BarMixVah Plus

Source now staged:

https://github.com/marktoml/BarBot

It is more or less the same as pictured, some minor tweaks.
Is working fine at this stage, if you run into issues log them on github and I'll fix

Things we have on the 'plate' for this:
1)enhance the UI as discussed for more control/options
2) put the recipes in a SQL db on the Pi (one of the small ones like SQLLite)
3) allow drinks to be 'queued' then an entire order can be dispensed
4) add a turn-table for dispensing multiple drinks (with sensors to know when a glass is present and empty or not)
5) Android phone app to control it (already started spelunking with that)
6) lastly an Alexa skill to control it (because telling Alexa to have the BarBot pour me a drink from anywhere is cool)

 

An example properties file is included that demos the recipe list. Eventually I will move this to an actual db, but for now it functions. The format is this:

motor2pin=GPIO 22
motor5pin=GPIO 25
motor1pin=GPIO 21
motor4pin=GPIO 24
motor3pin=GPIO 23
ingrediant1=Scotch
ingrediant2=Soda
ingrediant3=Gin
ingrediant4=Vodka
ingrediant5=Tonic
recipe1=Scotch N Soda:Scotch;1.5:Soda;1.5
recipe2=Vodka N Tonic:Vodka;1.5:Tonic;1.5
recipe3:Gin N Tomic:Gin;1.5:Tonic;1.5

The motor pin definitions are needed to know which GPIO pin the motor is attached to.
The ingrediant list matches the motor number (so ingredient 1 is on motor 1).

The recipes are just sequentially numbered and broken down like this:

NAME:Ingredient1;ounces:Ingredient2;ounces

You can carry on with as many ingredients as you have motors to support.
NAME is what shows up in the list box on the UI.

The program matches the ingredient on the recipe to the ingredient on the list of available ones (so spelling matters)

 

When you deploy this to the Pi the properties file needs to be in the "default" directory.
Some experimentation may be needed

That may be the "home" directory for the user (/home/pi) if you have it auto-launch at boot time or the home directory for the project if you are debugging it (for example -- /home/pi/NetBeansProjects/BarBot)

Naturally if you have questions, ping me. I am getting ready to expand the hardware for grins (7 motors). Sadly we can't go unlimited as the screen eats up some of the GPIO port which leaves only about 15 motor pins available.

 

Committed a small fix where we were calling the Motor I/O via Pi4J synchronously rather than async. SO while your ingredients COULD be mixed at the same time, they were not. This made mixing a given drink a lot longer that needed*.

The threads now run in parallel (or as parallel as java supports).

(change pushed to github)


*totally NOT what you want.

 

There have been a few recent bug fixes and I still have a couple of rare/odd ones being chased. I will update everything in the GITHUB site rather than keep this one up-to-date.

 

For completeness this is the circuit between the Pi and the motor
Sorry, resistor values a bit obscure (blame Eagle) 2.2 kohms and 470 ohm
We also added a diode on the output to the motor just so visually you can see each one is "on"

I was concerned since the Pi uses CMOS (3.3v) logic, but the way the TIP120 is used in this there is more than enough for it to trigger (roughly 2v is enough forward bias to get it wide open collector-emitter). So a non-issue.



Other parts here:


20pcs 1N4007 1A 1000V Through Hole Axial Rectifier Blocking Diode
Sold by: uxcell Product question? Ask Seller
$5.70


E-Projects 100EP514470R 470 Ohm Resistors, 1/4 W, 5% (Pack of 100)
Sold by: E-Projects, LLC
$5.16


TIP120 NPN Power Darlington Transistors TO-220 10 Pack
Sold by: ELITE.CITI
$7.99


E-Projects 100EP5142K20 2.2kOhm Resistors, 1/4 W, 5% (Pack of 100)
Sold by: E-Projects, LLC
$5.16

Peristaltic pumps:
https://www.amazon.com/gp/product/B00F9MXFFQ

And we just used breadboard to build it on (a little hot glue when done to keep everything in place)

 

Today I would like to make a brief apology
I was adding two more motors/ingredients to the rig and when I went to modify the code I realized that my son had used 4 different variables that were all named some flavor of "motor" -- not confusing at all. So, added the motors and code to support them and cleaned it up a bit before checking it back in....

 

So I had to recreate this image again. I could have sworn I made a backup and apparently I would have been wrong. Oh well, just needs Raspian and the 5" LCD driver and the BarBot software.... I think this must have been directed at me:

https://www.tvfanatic.com/quotes/im-thinking-you-werent-burdened-with-an-over-abundance-of-school/

(I swear by my pretty floral bonnet I am making a copy this time)

 

I love Firefly.

 

Aw, now see ... you just made me have to go back and mess with it again

 

NOW I has a back up
I also gave up trying to flip the screen and the touch screen and just have it mounted the other way. Not a big deal.


While adding a IceMaker is an idea, I can add an Alex skillset for this and have it mix drinks on-demand via the Echo. THAT is on the list ...

"Alexa, POUR ME A DOUBLE..."

 

config.txt

display_rotate = n

0 = Default
1 = 90°
2 = 180°
3 = 270°
0x10000 = flip horizontally
0x20000 = flip vertically​

 

That flips the screen but not the touch screen part of it.
So then the touch is 180 degrees away from where you want it.

 

/etc/modules

Change swap_xy=0 to swap_xy=1

 

Cool, I will give that a whirl, thanks!

 

OK, the CONFIG entry is the same as what I use, but that as before will leave the touchscreen part unreversed

The /etc/modules has nothing remotely resembling that...

I did find this:

https://www.raspberrypi.org/forums/viewtopic.php?f=108&t=120793

which talks about using XINPUT and I will try that.

 

pi@BARBOT-PI:/etc $ xinput --list
⎡ Virtual core pointer id=2 [master pointer (3)]
⎜ ↳ Virtual core XTEST pointer id=4 [slave pointer (2)]
⎜ ↳ ADS7846 Touchscreen id=6 [slave pointer (2)]
⎜ ↳ MemsArt MA144 RF Controller id=8 [slave pointer (2)]
⎜ ↳ MemsArt MA144 RF Controller id=9 [slave pointer (2)]
⎣ Virtual core keyboard id=3 [master keyboard (2)]
↳ Virtual core XTEST keyboard id=5 [slave keyboard (3)]
↳ MemsArt MA144 RF Controller id=7 [slave keyboard (3)]

So that driver is not responding to the axis swap commands. I will need to dig into it a bit further. Reverted it all back for now.

 

Yep, it works as-is. Only discussing the screen orientation now.
Pumps are not touching the beverages... only the tubing.