Page 13 of 13 FirstFirst ... 3 11 12 13
Results 601 to 649 of 649
Like Tree632Likes

Thread: Decent Espresso Machines (DE1) - Any thoughts?

  1. #601
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Spy the icon

    I'm working on the app icons and default wallpaper for the tablet on our Decent Espresso machine. You'll see this page when you turn the tablet on, swipe to unlock, and then run our app.

    By default, there's only one app on the page, to make it simple. You can tap the (...) icon on the middle bottom of the tablet to see all the other apps, which are the default Google-licensed (ie, clean!) apps.

    I'm thinking that it'd be fun for each "manufacturing run" of our espresso machines to have their own default wallpaper.

    For the first 300 machines, these customers are "in on a secret" and so I thought I'd use our "spy theme", which shows secret dossiers, books about spying, as well as an espresso, glass of water and steamed milk drink, on the desk.

    Of course, you can change it if you don't like my choice, and this Android installation is unlocked so you can install your own apps as well.

    I originally thought I'd auto-run the app on startup, but in practical use, I found this annoying, because I had more than one espresso app (DE1 vs DE1+) and if an open source ecology builds up around our machines, I'd like my app to not have an unfair (ie, auto-run) advantage over alternatives.

    Is this wallpaper too artsy, or do you like this idea?

    spy_wallpaper.jpg IMG_6979.jpg
    dilbert_user_friendly_computer copy.jpg wallpaper2.png

  2. #602
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Working with factories: PASS/FAIL check lists.

    Over the past two years, I've learned how to work better with factories, and one thing has emerged: photos, clear pass/fail criteria, and checklists are the best way to improve quality.

    The factories also really like these, because they are very clear, both sides agree on what to be delivered. Typically, a "quality control" person is assigned to our order, and they do the checklist for every item.

    While some English is needed, it's best if photos are used. Generally, only the management and salespeople speak English.

    As our knockbox is a fairly simple product, and this is our first manufacturing run, I haven't seen that many flaws, thus my QC checklist is only 4 pages long.

    Each manufacturing run, there is inevitably some new way discovered that the product can be badly produced, and this is then added to the QC checklist for next time.

    Our milk jugs are incredibly complicated to make, and the QC checklist runs to 24 pages. This is also because we've done so many manufacturing runs of it, and so we've seen so many ways to screw the job up.

    In case you're curious, this is what the QC checklist for the knockbox now looks like.

    When items are delivered to us, we re-check each one and separate them into "full price", 30% and 50% off, with those discounted items ending up on our Discount Sale at Decent Espresso page. I don't want to throw away almost-good-enough items, and some people are less fussy and happy to get a better price.
    QC Checklist for Decent Espresso Knockbox2.jpg

    Dimal, gc and RavenMad like this.

  3. #603
    Junior Member
    Join Date
    Feb 2017
    Posts
    22
    Quote Originally Posted by decentespresso View Post
    I'm working on the app icons and default wallpaper for the tablet on our Decent Espresso machine. You'll see this page when you turn the tablet on, swipe to unlock, and then run our app.

    By default, there's only one app on the page, to make it simple. You can tap the (...) icon on the middle bottom of the tablet to see all the other apps, which are the default Google-licensed (ie, clean!) apps.

    I'm thinking that it'd be fun for each "manufacturing run" of our espresso machines to have their own default wallpaper.

    For the first 300 machines, these customers are "in on a secret" and so I thought I'd use our "spy theme", which shows secret dossiers, books about spying, as well as an espresso, glass of water and steamed milk drink, on the desk.

    Of course, you can change it if you don't like my choice, and this Android installation is unlocked so you can install your own apps as well.

    I originally thought I'd auto-run the app on startup, but in practical use, I found this annoying, because I had more than one espresso app (DE1 vs DE1+) and if an open source ecology builds up around our machines, I'd like my app to not have an unfair (ie, auto-run) advantage over alternatives.

    Is this wallpaper too artsy, or do you like this idea?

    spy_wallpaper.jpg IMG_6979.jpg
    dilbert_user_friendly_computer copy.jpg wallpaper2.png
    Nothing wrong with artsy. It shows thinking outside the square. I like the first WP and will have that in a flash.

  4. #604
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    Quote Originally Posted by decentespresso View Post
    I'm working on the app icons and default wallpaper for the tablet on our Decent Espresso machine. You'll see this page when you turn the tablet on, swipe to unlock, and then run our app.

    By default, there's only one app on the page, to make it simple. You can tap the (...) icon on the middle bottom of the tablet to see all the other apps, which are the default Google-licensed (ie, clean!) apps.

    I'm thinking that it'd be fun for each "manufacturing run" of our espresso machines to have their own default wallpaper.

    For the first 300 machines, these customers are "in on a secret" and so I thought I'd use our "spy theme", which shows secret dossiers, books about spying, as well as an espresso, glass of water and steamed milk drink, on the desk.

    Of course, you can change it if you don't like my choice, and this Android installation is unlocked so you can install your own apps as well.

    I originally thought I'd auto-run the app on startup, but in practical use, I found this annoying, because I had more than one espresso app (DE1 vs DE1+) and if an open source ecology builds up around our machines, I'd like my app to not have an unfair (ie, auto-run) advantage over alternatives.

    Is this wallpaper too artsy, or do you like this idea?
    While appreciating your personal touch with the spy theme, I'd find it a bit busy for me [I choose black on my computers and tablets for visibility! ], but as you say, its easy to change. Because the tablet has a fairly specific use, maybe consider the [very minimal] pico Gapps [eg The Open GApps Project ] preinstalled in the ROM to minimize the Google apps to distract from your machine. If people want to multitask your tablet and want gmail, hangouts and media apps, they can still download them from the PlayStore.

    Autorun as a configurable option would still be good, so if another open source option were preferred it could be disabled with a switch/radio button in the app settings - or the other way round. Also, for non-techie users a really intuitive and accessible [like an icon from the home page or all apps screen] option to change the wallpaper, with some easily accessible choices.

  5. #605
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by gc View Post
    While appreciating your personal touch with the spy theme, I'd find it a bit busy for me [I choose black on my computers and tablets for visibility! ], but as you say, its easy to change. Because the tablet has a fairly specific use, maybe consider the [very minimal] pico Gapps [eg The Open GApps Project ] preinstalled in the ROM to minimize the Google apps to distract from your machine. If people want to multitask your tablet and want gmail, hangouts and media apps, they can still download them from the PlayStore.Autorun as a configurable option would still be good, so if another open source option were preferred it could be disabled with a switch/radio button in the app settings - or the other way round. Also, for non-techie users a really intuitive and accessible [like an icon from the home page or all apps screen] option to change the wallpaper, with some easily accessible choices.
    While Gapps is cool and I like your advice on many levels, the direction you're going in is different from where I want to go with the tablet and the espresso machine.

    My design goals are:
    1) we're in the espresso machine business, not the android tablet business. Don't put a lot of R&D into making a highly customized tablet, put that time into making a better espresso machine and espresso making app.
    2) a general purpose tablet with google apps preinstalled is useful. Don't strip everything out, instead let the tablet be useful if people want to use it for other things
    3) don't rewrite things that are already done for you (ie, wifi, wallpapers, app download and updating). Assume we can't do a better job than Google has.
    4) don't make an unusual tablet, make it very, very boring. No crudware, no malware, but ready to use.

    FYI I plan on eventually using Chrome on the browser and the built in web-standards-compliant video calls it can do, to provide video tech support, and also use the tablet to access to our eventual cloud services.

  6. #606
    Senior Member
    Join Date
    Aug 2013
    Location
    Rockingham W.A.
    Posts
    946
    Quote Originally Posted by decentespresso View Post
    While Gapps is cool and I like your advice on many levels, the direction you're going in is different from where I want to go with the tablet and the espresso machine.

    My design goals are:
    1) we're in the espresso machine business, not the android tablet business. Don't put a lot of R&D into making a highly customized tablet, put that time into making a better espresso machine and espresso making app.
    2) a general purpose tablet with google apps preinstalled is useful. Don't strip everything out, instead let the tablet be useful if people want to use it for other things
    3) don't rewrite things that are already done for you (ie, wifi, wallpapers, app download and updating). Assume we can't do a better job than Google has.
    4) don't make an unusual tablet, make it very, very boring. No crudware, no malware, but ready to use.

    FYI I plan on eventually using Chrome on the browser and the built in web-standards-compliant video calls it can do, to provide video tech support, and also use the tablet to access to our eventual cloud services.
    I fully agree with your points 1, 3 & 4 above. As for your second point, here is another point of view on this subject.

    When my DE1+ arrives the first thing I will do is strip out virtually all google apps (just like my Android phone). As an "old fart IT guy", I only install exactly what I need on all my own machines with no extras. Most IT guys of my vintage happen to regard a lot of google apps as crudware, spyware (indisputable) and arguably verging on malware. Despite the ability to share useful shot data, which I fully support, if it has too much other rubbish I will also (reluctantly) block all access to the internet. Reminds me of my friend's Samsung TV - he went from boasting how good connecting it to the 'net was to being rather sheepish when it emerged that the thing also had a microphone which could be controlled remotely... 1984 has finally arrived in his case.

    Perhaps it would be better to ship it fairly bare in terms of other apps and just provide a link for those who cannot help but check facebook whilst preparing a shot?

    My opinion: This is an espresso machine which happens to come with a tablet attached to control it. I just don't see it as a tablet that can do your emails, remotely control your house and do your dishes while making a coffee.

    TampIt

    "Inappropriate use of technology has finally reached the point that if you want to stop a draft coming in from under your door you buy a robot with a thick foot". Isaac Asimov (possibly inadvertently paraphrased from an interview I heard a long time ago).
    Last edited by TampIt; 3 Weeks Ago at 03:27 AM.

  7. #607
    Senior Member
    Join Date
    May 2005
    Location
    Perth
    Posts
    961
    Make sure you have antivirus on it. I'd also recommend a subscription to Office to do word processing and maintain spreadsheets...

  8. #608
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Inside the DE1v3 manufacturing prototype

    We just received the parts to build version 3 of our manufacturing prototype. We'll show you many of the parts that go into the machine and some of the things that have been improved.

    In a few days we're receiving a slightly modified group head design (we're calling that machine version 3b), to further isolate the pre-heated group head from the rest of the machine, and then hopefully we're done with R&D!

    But first, we have to build the 110V and 220V versions of this machine and make sure we're happy with all the little changes from a month ago when we built our manufacturing candidate v2.

    Dimal likes this.

  9. #609
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Coffee Jewelry

    About a year ago, Joo and played around with some "barista pin" ideas, jealous as I was that sommeliers have a wide variety of grape vine inspired pins, and looking to have an unusual "free gift" for our clients.

    We gave up on the effort, as everything we drew was pretty lame.


    A few days ago, a pin manufacturer contacted me with a variety of designs, and I wrote back "sure, ok, send me some samples" as I liked the look of the 3-bean pin. I did howl with laughter over the "drip coffee maker" pin. Well done, drip coffee maker dude!

    I received the pins yesterday and decided to "model" it on some of my clothes, to see "if it works". I'm not so sure it does.

    The problem, I think, is that sommeliers dress up a lot fancier than baristas do. A "jewelry pin" looks nicer on a fancy dinner jacket than on a collared shirt, but what barista would dress in a nice jacket to make coffee?

    I know this is a really trite topic, but it's an amusing diversion from my daily being-very-practical way of being.

    What do you think?

    IMG_7034.jpg IMG_7048.jpg IMG_7049.jpg

    RavenMad likes this.

  10. #610
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    Quote Originally Posted by TampIt View Post
    I fully agree with your points 1, 3 & 4 above. As for your second point, here is another point of view on this subject.

    When my DE1+ arrives the first thing I will do is strip out virtually all google apps (just like my Android phone). As an "old fart IT guy", I only install exactly what I need on all my own machines with no extras. Most IT guys of my vintage happen to regard a lot of google apps as crudware, spyware (indisputable) and arguably verging on malware. Despite the ability to share useful shot data, which I fully support, if it has too much other rubbish I will also (reluctantly) block all access to the internet. Reminds me of my friend's Samsung TV - he went from boasting how good connecting it to the 'net was to being rather sheepish when it emerged that the thing also had a microphone which could be controlled remotely... 1984 has finally arrived in his case.

    Perhaps it would be better to ship it fairly bare in terms of other apps and just provide a link for those who cannot help but check facebook whilst preparing a shot?

    My opinion: This is an espresso machine which happens to come with a tablet attached to control it. I just don't see it as a tablet that can do your emails, remotely control your house and do your dishes while making a coffee.

    TampIt.
    I was beginning to feel alone in my thinking when I suggested a similar view above. Thought John missed my point completely, but later thought maybe he's aiming at a younger generation who might think its really "cool" to stand at the coffee machine while its heating up and do their email! The excuse was that it was too complicated/distraction to have their tablet with minimal Google stuff, but the truth is that the manufacturer will flash just about any image they require - especially if they compile it from source with their own specs (pretty easy for any programmer to do - I'm a non programmer and have done it on a Linux machine). So I'm with you Tampit about a minimal device but with easy update options which was the suggestion I was making - afterall, its a coffee machine that happens to have a an Android controller. And an option to alter the default startup status of the decentespresso app is simple for any app developer to implement. So I can only assume they are not really interested in the underlying principles of function and purpose we are espousing. Same with the minimal PS chosen.

    I'll probably strip out all the junk apps too. But its good to see the continuous refinement of the production versions as its all coming together.

  11. #611
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Describe that espresso shot

    I'm working on the design for a "DESCRIBE" button that appears on the DE1+ after you make an espresso.

    The goal is to replace the paper notebook that some (very organized) people keep of each espresso they make, in their efforts to improve the quality of their drinks.

    The DE1+ already stores everything about every espresso, 10x per second (pressure, flow, temperature, and all settings) so that you can recreate it, but in addition, I'd like to give you the option to enter in the most relevant data points.

    I want to keep this as short as possible but not miss anything vital.

    My current list of fields for each shot is:
    - type of grinder
    - grinder setting
    - coffee dose weight
    - brand of coffee bean
    - type of beans
    - roast date
    - other notes
    - TDS? EY?

    I could also include sliders for 1-10 ratings for:
    - Quality 1->10
    - Acidity 1->10
    - etc...

    but I haven't give that much thought yet.

    I know there are taste grading sheets out there, and I'd LOVE to hear what you guys think works well. For example, the SCAA Q grader cupping form (picture below)

    Also, I will shortly be announcing a partnership with SCENT ONE, a new company that has an "Aroma Standard" with vials for 100 flavors. The DE1+ will have a GUI for quickly describing the characteristics of an espresso (ie, vanilla, grassy, anise, etc...)
    scaa coffee cupping form copy.pngScentone.jpg

    MrJack likes this.

  12. #612
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    I was just looking at my hand written notes and the only additional items I have is the date poured and an estimate of tamping pressure -eg light/medium/hard. My only other query is will the shot data in graph or tabular form be available beside/above/below the shot notes? I think that would be helpful too.

  13. #613
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by gc View Post
    I was just looking at my hand written notes and the only additional items I have is the date poured and an estimate of tamping pressure -eg light/medium/hard. My only other query is will the shot data in graph or tabular form be available beside/above/below the shot notes? I think that would be helpful too.
    My plan is to export (automatically) all the shot data to the "Decent Cloud" and there offer various ways to slice and dice the data, as well as offering both downloads and API access, so that others can do "magic things" with it. A web browser (and a computer) is a better place for data analysis than an 8" touch tablet.

  14. #614
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    Quote Originally Posted by decentespresso View Post
    My plan is to export (automatically) all the shot data to the "Decent Cloud" and there offer various ways to slice and dice the data, as well as offering both downloads and API access, so that others can do "magic things" with it. A web browser (and a computer) is a better place for data analysis than an 8" touch tablet.
    I wasn't thinking serious data analysis, I agree a bigger screen is better for that - just an easy to see summary of the shot and notes on the same screen as you pull shots, or even choose what coffee for the morning java.

  15. #615
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by gc View Post
    I wasn't thinking serious data analysis, I agree a bigger screen is better for that - just an easy to see summary of the shot and notes on the same screen as you pull shots, or even choose what coffee for the morning java.
    My concern is mostly that lots of people have their own ideas for how best to display the shot info to best use, and instead of my deciding "one true way" for everyone, I'll take a stab at it but also let others make their attempts too.

    Your request to show shot charts along with metadata, together, is reasonable enough, and likely to be how I'd show it. I prefer to do that in a web browser, because the page will likely run long and want for filtering capabilities.
    gc likes this.

  16. #616
    Senior Member
    Join Date
    Aug 2013
    Location
    Rockingham W.A.
    Posts
    946
    One of the best tools I know for determining tasting issues is the "Counter Culture Coffee: Taster's Flavor Wheel".

    The latest versions I have are 2014 (two main sections - good shots and problems separated) and 2015 (all integrated, https://counterculturecoffee.com/lea...s-flavor-wheel) and they are both brilliant for providing categories for flavours. Worth a look and possible incorporation at some level.
    gc likes this.

  17. #617
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    counter culture's chart

    Quote Originally Posted by TampIt View Post
    One of the best tools I know for determining tasting issues is the "Counter Culture Coffee: Taster's Flavor Wheel".
    I've mentioned it before, that Counter Culture's chart is also my favorite, because it allows people to get started describing their shots (ie, chocolate & fruity) at a beginner level, and as they progress, they can move outwards on the wheel (lemon zest, anis, dark bitter chocolate). I also think their categories are really useable.

    I have yet to see any system that separates the citrus flavors into (a) juice (b) zest and (c) pith. It's very common to find grapefruit pith, or lemon zest flavors, and those are really different from the juice flavors.

    My guess is that I will need to make a text file that explains the desired hierarchy, and let users edit them, perhaps with a sharing ability too.

    It will take me several years to code it all, but I'd like to support multiple systems, and multiple analysis methods, and really work with the coffee geek community to try to data mine this stuff so that it's actually useful to anyone passionate about espresso. For now, I'm making a first pass.

    CCC_Tasters_Wheel_LR-1.jpg
    gc and trentski like this.

  18. #618
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    What's in Release Candidate 3

    This iteration of the Decent Espresso machine sees us tidying a lot of things (everything should look neater) and also (importantly) we're going to be testing in real use our (hopefully final) heater insulation strategy, using a fiberglass box stuffed with insulating material. This also has the 3rd revision of our pump mounts, to dampen their sound. A more heat-isolated brass group head should arrive in a few days, and we're hoping that we're then at the end of this long R&D journey.

    MrJack likes this.

  19. #619
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    It's OK to go a little slower in your video tours if you want. You're probably a bit more familiar with what your looking at and we're only seeing it for the first time

    I had to replay vid several times to really take it all in.

  20. #620
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    why so fast?

    Quote Originally Posted by gc View Post
    It's OK to go a little slower in your video tours if you want. You're probably a bit more familiar with what your looking at and we're only seeing it for the first time
    I had to replay vid several times to really take it all in.
    I'm definitely guilty of that in this video.

    However, I had very recently covered the not-yet-built guts of this machine in another video, and didn't want to cover the same ground so soon. Here's the earlier (much longer) video of the same machine, not yet assembled.

    gc likes this.

  21. #621
    Senior Member
    Join Date
    Sep 2005
    Posts
    458
    This machine is looking more and more impressive as it gets closer to production..

    I noticed that you said "Lots of screws" in your listing of the parts in this last YouTube clip.
    Are you using all stainless fasteners in your machine?

  22. #622
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Just checked with Ray, and all screws are either stainless steel or--when heat isolation is desired--nylon.
    MrJack and Paolo like this.

  23. #623
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    Thoughts on avoiding bi-metallic corrosion with those stainless screws on aluminium (I was once a yacht owner where such things mattered) moist environments and all that?

  24. #624
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by gc View Post
    Thoughts on avoiding bi-metallic corrosion with those stainless screws on aluminium (I was once a yacht owner where such things mattered) moist environments and all that?
    Unless something is leaking, there shouldn't be any water inside the machine. There is a fan flowing air into the machine from the back, and thus out the side-apertures, so moisture should not remain, though moisture shouldn't even start to exist there in the first place. Since an espresso machine is a pressurized system, you really don't want any water leaks. No salt, either, which your yacht had.

    There is a possibility of "sacrificial ion exchange" between different alloys, but generally this creates an unsightly finish (inside the machine) rather than destroying it. We had that problem with our milk jugs initially, being a different stainless steel alloy than the tips of our digital thermometers. That caused a little black dot on the inside of the jugs. Now, the jugs use the same alloy as the thermometer tips. But again, you need something to help the ions move, such as an acid or salt, which won't be inside the DE1.

    So... that's the long answer. The short answer is "we've thought about it, not too worried about it, it's something to look out for, and if it's a problem in the field over several years, it's just a screw and easily changed". After all, we're not the only ones bolting metals together.
    Paolo likes this.

  25. #625
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    Thanks for the long answer! Clearly, its been thought about - and I'd forgotten about the fan.

  26. #626
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    DE1 tablet software updates

    Going from "beta" to "release" is always a big step, and it also presents my last chance to change anything that has been really bugging me, but I've been putting off. Also, most of my programming effort in the past year has been on the DE1+ (to wow! people) and now I need to do as good a job on the DE1 tablet software.

    screen 2017-09-05 at 6.32.53 PM.jpg

    In that vein, I have given espresso "presets" their own tab, and it's the first (default) tab you'll see. There's now a sentence or two about each preset, to help you understand why you'd want to use it. Espresso temperature now is set with clear +/- buttons rather than an obscure graphic that wasn't clearly for tapping.

    screen 2017-09-05 at 6.32.55 PM.jpg

    Making profiles: when I would let people make their own profiles on my demo machine, I noticed that sometimes they'd choose settings that wouldn't actually work. The chart made it seem like it would, but it wouldn't. The chart now much more accurately simulates the shot (taking into account how quickly pressure can be ramped up, for instance) and prevents you from doing things that are illogical (such as declining in pressure from 9 bar to 11 bar).

    What was previously labelled as step 2: "hold" is now "rise and hold" because the mechanics of what happened in this step confused people. You need to build pressure up after preinfusion, and then you can hold it. Previously, people could go from "preinfusion" to "decline", the chart would look like a Lever machine, but the shot wouldn't actually work, because there was no pressure rise. The software now always includes a "rise" step, to avoid that confusion.

    screen 2017-09-05 at 6.33.00 PM.jpg

    The "tablet styles" have been broken out onto their own half-page tab, and given a lot more vertical space. They were really cramped before.

    Hot water (for Americanos) now has a clear-to-understand slider control rather than my too-clever graphic.

    screen 2017-09-05 at 6.33.05 PM.jpg

    The Machine Maintenance page now has separate buttons for CLEAN vs DESCALING as these are different processes.

    Bluetooth pairing is now done with a dedicated button, rather than being magical. That makes it much easier to have multiple DE1 machines next to each other.

    I'll be making a new video, showing this new (and close to finalized) tablet gui in action.
    gc and RavenMad like this.

  27. #627
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    I could see how to use the machine pretty much immediately from the screenshots, so I think it is fairly intuitive at this point. In the early stages help popups/descriptive text when hovering over an item might be useful as people familarise themselves with a different way of doing things. Although I'm sure detailed videos will be pretty popular too. It really looks like its coming together now.

  28. #628
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by gc View Post
    I could see how to use the machine pretty much immediately from the screenshots, so I think it is fairly intuitive at this point. In the early stages help popups/descriptive text when hovering over an item might be useful as people familarise themselves with a different way of doing things. Although I'm sure detailed videos will be pretty popular too. It really looks like its coming together now.
    Unfortunately, tablets have no concept of "hover" which desktops with mouse control have. There is the possibility of "tap and hold to see help" but I think this might confuse more than help.

    Videos are indeed how I want to teach, because it allows me to communicate so much clearer.
    gc likes this.

  29. #629
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Sensor Basket Progress

    A few weeks ago I wrote about our "Decent Sensor Basket", which is an all-digital alternative to the Scace 2.

    The two main uses for this are:
    1) to digitally measure the temperature and pressure coming out of your espresso machine
    2) to calibrate your machine (because we use very accurate (and expensive) calibrated flow constrictors) to the flow rate you want

    Our first attempt at this has now come back from the machinist, and we'll be assembling and testing it over the next few days.

    sensorbasket.jpg

    Dimal, Angmav, level3ninja and 1 others like this.

  30. #630
    Junior Member TampPolice's Avatar
    Join Date
    Aug 2017
    Posts
    21
    Hi John

    Let me start off by saying thank you for being so open and explaining everything in so much detail. Over the past few weeks, this thread and the one on home-barista have been a pretty good bed-time reading material (not finished yet) and it feels like I've been part of your journey in developing this incredible machine.

    One of the reasons that most "aficionados" love their traditional gear is that they understand how it works and appreciate the complexity of process that goes in behind crafting a tasty espresso. Some (including myself) feel a little intimidated when we see a whole bunch of electronics and new design methods and we can't fully understand how it works. By sharing your work in such depth on these forums, there's enough material (at least for me) to get confidence and to gain appreciation of this marvellous machine's inner working and the hard work that has been put in behind the scenes.

    There is something that bugs me a little every time you mention the "220V" machines so I thought I'd ask the question.
    In Australia, the nominal supply voltage is 230V +10%, -6%. Most Ulka pumps used in Australian machines are rated at either 230V or 240V. A google search on rotary pumps shows similar ratings.
    At my home (inner Sydney) I usually get voltage between 240V and 250V.

    My two questions are:
    1) Are the pumps in DE machines rated to 220V? If so, will using a voltage upwards of 240V have any detrimental effect on the design life of the machine. If yes, what would be your estimate on design life with using 245V?

    2) How will the electronics perform with a constant use under 245V? Is all certification/testing done around 220V? What would be (if any) the failure rate when operating under higher voltage?

  31. #631
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by TampPolice View Post
    it feels like I've been part of your journey in developing this incredible machine.
    Good, I'm glad to hear that, because absolutely, this machine has been co-designed with some dozen people on these forums.

    Quote Originally Posted by TampPolice View Post
    One of the reasons that most "aficionados" love their traditional gear is that they understand how it works and appreciate the complexity of process that goes in behind crafting a tasty espresso.
    Some (including myself) feel a little intimidated when we see a whole bunch of electronics and new design methods and we can't fully understand how it works. By sharing your work in such depth on these forums, there's enough material (at least for me) to get confidence and to gain appreciation of this marvellous machine's inner working and the hard work that has been put in behind the scenes.
    That makes sense: everyone (including me) has a life of experience with boilers, such that we literatally can't think of how things might work with a different heating approach. Similarly vibratory pumps, usually only seen on low end machines, can be controlled to act like a gear pump, except that they provide much faster flow rate changes, thus enabling water mixing.

    There's a very famous computer UI design book by Brenda Laurel entitled "Computers as Theater" that really influenced me.
    https://www.amazon.com/Computers-The.../dp/0321918622

    The argument she makes is essentially that your UI either
    1) has to provide a perfect, no flaws illusion
    2) or it has to show the audience how things actually work

    My approach is #2, but granted, the can cause information overload if not handled well.

    Quote Originally Posted by TampPolice View Post
    There is something that bugs me a little every time you mention the "220V" machines so I thought I'd ask the question.
    In Australia, the nominal supply voltage is 230V +10%, -6%. Most Ulka pumps used in Australian machines are rated at either 230V or 240V. A google search on rotary pumps shows similar ratings.
    At my home (inner Sydney) I usually get voltage between 240V and 250V.

    My two questions are:
    1) Are the pumps in DE machines rated to 220V? If so, will using a voltage upwards of 240V have any detrimental effect on the design life of the machine. If yes, what would be your estimate on design life with using 245V?
    Have no fear, the pumps and heaters are rated at either 120V or 240V, depending on the machine voltage you're ordering.

    Quote Originally Posted by TampPolice View Post
    2) How will the electronics perform with a constant use under 245V? Is all certification/testing done around 220V? What would be (if any) the failure rate when operating under higher voltage?
    I'm not so worried about the electronics, because we use a fairly expensive high end Meanwell power supply in front of our electronics, to get a clean low voltage feed.

    Here's the spec sheet:
    LPV-35-MEAN WELL Switching Power Supply Manufacturer

    Here are its input voltages:
    1.1A/115VAC
    0.7A/230VA

    it looks to me like they've rated it smack in the middle of the standards (110/120 and 220/240).

    Its operating voltage range is:
    90 ~ 264VAC

    Here in Hong Kong we're running our machines at both 220V/50hz and 120V/60hz (and occasionally 240/60hz for South Korea), but Intertek, our testing company, will subject us to a wide variety of bad and dirty electricity.

    For UL safety certification, Intertek will have 8 machines operating at 110V/120V.

    For CE safety certification, Intertek will have 4 machines operating at 220V/240V.

    For the certifications, they will completely destroy all 12 machines, and so if we pass that we'll know what our safety tolerances are.
    gc, LeroyC and Lav like this.

  32. #632
    gc
    gc is offline
    Member
    Join Date
    Dec 2008
    Location
    Kuranda, Australia
    Posts
    90
    I was a bit worried about the 240V in Oz too, so thanks for the clarification.

  33. #633
    Junior Member TampPolice's Avatar
    Join Date
    Aug 2017
    Posts
    21
    Quote Originally Posted by decentespresso View Post
    There's a very famous computer UI design book by Brenda Laurel entitled "Computers as Theater" that really influenced me.
    https://www.amazon.com/Computers-The.../dp/0321918622

    The argument she makes is essentially that your UI either
    1) has to provide a perfect, no flaws illusion
    2) or it has to show the audience how things actually work

    My approach is #2, but granted, the can cause information overload if not handled well.
    Couldn't agree more!

    Quote Originally Posted by decentespresso View Post

    I'm not so worried about the electronics, because we use a fairly expensive high end Meanwell power supply in front of our electronics, to get a clean low voltage feed.

    Here's the spec sheet:
    LPV-35-MEAN WELL Switching Power Supply Manufacturer

    Here are its input voltages:
    1.1A/115VAC
    0.7A/230VA

    it looks to me like they've rated it smack in the middle of the standards (110/120 and 220/240).

    Its operating voltage range is:
    90 ~ 264VAC
    Ok, that gives me a lot of comfort knowing that you are using MW supplies. I have seen hundreds of them used across the industrial and chemical plants in Australia. A few years ago I was involved in the electrical design for some major upgrade works across 100+ sites (pump stations, water treatment plants) for the largest Water Supply Authority in Queensland. They insisted on using Meanwell power supplies (a different larger model of course) for all of their control gear. It goes without saying that reliability is not a joke in that industry. So it doesn't really surprise me to see an MTBF figure of 80+ years published by MW on the model you are using.


    Quote Originally Posted by decentespresso View Post
    For the certifications, they will completely destroy all 12 machines, and so if we pass that we'll know what our safety tolerances are.
    What a shame eh!

  34. #634
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by TampPolice View Post
    Ok, that gives me a lot of comfort knowing that you are using MW supplies. I have seen hundreds of them used across the industrial and chemical plants in Australia. A few years ago I was involved in the electrical design for some major upgrade works across 100+ sites (pump stations, water treatment plants) for the largest Water Supply Authority in Queensland. They insisted on using Meanwell power supplies (a different larger model of course) for all of their control gear. It goes without saying that reliability is not a joke in that industry. So it doesn't really surprise me to see an MTBF figure of 80+ years published by MW on the model you are using.
    That's quite interesting and helpful to know, thanks for sharing that.

    I initially learned about Meanwell power supplies by having a long discussion with the engineer behind the "Disorient" light shows at Burning Man, running on several Beaglebone Black computers. Burning Man is a really hostile environment to tech, and this guy swore by MW and never had a failure in a decade.

    NeoMonster2.jpg

  35. #635
    Junior Member TampPolice's Avatar
    Join Date
    Aug 2017
    Posts
    21
    Quote Originally Posted by decentespresso View Post
    I initially learned about Meanwell power supplies by having a long discussion with the engineer behind the "Disorient" light shows at Burning Man, running on several Beaglebone Black computers. Burning Man is a really hostile environment to tech, and this guy swore by MW and never had a failure in a decade.

    NeoMonster2.jpg
    Thats pretty incredible. Looks like a really tough environment for the IGBTs.
    Speaking of which, I just looked at the datasheet in your earlier post and couldn't help myself to do some numbers so here they are:

    Assuming that the 5V model is used, the efficiency of the unit is 77%
    The graph at the end of the datasheet shows a temperature derating of about 58% at 60 degrees ambient.

    From your past posts I can see that you are providing a USB port. For Samsung tablets, this limits the charging power to 2.5Watts (not sure about other tablet manufacturers)

    To put it all together:
    De-rated maximum available power: 30W x 77% x 58% = 13.4W
    Power available after subtracting tablet's charging current = 13.4 - 2.5 = 10.9W

    A good practice would be not to load this small of a unit more than 80% of the derated wattage. This can help with any defect claims down the track from MW.

    However even looking at the 100% derated wattage, this will leave you about 11W of power for the circuit board and flow meter. Is that sufficient?

    From previous posts I can see that the electronics are isolated from boiler with a metal barrier. Is the expected ambient temperature around the electronics and the power supply less than 60 degrees? If so, that can help.
    gc likes this.

  36. #636
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by TampPolice View Post
    From previous posts I can see that the electronics are isolated from boiler with a metal barrier. Is the expected ambient temperature around the electronics and the power supply less than 60 degrees? If so, that can help.
    I'll do my best to answer your questions, but keep in mind that while I'm the project leader, I'm a programmer, not an EE, and I kind of don't want to bug Ray at the moment as he has a new approach to steam that he is coding at the moment. Instead of constant water flow, and controlling the heater, now that he has good data from the flow meter (we've struggled with that) he wants to keep the steam heater power at 100% and vary the flow via PID. That should give us a bit more steam power.

    But speaking to your points:
    - we stop charging the tablet when making espresso, because you are right, we'd be close to the edge if the electronics were at 60, the machine were making steam, and we were charging.
    - yes, we have a heat shield between the electronics and the rest of the machine, but we also have a fan inside the electronics area, pulling cool air from the back into the PCB area
    - the non-electronics area is running at about 60C
    - the electronics area is currently running at a peak of 52C, if you really stress the machine out (lots of steam) but its normal constant temp is 44C
    - thus we are within the power supply's capabilities

    Now that you've done the calculations, you can look a few screens previous and see why I mentioned that tablet-charging is disabled during espresso making.

    On the DE1CAFE we plan on moving to the Meanwell 60W power supply, so that we can:
    a) charge the tablet constantly
    b) have sufficient power to run more cooling fans
    MrJack and trentski like this.

  37. #637
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    A "Slayer Shot" : two approaches

    I'm currently reworking the tablet software for our DE1+ espresso machine, and one of my major goals is giving you the power to make very sophisticated espressos, such as the "Slayer Shot", but with accuracy and repeatability.

    On a Slayer espresso machine, you can adjust a mechanical "needle valve" to control the pre-infusion flow rate. People have found that slow pre-infusions work very well with lightly roasted beans.


    On the DE1+ we give you precise control over the flow rate, so you can exactly set it to what gives you the best flavor result.

    A few days ago, the distributors of Slayer in Thailand spent the day with me here in Hong Kong. Their tech guy came along and really helped me understand how this style of espresso is made. In the end, we achieved shots that generated many Thai words that sounded like "yum".

    My experience is that ground espresso can hold about 2x its weight in water. We set the pre-infusion to 2 mL/second, for 18 seconds, to saturate our 18g espresso coffee puck.

    On our first attempt, water leaked through before pre-infusion was over, and the tech guy told me that the grind has to be really fine for this to work. So, adjust grinder, repeat.

    After pre-infusion, rise to full pressure (9 bar) or 25 seconds. You should have 36g in the cup.

    Below is the screen shot on the DE1+ showing this shot program using pressure profiling. Note that the chart tries to guess that pressure will be rising during pre-infusion and charts it. However, it's just a guess, and if your coffee is coarse, that won't happen, but still, this chart is my best attempt to let you see what will happen with this shot program.

    slayer1.jpg

    Another approach, with different pros and cons, is to use flow profiling. This is the part of the tablet software I've been lately working on improving.

    slayer6.jpg

    Since pre-infusion on our machines is always flow controlled, there's no difference there.


    After pre-infusion, though, instead of rising to 9 bar (with whatever flow rate that might cause), I instead opt to "rise" the pressure to 4 bar, which guarantees that the puck has been compressed and will now make espresso. You can see that the chart represents this by raising the flow rate the 6 mL/s for a few seconds to achieve the desired pressure.

    Next, this shot maintains a constant flow rate of 2 mL/s for 18 seconds, to achieve 36g in the cup. Because the coffee puck is fully saturated at this point, each gram of water in should (more or less) come out. This might not be completely true (coffee is not that reliable) so you might need to change the "hold" period to get the in-cup volume you asked for.

  38. #638
    Junior Member TampPolice's Avatar
    Join Date
    Aug 2017
    Posts
    21
    Quote Originally Posted by decentespresso View Post
    I'll do my best to answer your questions, but keep in mind that while I'm the project leader, I'm a programmer, not an EE, and I kind of don't want to bug Ray at the moment as he has a new approach to steam that he is coding at the moment. Instead of constant water flow, and controlling the heater, now that he has good data from the flow meter (we've struggled with that) he wants to keep the steam heater power at 100% and vary the flow via PID. That should give us a bit more steam power.

    But speaking to your points:
    - we stop charging the tablet when making espresso, because you are right, we'd be close to the edge if the electronics were at 60, the machine were making steam, and we were charging.
    - yes, we have a heat shield between the electronics and the rest of the machine, but we also have a fan inside the electronics area, pulling cool air from the back into the PCB area
    - the non-electronics area is running at about 60C
    - the electronics area is currently running at a peak of 52C, if you really stress the machine out (lots of steam) but its normal constant temp is 44C
    - thus we are within the power supply's capabilities

    Now that you've done the calculations, you can look a few screens previous and see why I mentioned that tablet-charging is disabled during espresso making.

    On the DE1CAFE we plan on moving to the Meanwell 60W power supply, so that we can:
    a) charge the tablet constantly
    b) have sufficient power to run more cooling fans
    Thank you for clarifying this John.
    I'm only up to page 9 of this thread. Will get there soon.
    The intent of the comment was merely to highlight that higher temperatures do in fact decrease the life of switch-mode power supplies and other electronics exponentially. With your answer I'm satisfied that it is appropriately sized.

    The terms 'reliability' and 'durability' are very subjective and prolonged use under higher temperature and humidity will reduce the life of just about anything where transistors and IGBTs are used.
    Most manufacturers only publish data based on the ideal environments i.e room temperature and as you pointed out, the temperature around the electronics within the espresso machine can go up to 50C. How much direct and cascaded indirect effect will it have on the overall reliability of the machine is unknown at this stage and without knowing how much redundancy is there in the system for critical paths or 'more likely' points of failures, it is extremely difficult to extrapolate the expected design life of the electronics portion of the machine.

    Why do I care about all of this? I don't expect my machine to last me 25 years. If I'm spending $4k on an espresso machine (DE1+Pro), I would like for it to last at least 5-7years without causing any grief and then hold a re-sale value of 60%. If someone is paying $2.5k for a 7 year old machine, they would want to know how many parts they need to get replaced for it to last them another 7years. If the answer is between $200 and $600 then they could justify buying a second-hand machine to save $1k. But since a 'tablet' is likely to be the first item in the list of things to replace, this leaves very little room for any other parts to be replaced for a pronlonged hassle-free operation.

    Can this machine produce shots better than ANY other prosumer machine available in the market (Slayer included)? I don't think anyone will have any doubts on that.
    Can it do so without any electronics or software issues? That's the big question, right?

    Many forum members with a variety of skill-set and expertise have already provided some good positive feedback earlier in this thread. And I can see that you go into great level of detail to ensure that reliability is not compromised.

    So once again, thank you for your time with answering all queries and concerns.

    From this thread I can see that building a competitive espresso machine in 2017 is very much like large multi-disciplinary project. You need to know just about every subject matter to succeed and you've done a tremendous job so far.
    Last edited by TampPolice; 5 Days Ago at 02:40 AM.

  39. #639
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by TampPolice View Post
    Most manufacturers only publish data based on the ideal environments i.e room temperature and as you pointed out, the temperature around the electronics within the espresso machine can go up to 50C.
    Agreed, and that's why during the past 2 months, as we've mostly gotten a handle on making espresso, we've been focussing on waste heat. Besides costing money, increasing machine warmup time, and being un-green, that waste heat can lower electronic's lifespan, especially (as you wrote) if the electronics go higher than 60C. That's one reason why we have compartmentalized the electronics and are blowing outside air into that compartment. Besides cooling the electronics, this also keeps hot air from leaking in from the rest of the machine ("positive pressure"). I believe the last thing we have on our todo list here is to stick an insulator on the back of the separating plate, to inhibit radiative heating of the electronics.

    Quote Originally Posted by TampPolice View Post
    But since a 'tablet' is likely to be the first item in the list of things to replace, this leaves very little room for any other parts to be replaced for a pronlonged hassle-free operation.
    My hope is that the tablet we provide will last longer than its technical obsolescence. What I mean by this, is that in a few years, if you can buy a "retina display tablet with a 10x faster processor" for A$200, you might want to upgrade the tablet that came with the espresso machine, even if there's nothing wrong with it. Most of my past mobile phones are in this category. They still function, but I don't use them.

    FYI if the tablet we provide should happen to break after your 1 year warrantee, we sell the replacement at A$161. I know people have a worry about this component, and that's why we're selling the replacement at a reasonable price. Given your stated budget for repairs, this tablet price should leave room in your budget for other repairs.

    And just to be clear, the current tablet works perfectly fine. It's fast, the screen is clear and has nice colors. There's no "problem" with it. But, it is a 1GB RAM/720p resolution tablet, and better tablets are available on the market, if you want to spend the money.

    ---

    We will be selling all the parts to make our espresso machines, online, as part of our shopping cart. The parts will be organized hierarchically based on the assembly they are part of. We'll also be releasing all the CAD drawings to the machine via onshape.com, to help you (or a repair person) understand and fix the machine.

    Here's a small sample of the parts database that will be online soon:

    screen 2017-09-15 at 10.02.07 AM.jpg


    Quote Originally Posted by TampPolice View Post
    Can it do so without any electronics or software issues? That's the big question, right?
    As to software issues, there will certainly be some bugs, but since I can push out new software versions over the Internet, I'm hoping any bugs discovered will be short lived.

    One BIG reason we're supplying a tablet is that having control over the hardware removes a massive source of Android bugs. I have a pile of tablets that I've tested that have little niggly bugs in their Android distribution. Since Android 6, Android has been getting a lot less buggy, but tablets that support that OS are much more expensive than the Android 5.1 tablet we're using, and this 5.1 tablet happens to be well made (thanks to Argos, the UK-based electronics chain that developed it a few years ago).

    Android 8, recently released, has me very excited, as Google is forcing all tablets to be compatible with a generic version of Android, and so this will allow us to make our own crud-free Android distribution. It will also allow us to buy a less expensive tablet but still run current software on it.
    MrJack, dlight, Lav and 1 others like this.

  40. #640
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Help me figure out flow profiling

    Im working on the tablet software for flow profiling and could use your opinion about if what Ive done makes any sense.



    Just one correction: yes, I know the Mina espresso machine has flow profiling, so the DE1+ is the 2nd flow profiling machine, but what I meant was that there haven't been any GUIs trying to help people make flow profiling shots.

  41. #641
    Junior Member TampPolice's Avatar
    Join Date
    Aug 2017
    Posts
    21
    Quote Originally Posted by decentespresso View Post
    Im working on the tablet software for flow profiling and could use your opinion about if what Ive done makes any sense.



    Just one correction: yes, I know the Mina espresso machine has flow profiling, so the DE1+ is the 2nd flow profiling machine, but what I meant was that there haven't been any GUIs trying to help people make flow profiling shots.

    My only concern here is an inadvertent inception of a channel due to a sudden drop in pressure.

    For example with the following settings:
    1: Preinfuse
    Time: <5sec
    Flow: 3.5ml/s
    Pressure: 4.5bar

    2: Hold
    Time: 30sec
    Flow: 2ml/s

    The flow is likely to rise at 6ml/s for approx. 3 to 4sec and then drop to 33% of 6ml/s.
    Is 6ml/s the max threshold for it to achieve the desired pressure or could it be higher?

    I guess only way to find out is to try it out with a darker and a lighter roast to see if you can spot any channels during the hold period.

    Otherwise I think it's an excellent option to have. I can only imagine how much fun it would be writing a manual for DE1's operation
    You should probably include the manual on the tablet and update with every firmware update.

  42. #642
    Junior Member
    Join Date
    Sep 2017
    Posts
    1
    Release Candidate 3 looks amazing. What can you tell us about the metallurgy of the brass in contact with the brew water in the brew head and brew water dispersion plate?

  43. #643
    Senior Member level3ninja's Avatar
    Join Date
    Nov 2013
    Location
    Casula, NSW
    Posts
    406
    Makes sense to me, the only thing you could do to make it clear would be to add an "OR" somewhere to show it stops when either 4 bar or 10s is reached. Also I find it a little confusing that without the guarantee ticked you show a <10s and >4bar, when it might not reach 4bar. Can I suggest some sort of symbol denoting what will happen, like <10s and ➚4bar (an arrow pointing up right to symbolise towards 4bar)? And maybe a different symbol when the guarantee is checked to indicate a definite 4bar, like ✓4bar or =4bar? Just to give an extra reminder of the expected function.

  44. #644
    Senior Member
    Join Date
    Oct 2011
    Posts
    1,315
    Quote Originally Posted by decentespresso View Post
    I’m working on the tablet software for flow profiling and could use your opinion about if what I’ve done makes any sense.

    ...

    Just one correction: yes, I know the Mina espresso machine has flow profiling, so the DE1+ is the 2nd flow profiling machine, but what I meant was that there haven't been any GUIs trying to help people make flow profiling shots.
    I personally think it might confuse people unnecessarily.

    As soon as you introduce the pressure specification, it's not really flow control any more (looks something like cascade control).

    If the grind is too coarse to reach the target backpressure at the specified flowrate, will a temporary deviation from the spec give you the same result as just tightening up the grind and trying again? I'll go out on a limb and say probably not.
    That's not to say that it won't produce a nice shot, but it might be hard to reproduce. Guess it depends on what you are trying to achieve really.

    A couple of general thoughts/observations:


    • If your target is reproducible shots, implementing a preinfusion volume specification might be better than a time specification.
    • It is interesting that you typically pre-infuse at a greater flowrate than your peak extraction flowrate (I do the opposite)
    • I'm not sure that exceeding a minimum backpressure is required for puck compression - I've not seen any evidence of a threshold in the data*. In my own experimentation I have observed that low flow/pressure can impact crema formation, but I think this is linked to CO2 dissolution and the velocity of the fluid through the holes in the basket.


    *This is 3D visualisation of plot of pressure, flowrate and volume, in case you're interested: https://www.instagram.com/p/BYsVDhDh45R/

  45. #645
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278

    Sensor Basket thoughts.

    Here's what the Sensor Basket looks like after being used for a few dozen shots. This was made in sandblasted aluminum and looks a bit industrial.

    Because this prototype was made by carving out a block of aluminum, the "lip" is thicker than a traditional basket. This made the fit in the espresso machine too tight, and a portafilter with the Sensor Basket had to be wedged hard. To solve this, I sanded off a bit off the top tip (you can see the sanding marks). In the next iteration, we'll try to take a millimeter off the thickness, or else flatten the top on purpose, like I did with the sander.

    I haven't yet installed the shield on the pressure sensor since it's not wired in. You can see where that'll go in the photo below, and it's there to prevent splashing water from giving you temporarily high pressure readings.

    One thing that was not great about the Scace was that its flow rate was much faster than a good espresso actually is, making it a not-great simulation for our testing. The Scace's flow rate is more like a "gusher" espresso.

    What I've found, however, is that a smaller hole, that properly simulates real espresso, is much more likely to get clogged with a coffee particle. So, there might be a method to Scace's madness. <grin>

    On the positive side, because the water comes out of a "barb" you can (ahem) put your lips around the exit barb and blow hard to push the particles out. Then give it all a nice rinse and try to forget what you just did.

    If you'd like to be more professional, you can buy a very thin guitar string and use that to push the jammed particle out.

    Or, you can clean out your espresso machine better than I did, before using this testing tool.

    (update 1h later) or you actually install the mesh filter that was designed to prevent this clogging problem (sigh).

    IMG_7236.JPG IMG_7237.JPG
    IMG_7238.JPG IMG_7239.JPG

    MrJack likes this.

  46. #646
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by TampPolice View Post
    My only concern here is an inadvertent inception of a channel due to a sudden drop in pressure. The flow is likely to rise at 6ml/s for approx. 3 to 4sec and then drop to 33% of 6ml/s.
    Is 6ml/s the max threshold for it to achieve the desired pressure or could it be higher?
    6mL/s is the current maximum flow rate the DE1+ can achieve. When we require 220V (as on the DE1CAFE) and put even larger heaters (and heat dissipation) in we'll be able to go to higher flow rates.

    For comparison, my GS/3, full throttle, pumps out 8 mL/s, but that's a punishing flow rate for a barista to use, as it demands very good puck preparation or you get terrible channeling.

    Generally, if you skip preinfusion and go straight to 6 mL/s, unless your puck is really well prepared, you'll have some channeling. Preinfusion around 4 mL/s really is a good thing, even if it's just for 4s.

    Quote Originally Posted by TampPolice View Post
    Otherwise I think it's an excellent option to have. I can only imagine how much fun it would be writing a manual for DE1's operation
    You should probably include the manual on the tablet and update with every firmware update.
    My plan is to focus on videos, a few minutes long, to explain these things. I can do a much better job of explaining things that way, but also people can ask questions in the youtube comments, which then become part of the permanent record.

  47. #647
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by HBDecentRefugee View Post
    Release Candidate 3 looks amazing. What can you tell us about the metallurgy of the brass in contact with the brew water in the brew head and brew water dispersion plate?
    Attached are the lab test results for the brass pieces, all of which are in contact with water. I've also included a drawing of one of the pieces, so we're all on the same page as to what we're talking about.

    Note the "Pb < 0.0005" number, which is the crucial one for Germany and Japan, as they are very tough on lead content.

    FYI, here is a complete list of the materials that come into contact with the water, in "flow order"
    - ceramic water tank
    - stainless steel uptake tube
    - medical grade silicone rubber tubing (to bring water to the pumps)
    - (I believe) ABS plastic is what the ODE pumps use. Need to look this up, but these come from an Italian espresso machine parts company and have certifications.
    - PTFE (solid teflon) tubing
    - stainless steel ferrules inside the PTFE tubing
    - stainless steel tubing (inside the water heaters).
    - ULTEM resin (from Switzerland) for the mixing chambers and sensor mounts.
    - brass (group head)
    - stainless steel portafilter basket (I can supply testing certificate, if you're curious).
    - aluminum drip tray cover
    - ceramic drip tray

    We currently (or will) have certification at each step, from the manufacturer, except for the ceramic, which we are sending to a lab ourselves for testing (though the company we're using for the ceramics is exclusively in the cookware business, and sells their products in the USA, so we are not especially worried)

    lead1.jpg lead2.jpg

  48. #648
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by level3ninja View Post
    Makes sense to me, the only thing you could do to make it clear would be to add an "OR" somewhere to show it stops when either 4 bar or 10s is reached. Also I find it a little confusing that without the guarantee ticked you show a <10s and >4bar, when it might not reach 4bar. Can I suggest some sort of symbol denoting what will happen, like <10s and ➚4bar (an arrow pointing up right to symbolise towards 4bar)? And maybe a different symbol when the guarantee is checked to indicate a definite 4bar, like ✓4bar or =4bar? Just to give an extra reminder of the expected function.
    In discussions on other fora, I've agreed to these suggestions from others:
    - both symbols should actually be <
    - that the "guarantee" word should be changed to "force"
    - that the "rise to 6 mL/s" in the chart should always be displayed so as to avoid confusion ("why is it disappearing??!?!?")

    and I hope that will address your concerns too.
    MrJack likes this.

  49. #649
    Site Sponsor
    Join Date
    Aug 2016
    Location
    Hong Kong
    Posts
    278
    Quote Originally Posted by MrJack View Post
    As soon as you introduce the pressure specification, it's not really flow control any more (looks something like cascade control).
    True: my goal is less to be "pure" about flow vs pressure profiling, and instead to make good coffee. If pressure is a useful concept, I'm ok to using it, even in the beginning of what I call a "flow profile"

    I haven't mentioned this before, but instead of "exit preinfusion when X bar is reached" we could instead "exit preinfusion when flow slows to under Y mL/s".

    This produces identical results, but in my presenting these concepts to people, this alternative approach was even more confusing to people, so I've avoided promoting it (though this capability is available in the ADVANCED shot editor).

    And just to be clear, the way this other approach works is that
    - initially the puck can absorb water at a fast rate
    - as the puck saturates it resists accepting water, water builds up and fills the group and then the espresso machine notices (charts) the increased pressure, and actual flow rate decreases below some set threshold.
    - at that point, the pack can be considered "compressed".

    Quote Originally Posted by MrJack View Post
    If the grind is too coarse to reach the target backpressure at the specified flowrate, will a temporary deviation from the spec give you the same result as just tightening up the grind and trying again? I'll go out on a limb and say probably not. That's not to say that it won't produce a nice shot, but it might be hard to reproduce. Guess it depends on what you are trying to achieve really.
    Agreed.

    Quote Originally Posted by MrJack View Post
    If your target is reproducible shots, implementing a preinfusion volume specification might be better than a time specification.
    That's a really interesting suggestion, and that is explicitly possible in the ADVANCED editor, viz the arrow in the screen shot:

    Attachment 17093

    A major challenge I have, as the GUI designer for this machine, is that so much of what the DE1+ presents is new to people, that I have to pick between "what people will easily understand, because it's similar to what they already know" vs "what makes the most sense objectively".

    I absolutely agree with you that "volume" makes more sense than "time" during preinfusion. But people have always thought of preinfusion in terms of time. Can they cope with the change?

    So yeah, I could absolutely change preinfusion to have these sliders:
    - Water volume: 36 ml
    - Flow rate mL/s: 4.0
    - (calculated for you) Estimated preinfusion time: 9 seconds

    Super interesting suggestion, thank you.

    Quote Originally Posted by MrJack View Post
    It is interesting that you typically pre-infuse at a greater flowrate than your peak extraction flowrate (I do the opposite)
    I think you and I are talking about different numbers.

    Are you measuring actual flow rate through the puck when you write that?

    I agree, that with existing espresso machines, people generally do something like this:
    - preinfuse at 4 mL/s
    - infusion at 6 mL/s to 8 mL/s

    Or, with a paddle, they open it up a bit until there are some drops, and open the paddle up all the way.

    However, if you were to measure actual water flow rate into the puck, they'd find that during infusion, flow is actually in the 2 mL/s to 3 mL/s range.

    The proof of this, is that if you really were getting 6 mL/s through your puck during infusion, that you'd complete your 36g espresso shot in 6 seconds.

    Do you agree?

    Quote Originally Posted by MrJack View Post
    I'm not sure that exceeding a minimum backpressure is required for puck compression - I've not seen any evidence of a threshold in the data
    Perhaps we are disagreeing because of our use of different words. Instead of "minimum backpressure" what if we say "pressure measured inside the group", do you then agree?

    For me, the very definition of a compressed puck is that it is creating pressure inside the group. If the pressure in the group is at 4 bar, the puck must be compressed for this to occur. If the pressure in the group is 1 bar, the puck might or might not be compressed, it's hard to tell.

    In my experience, a puck is fully compressed around 3 to 4 bar.

    By "a fully compressed puck" I might suggest this test:
    - if you slow the flow rate down to 1 mL/s for 20s, does pressure in the group increase or not?

    If not, then I would say the puck was not compressed (water is dripping through the puck).

    If pressure does increase, for me the puck was compressed (water is building up behind the puck, increasing pressure).

    Interesting conversation, thanks.
    level3ninja likes this.

Page 13 of 13 FirstFirst ... 3 11 12 13

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •