional on Mac, Windows and Linux, thanks to the USB connection. Previously, it was mostly functional, but lacking the Bluetooth functionality to actually make coffee.
- The USB interface is *much* *much* faster than bluetooth, and can access all the internal sensors and data, at much more useful speeds than bluetooth could.
- Reed's plan is for his DAYBREAK board to be something you can simply buy and plug into your espresso machine. You won't have to be a tech genius to use it.
- A tablet app update release I made a few weeks ago, solved the outstanding bugs with this approach.
Reed still has problems to work through before it's ready, but his progress has been steady and impressive.-john
Could you plug the supplied tablet in and get better response times that way?
Yes, but you'd want to buy a tablet that can charge via USB at the same time that it's connected to a USB device. They're not that common.
Originally Posted by level3ninja
The way to do this is to get a tablet that is "On the go" (OTG) compatible, and a special cable:
Note that response time is not bad at all, at the moment, on our Android app, because the chart refreshes at 10x per second, the max that Bluetooth can do. You'd want to "do more" with the app, to take advantage of the increased bandwidth. Might be useful, for instance, to visualize very fast puck channels better.
Bigger BROR IKEA Coffee Cart Progress Report
I've wired and plumbed everything into my Bigger Bror cart. Here's a quick tour.
Here is a link to a shopping cart with everything that is on my coffee cart:
except for these parts, which you will buy elsewhere:
iPad & Mac demo - running the Decent Espresso App
In this video, you can see the Decent tablet app running as a "web browser app". This allows you to run the app from any device that has a modern web browser.
In this demo, you'll see me use the app from my Mac, and my iPad.
- A Z button floats on screen to zoom the app in your web browser. This works both on my iPad and Chrome on my Mac. You can choose to go completely full screen, or just full-inside-the-web-browser, by repeatedly pressing Z.
- A K button floats on the screen of the iPad and Android tablet (under web browser use) to show the on-screen keyboard.
Still to do:
- you need to run the tablet app on a computer (windows/mac/linux), which then broadcasts the app through a built in web page. So... you need a computer around.
- at the moment, running the app on a computer means it doesn't use bluetooth.... which means it can't yet actually control the espresso machine. That's getting addressed in two ways: future bluetooth support in the app on windows/mac/linux, and USB support via Reed Taylor's DESIRE accessory
- I'm fairly sure that by using Jasonelle https://jasonelle.com I will be able to make a native app on iOS that automatically launches the Decent app on your iPad. See https://www.freecodecamp.org/news/ho...-631c9c9895f5/ - that will be a little bit more convenient than loading a web page in Safari.
- Typing on iPad does not work right. Typing does work from a computer or from Chrome on Android, but not on iOS. This is the only major bug I observed.
- A minor issue is that the web app is a bit dimmer than when run natively. I'm working on both those issues, but neither is a deal-breaker.
John just saw a Starbucks getting smashed up in the riots on TV. Hope you and your staff are ok and not being inconvenienced. Though I am sure you would be taking every care with their safety and your own.
They just need to make better coffee.
Originally Posted by 338
It's a bit more excitement than we'd all thought we'd signed up for, being in HK, and on Friday I let the staff leave early as they (rightfully) feared the subway closing down.
Originally Posted by 338
We're located in a family-friendly community named Tsuen Wan, lots of kids and schools. And... quite a bit of protests here too.
DE1XL Suitcase Prototype
I just received my DE1XL in some hand-cut foam to see if it could travel well in our suitcase. My goal is to make the slightly-larger and more capable DE1XL model as transportable as our DE1PRO and DE1+ models.
Here, we're also trying new protective corners, made out of hard plastic.
And a new, more opaque back panel for the white model, so that the magnets don't show through.
If we can pull this off, we'll be able to use the same suitcase, but with different foam, for all our espresso machine models. And I'll be happy to send (for free) the updated foam to those of you who already bought our DE1XL model. FYI, the DE1XL (launched in March 2019) now represents 18% of our machines sold, almost 1 in 5.
Testing at the source
One of the things we've discovered, as we've learned "how manufacturing works" is that if you've ordered something custom-made, once it's in your hands, it's your problem if it doesn't work. All contracts stipulate that responsibility ends as soon as the item gets loaded on a truck.
For the past two years, we've had our PC boards made by a company in Shenzhen, where almost all mobile phones in the world are made. We've experienced defect rates between 15% (the AC board) to 30% (the DC board, aka the "computer"). At $50 each (approximate), that's (150pcs x $50) + (300pcs x $50 = $7,500 + $15,000 = $22,500. Ouch. This high defect rate forces us to over-order each part, so that after quality-checking, we have enough parts to make our goal quantity of espresso machines.
Part of the cause of the high defect rate, is that our quantities are considered small. Anything under 20,000 pieces is small in Shenzhen. We're ordering 300 to 750 pieces, so we're "real small". These small order quantities mean that it's easier to do a lot of the work by hand, instead of by robot. And by hand, means lots of variation, and a high error rate.
For our upcoming v1.3 espresso machines we've changed to a Hong Kong based company called PDSTE, who is based in "Science Park", an incubator for tiny high tech companies. For PDSTE, we're a huuuuge client, which is a nice change! And... everything they do is 100% by robot, because in Hong Kong, labor is not cheap: about the same as in the USA or EU. It's either Robotic or it's not Made in Hong Kong (well, except for our Decent Espresso Machines).
So... as part of moving to a local supplier, we've built a "testing computer" that is essentially everything the espresso machine is, but as a single board computer that does nothing but run the tests.
We're building a few dozen of these test boards, so that PDSTE can test our "espresso computer" before they deliver it to us. Crucially, any board that fails at PDSTE, before it is delivered to us, is *their* problem, not ours. So, it's in our interests to make this test as thorough as possible.
For each board to test, all the sensors, pumps, valves and more are connected to the board that needs testing, and then the test board connects to that. Then.... all the lights go GREEN or they don't. If they don't, they can't be shipped to us.
This also highly motivates the manufacturer to reduce their defect rate, as they feel the pain of each defective board, to the tune of about $35 in parts that they just lost, not to mention their labor cost.
When we started down this path almost 5 years ago, we knew nothing. We're slowly learning.