[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/diy/ - Do It Yourself



Thread archived.
You cannot reply anymore.



File: DrinkMixer.jpg (36 KB, 1208x685)
36 KB
36 KB JPG
So I'm trying to build a drink mixer. It'll have to be cheap and it'll be based around a Rpi.
Pic related, it's how it's supposed to look.
I decided against a air pressure or gravity powered pouring system, so it'll come down to pumps.
Now other projects that I looked at were incredibly obsessive about exact pouring, honestly, I don't care that much. I mean obviously it's supposed to be reasonably accurate but that's enough.
Problem 1: The pump:

There's 3 pump choices: Centrifugal -> Good Volume per time, really bad accuracy, cheap
peristaltic pump -> awesome accuracy, little volume per time, costs more
Diaphragm pump -> okay accuracy (i guess?), okay volume, reasonably cheap
As you may have guessed by now, I'm in favor of a diaphragm pump atm.
>is that a good or a stupid idea? Will the diaphragm be okay in terms of accuracy or will I need to get a special one with a stepper motor for that?
Problem 2: Tubing
So I can either:
Use one Pump for each liquid type. -> Pros: Faster Liquid serving, same price if cheap pumps, otherwise more expensive.
Or:
Use one Pump only and a solenoid valve for each liquid -> Pros: Cleaning can be automated, cheap. (Cons: liquid will mix to some extend)
>What do you think makes more sense?

Also only way to make the centrifugal pump work would be by measuring the liquid container. Set variable to 0 when container is placed. Measure the force on the sensor when liquid is poured, this should be the force of the liquid (mass) + the pressure of the liquid shot into the gllass/container. If you'd take a plastic tube, sealed the end and added holes it would squirt the liquid to the walls of the container so it'd almost eliminate the pressure aspect, thus only leaving the mass to measure. All in all pure centrifugal pumps are hardly worth the effort.

Pic related, rough sketch of the machine's exterior
>>
I am not sure if you can draw 12V from a raspberry pi, but you might want to have a look at these pumps. I think they suit your project nicely.
https://m.aliexpress.com/search.htm?keywords=priming+diaphragm+pump+spray+motor+12v#/
>>
http://lmgtfy.com/?q=diy+raspberry+pi+drink+mixer
http://lmgtfy.com/?q=diy+robot+drink+mixer
>>
>>1307710
Obviously I'll work with relais, directly driving pumps won't work
>>
>>1307714
Yeah thanks, that's obviously why I wrote the stuff up there, because I can't google it.
The questions I asked I asked becase the problems I face have not been well adressed or ignored in the classical google respones.
>>
>>1307710
Thanks though for the pumo, the problem as mentioned above though is this: Those diapharagm pumps are likely run by cheap motors, not stepper motors or such, so I'm feeling like accuracy would become quite the issue.
>>
>>1307705
why are the bottles not upside down and therefore gravity fed, gravity is free, why are you making things difficult for yourself having to pump liquids up when you could use gravity and timed electronic solenoid valves.

>turn bottles upside down
>add electronic solenoid valves to bottles
>make rpi open and close the valves. timing dependant on the amount of liquid from each bottle required to make the mix

should be fairly simple.
>>
>>1307705
why not air pressure? it's pretty cheap to put an aquarium pump on an SSR, vent the output into a two-hole stopper, and push the liquid up a dip tube in the other hole. no worrying about material compatibility or viscosity as long as the stopper seal is good
>cleaning
would be hell with a diaphragm pump that's moving the liquid
>>
>>1307721
Yeah I considered gravity fed (Or pressure powered) systems. The main problems with Gravity are these:
Inconsistent flow rate as bottles drain
harder to build (not the electical part but the case), and I want to add other stuff that's not in the sketch so the design is impractical.
>>1307727
Air pressure was another option but to be honest, I'll need the same amount of parts (Pressure pump instead of liquid pump), it'll cost roughly the same, it'll be slightly more effort to build because everything needs to withstand the pressure and wouldn't air pumps that build pressre likely be louder?
>>
>>1307719
Your questions are fucking retarded. Google a project and follow it.
>>
>>1307737
Thanks for the input, you can stop sharing your opinion now.
>>
>>1307729
you don't need several bars of pressure. you probably wouldn't even need 1 bar. you just need to inject a volume of air and keep it from coming back out except as the liquid it replaces. aquarium pumps are relatively quiet and about 5 burgerbux each. you probably don't even need an SSR, a triac/optotriac would probably be enough
>the only way to make centrifugal work
consider a strain gage under each bottle or under the whole assembly
also you really should consider reviewing the existing literature
>>
>>1307741
Thanks, I'll consider the pressure Idea, but how is that better than a "normal"pump? I don't really get how it's better.
I checked some literature already, but the focus is always set very different from my project. It's not my first microprocessor project, so I'm confident it'll work out. Any Ideas where else I should look? I checked the basic duckduckgo/google results and in regards to literature some books on arduino but apart from programming it's the same I'd guess.
>>
>>1307743
you have no need to clean the pump itself, no worry about syrup clogging flap valves or strong flavors hanging around or carbonated beverages flattening or material incompatibility. the fluid handling portions can be easily washed or just replaced.
you could use a low voltage dc-motor-driven air pump, if you prefer, and reap basically the same advantages.
you might build a single channel of a few different types for research
>>
>>1307747
That's actually a really good argument that I didn't consider as much as I think I should have...
I mean the solenoid valves would still need to be in contact with the liquid but that's all. Thanks! I'll test that.
>>
File: 1508046710075.jpg (170 KB, 1024x683)
170 KB
170 KB JPG
>>1307748
>the solenoid valves would still need to be in contact with the liquid
nope
>>
>>1307748
besides that, you wouldn't need solenoid valves of any kind if you use multiple air pumps and plumb the output tubes straight to the spigot
>>
>>1307751

I'm too cheap though, so I'll likely get pic related or something. Your point still stands though.
>>1307754
multiple air pumps seems redundant though, I mean then it'd still cost. Also if I don't use valves the pumps cant create pressure so it'd be inconsistant again...
thanks for the help already dude! Helped me out a bunch
>>
>>1307775
also I'm retardedm pic related:
>>
File: 1506060996138.png (48 KB, 800x800)
48 KB
48 KB PNG
>>1307775
>Also if I don't use valves the pumps cant create pressure
they don't need to create much pressure at all, they just need to ***displace*** the potent potables in the bottle. pressure will work its way back to equilibrium once the pump goes off, preferably by displacing delicious alcohol from the bottle into the cup. in fact, the less pressure that builds inside the bottle, the better -- maybe a larger ID exit tube is best.
assuming your solenoid valves are indeed leak-tight, the pressure in the bottle shouldn't be more than the few inches of water column from the surface of the liquid up to the spigot which is on the order of 1psi
if the solenoid valves are only $2 that's probably a better way to go.
9001.5 hours in MS Paint related
>>
>>1307813
>vinyl tubing on the outlet
gross
just use glass
>>
>>1307823
consider it a consumable
>>
>>1307824
i meant gross as in your liquor will taste like PVC
>>
>>1307813
First of all, that pic is pure art.
Thanks again for the help

>>1307826
glass may be a good idea, I'll consider it.
>>
>>1307705
Don't be a fool. Use parastaltics and steppers. Theres not much else out there that makes the project food safe and controllable.

Google bartendro.
>>
>>1307850
Wel peristaltics cost ~6$ per pump without steppers, far more with, they are hard to clean, they have very low volume output in comparison.

If I'm willing to sacrifice accuracy for speed and cost, all other options seem better. Even the Diaphragm pumps.
>>
>>1307850
also bartendro is over engineererd imo. It's the basic "If we simply put a lot of peristaltik pumps in a row it'll work" approach.
While technically true, I think there's better ways to design a drink dispenser. I have other needs from mine then the bartendro people had for theirs, so the design approaches are different.
>>
Here you go OP
https://www.youtube.com/watch?v=Z7GkGeZrb2Y
>>
>>1307851
Man if you can find a food safe pump that can dish out 0.75 to 1.25 Oz of fluid a squirt and be food safe at a low price point all the power to ya.
>>
>>1307705

>ctrl-f: "peristaltic"
>1 match

Really?

Peristaltic pumps are, as far as I'm concerned, the ONLY way to go if you insist on a pump. They're accurate, pretty inexpensive, and sanitary. They are kind of slow for their size, though.

Centrifugal pumps won't dispense reliable amounts without a meter on their output. Piston pumps need lubrication. Diaphragm pumps could work reasonably well, but you're going to struggle finding one with all food-safe and alcohol-resistant parts, at least for cheap. An air pump by itself isn't workable because of the compressibility of air causing issues with dispensing; you'd need to compensate for the variable volume of liquid in the bottle, dump the excess pressure when you're done dispensing, you'd need a filter for the incoming air and, like with a centrifugal pump, you're going to struggle to even ballpark the measurements without some kind of feedback.

Every pump-using bar-bot project I've seen has used peristaltic pumps. I'm sure there's at least one out there that doesn't, but there's good reason to do so.
>>
>>1307859
Thanks, but I've seen that already.
>>1307867
Well, I'm trying something different, that's why I thought about the Diaphragm pump.
Admittadly I forgot about the problems arising with air, but if I keep a constant pressure, wouldn't it be irrelevant how much air there is in the tank? If the pump creates a pressure of X and the dispensing clears a room of Y then the pump must just create enough pressure to get back to X again, air or not...
Again I'm okay with reasonable inconsistencies. If there's a few ml more or less, so be it.

If I just wanted to go the safe route, obviously peristaltic pumps would work. The thing is that I want to change stff though, I'm not interested in just following instructions and not improving upon the concept (or trying to and failing)

Thanks for the input, it's great to get so many people point out stuff that I didn't think about
>>
>>1307729
There are fairly simple, purely mechanical bar drinks dispensers that pour a single measure every time they're pushed. If you replicate that system a solenoid would be great.

You could also consider miniature bell siphons.
>>
I see all these robot bartenders and wonder what the purpose is. You're not saving any money in the long run because you need to build it and program it and clean it once in a while. You can't make something fancy because all it's doing is pouring different liquors into a glass. It's not even that much faster than doing it by hand because you have to wait for the booze to get pumped out of the bottle. Can someone enlighten me because this just seems like a silly project for no real benefit.
>>
>>1307872
>Admittadly I forgot about the problems arising with air, but if I keep a constant pressure, wouldn't it be irrelevant how much air there is in the tank?

If the valve is on the OUTLET, you should be able to get some modicum of consistency. But then you need to find a valve suitable for food contact, which may make up the difference in savings. And you still need to filter the air.
>>
>>1307901
>There are fairly simple, purely mechanical bar drinks dispensers that pour a single measure every time they're pushed. If you replicate that system a solenoid would be great.

I've seen a project that does something like that, it was gravity fed. Truth be told, I like it a lot but I don't know how to replicate it without 2 solenoid valves per liquid (valve one to fill minimal liquid increment container , valve 2 when 1 is closed to dispense it.)

>>1307904
Honestly, it's worse tan just doing it by hand. By hand you have flexibility and speed on your side. Only possible pro I see is how exact you can be.
I do it because building and designing it is fun. I want it to look good in the end and be something in the flat that is fun and that people think looks cool. I build a lot of electronics because they are fun. Miniature radios, prank circuits, etc.
Some have use, some don't. It's the fun that matters.
>>1307914
Outlet is what I thought, otherwise it's pretty impossible to achieve that (as you said already). The filtering I didn't consider, but it I don't pump the pressurized air in to the liquid but just into the top of the bottle, wouldn't air cleaned of particles be enough? Maybe a carbon filter + that stuff you use for Vacuums maybe...
>>
>>1307714
Fucking hell why does this guy show up in every single thread with this lmgtfy shit post. Could we just auto-ban people who post this?
>>
File: 1500590415384.jpg (202 KB, 628x1000)
202 KB
202 KB JPG
>>1307872
the compressibility of the air should be the same at all bottle levels. if your solenoid valve and stopper are airtight, the pressure in the vessel will always be determined by the height of your dispensing tube.
calibration (mL/sec) would probably vary slightly per material being dispensed. vodka vs. simple syrup, for example.
field-proven: this is basically the principle of Pic related.
>>
>>1307926
>The filtering I didn't consider, but it I don't pump the pressurized air in to the liquid but just into the top of the bottle, wouldn't air cleaned of particles be enough?

If it's going into the bottle, it's going to settle into the liquid eventually. It would be a good idea to filter it, doubly so if the bottles are meant to remain attached for long periods of time.

Not sure how well carbon filters handle airborne pathogens, but a sub-micron particulate filter would be advisable at the bare minimum.
>>
>>1307956
Most of these threads are made by literal retards where everything can be answered by a quick google search you fucking retard.
>>
>>1307956
http://lmgtfy.com/?q=why+does+this+one+retard+show+up+in+every+thread
>>
>>1307966
Yeah, that's what I thought. Thanks!
>>1307978
Thanks for the advice again, I'll consider it!
>>1307980
Well yeah but to be fair, if you googled it and read up on it then you should still be able to ask. The lmgtfy doesn't add to the discussion and the board is slow already so unless it's someone who asks how do I... Without having even a slight idea, frankly it's unnecessary. On the other Hand they don't really prevent the topics from being posted, in the contrary, they even bumb the younger threads
>>
>>1307978
>airborne pathogens
>alcohol
not really seeing a great hazard here
in any case, sanitary air filters used by the home brewing community that mate with typical tubing sizes are about $3, and industrial versions for much higher flow rates could be had for about $10

>>1307980
why do you hate conversation so much
>>
Anyone in this thread ever work in a bar?
>>
>>1307705
You could try to reinvent the wheel or look up Party Robotics and see what they've done. They've made all the rookie mistakes (like gravity feed) so that you don't have to.
>>
>>1308293
If you read the thread OP want to break the mould.. Probably with mold.
>>
>>1308293
While they are a cool resource, they were mentioned already. Following their way also leads to a convenient 3700$ solution. Yeah it's good but that's not why I'm doing this.
>>1308295
Exactly. I may fail, it may be a waste of time, but I'm happy to try new stuff.
>>1308277
No, but I know a few who did. Why? I'm fully aware that this doesn't solve any problems that exist, contrarily it likely creates problems.
>>
OP stop trying to be 100% original and turbo cheap or you'll never get your project done.
>>
>>1307705
ROBOT ARM
ROBOT ARM
6 DEGREES OF FREEDOM SEXBOT ARM
>>
File: Pump Daigram Color.jpg (190 KB, 746x662)
190 KB
190 KB JPG
>>1307705
Why not use a peristaltic pump?

Pros
>Low pressure
>No machinery needs to be in contact with liquid
>simple to maintain

Cons
>Loud
>Expensive
>Vinyl tubing is really the only effective medium. May impart a taste into your drinks.
>>
>>1308476
why not use silicone tubing?
>>
>>1307705
MAKE A DRINK ROBOT, THAT'S ALSO A BOUNCER, BEST NIGHTCLUB EVER
>>
>>1308423
Well I'm trying to stay within budget (which is pretty low, I know) and I want to experiment. The designing, planning and building are as much if not more fun than the final product is supposed to be.
I will get on with it though, I get what you're saying.
>>1308476
Mainly price point and serving speed.
I get that the advantages are immense for certain liquids (Bailey's or similar would wreck any other pump over time), but they loose when other Liquids are served.
>>
>>1308558
The poster might just be mistaken. You can definitely use silicone tubing.
>>
File: blue screen.png (2 KB, 640x400)
2 KB
2 KB PNG
Questioooonnnn:

Do you know what code will you use for the valve?

Because you need to declare variables for the timing and order of the drinks

I recommend using Python (it's easy to learn and Arduino)
>>
>>1309479
>python
>arduino
I'm not going to call you a scrub, but I didn't think it could be done.
>>
this is the IoT i totally approve - Aclomat, https://youtu.be/XmebXgXmXsM

details here: http://geektimes.ru/post/265722/


Guy also made shashlik machine
>>
>>1309479
I will actually use Python because I'm programming on the Raspberry pi and python for rpi is pretty good.
I would use C++ for Arduino (because of the IDE)

From what I talked about a Rpi would be overkill so far, a arduino mini would totally do the job. I'm using a Rpi though because it should later be integrated in the home network, people in a party should be able to use it. Also it'll be a music station, essentially you can connect to it, play music via it which is then fed to the speakers in the flat or you can order a drink.
>>
>>1310024
Like to see you make a decent gui with arduino.

Oh wait....
>>
>>1310025
>Like to see you make a decent gui with arduino.
Where in my post did I even say I'd create a UI with arduino?
The UI, IOT integration and Music player integration is WHY I'm using Rpi
>>
>>1309738
That'll be somewhat what I'm going for, but not with buttons but with a network integration
>>
>>1310030
Called an esp you pi sucking faggot.
>>
>>1310037
I already have a Pi, it was unused for quite some time except for a pihole. I want to use it for this becase it does what I need it to.
If you have better ideas, use of ressources, go for it, tell me.
>>
>>1308293
>all those nice-ass Welco pumps
geez that must cost a lot.
>>
>>1310068
That's why their Bartendro 7 costs 2499,99$ and their Bartendro 15 costs 3699,99$...
>>
File: image.jpg (18 KB, 378x378)
18 KB
18 KB JPG
Why not put a bunch of these in your frame and go from there, no need for pumps, food safe, easy cleaning and swapping of bottles. Either funnel them all to your glass or have the glass move underneath.
>>
>>1310091
There are pro's and cons to this, the inebriator does exactly that. For normal drinks I considered that because it's incredibly fast in comparison. Of course microajustments to amount can't be made anymore, but then again that's kinda a non issue because personally as long as it fits somewhat I'm okay with it. I'll likely combine the systems. At least that's what I'm going for right now.
>>
>>1310100
You can still do micro adjustments by measuring the flow rate for each liquor you will be using, should be fairly consistent since it is always the same amount from the same sized container.

Then just calculate the time to activate for the desired amount of that liqour. Just be sure to allow for refill time if you need more than 1 shot of that liqour.
>>
>>1310113
I actually didn't think about that. That kinda solves a issue I had. I'll have to stop a certain order...

Yeah it won't be accurate, but who cares as long as it is acceptably inaccurate
>>
>>1310125
The ones I worked with in a bar would allow 2cl to fill in, when activated they would start draining untill you stopped pushing up. When you fully let go the container will refill itself till 2cl.

With a simple cam for your Pie you could even read out the measure on the container, you would be about as accurate as your average barman.
>>
>>1310050
Already did. Read between the lines.

Keep the controller seperate. Use arduino or whatever the fuck, but set the limits outside the pi. This can be as simple as hard-wired limits as well.

You obviously never crashed a pi but when you do I hope you have a big cup.
>>
>>1310148
don't try to run your SBC off microUSB phone chargers and maybe you won't crash it
>>
>>1310215
Even a 3.5 amp source dont mean a pi will respond in time nor be stable, especially when developing your own code.

This is why I use pis only for hmi purposes.
>>
Google "Shurflo Brix Pump".




Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.