Kegbot - PowerPoint PPT Presentation
1 / 23
Title:
Kegbot
Description:
Beer flow from Keg is 100-140 Oz (.8-1.1 G) PM. Low PSI. Cv factor = GPM at 1PSI ... Minor supply shortfall : the beer guy. Major project ruinage: TekScan ... – PowerPoint PPT presentation
Number of Views:48
Avg rating:3.0/5.0
Title: Kegbot
1
Kegbot
- Steven Gangstead
2
Project Idea
- Im a senior
- Project involving beer
- Automatically pour one beer (ala McDonalds coke
machine)
3
Original Goals
4
Original goals
- Flexible on cup size as long as same volume
- Dont want to spill any
- Want Portable, Durable, Cleanable
- Standard Keg hookup
5
Goals added along the way
- Want to have a top off button
- Automatic cut off (if weight drops to zero, stop
pouring) - LEDs because blinking is cool
6
Technical
- Used 6811 processor because I didnt want any
problems (see Meng) and processor requirements
were low - Used a Solenoid valve to control flow
- Piezo-resistive weight sensors
- Lots of parts from home depot
7
Solenoid Valve SV3311
- Beer flow from Keg is 100-140 Oz (.8-1.1 G) PM.
Low PSI - Cv factor GPM at 1PSI
- Cv.63 ¼ thread
- Brass, Stainless Steel, Copper materials
8
(No Transcript)
9
Flexi Force Sensors
- Sound cool, only 8 mil thick
- Piezoresistive force varies according to load
- Well, well get back to these later
10
Final project
- FLOP
- Forces bigger than me conspired to ruin my
project - Minor supply shortfall the beer guy
- Major project ruinage TekScan
11
Thoughts on TekScan
12
FlexiForce
- Sensors sound good on paper
- 5 linearity
- /- 2.5 repeatability
- WRONG
13
FlexiForce
14
FlexiForce
- System not linear
- Weight fluctuates - always
- Hysteresis is rediculous making me hysterical
- Weights didnt add linearly
15
What went right
- Software works
- Instantaneous pour, Cup pour (just gets bad
info), cutoff - Awesome arcade style buttons
- Cool Plexiglas construction
- Cleanable (beer contact area all plastic)
16
What went wrong
- Weight sensors
- Cup size not really flexible (had to immobilize)
- Spillage (will come back to)
- Not really portable
- No LEDs
- Didnt get to test with beer
17
My Fonzerelli Fix
Partial Salvage
18
Fonz fix
- Weights didnt add linearly so just use one
sensor - Picked the sensor that fluctuated least by weight
- Weight still all over the place (cant just add
20 samples to it or anything) - Readings still inconsistent. But
19
Fonz fix. Ehhhhhhh!
- Noticed relationship between full cup weight and
starting weight - MOST of the time finishing weight was 2X starting
weight. Must be a huge coincidence but Ill take
it.
20
Results
- Did 40 trials
- 30 of them gave me between 2/3 and full cup (full
one drop from spilling over) - 8 overflows (had to use cutoff button)
- 2 underflows (only a little bit in the cup)
- 75 success from horrible sensors
21
Rev 2.0
- Dont use development board, Axiom has smaller
versions. - Possibly even use a different processor (program
is very small and would be easy to port) - DONT USE TEKSCAN PRODUCTS
- EVER
22
Rev 2.0
- Either get better weight sensors (with no
circuitry to build) or flow meters - Flow meter would be nice could write all in
software, but they are on a scale
23
The End
Fin
Recommended
CrystalGraphics Presentations
