Pressure Sensors? A little more info

Archiver

well this will be the 3rd time i try this, my email
does not seem to be working properly, here goes:

i hope this explins a little better.
overall problem: track volume of several hoppers of
different fluids to accurately pour out ounces.
Accuracy ~ some fraction of an ounce, volume ~2
gallons
Density = variable

i am buildig a robotic bartender that makes drinks by
looking up the mix from an EEPROM and then activates
the appropriate solenoid valves to pour out the proper
amounts. I am trying to come up with a way to
properly pour say 1 oz out of each hopper of fluid.
The problem is that the velocity flow rate is
linear(downward slope), not constant. So I need a way
to track how much is in each hopper to open the valve
for the correct amount of time, to properly match the
repective flowrate to volume that changes as the fluid
is poured out. Someone else mentioned that I need the
density, what I planned to do for that was measure
several types of say gin and average them. Anyway, I
examined several ways to measure the volume via
capacitance, weight, etc., but the best way both
functionally and economically, I thought at least, was
to indirectly measure it by finding the pressure from
a custom made manometer. see crude drawing below.
___ _ = sealed top
| | |x| x = pressure sensor -> Stamp
| | | |
|^^^^^| |^| ^ = fluid level
| | | |
| | | |
| | | |
| |__| |
|_ ______|
| |
| |
[noparse][[/noparse]x] [noparse][[/noparse]x]= solenoid valve(gravity feed)
| |
there will be at least 12 of these devices.

I believe that when fluid is placed in the large
hopper , it will flow into the small hopper joined to
it (3" PVC tube and 1" PVC tube) and because the
column of air is trapped in the small tube the
pressure will increase, and therefore I should be able
to fit a curve and get the equation of a line for
pressure vs volume regardless of liquid. If it was a
constant flow I could just use counters for each
hopper and open the valves for some constant time, but
I guess that is the ultimate problem at hand. Maybe I
have made this harder than it really is, ideally a
small pump would work the best, but I am trying to
minimize my costs, as most people do. I was hoping to
get the sensors for a couple 3 bucks maybe, but I
guess I'm dreaming. Maybe some approximations or
assumptions I should make,anyway,if anyone wants to
give input to my problem, I am more than happy to
listen. And thank you to those that have given input
already.
Ross Cooper

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/

Archiver

Hi Ross,

I've not been following this thread closely, so forgive me if this is
irrelevant.....................

I designed a system to dispense 500 ml of distilled water over and over. It
was based on a mariotte siphon

The accuracy was very good + - 1 or 2 ml.

I used a timer to control the open time of the valve.

Here's a link to show you how it would work

http://www.uswcl.ars.ag.gov/exper/mariotte.htm

-Steve




At 05:55 PM 4/17/02 -0700, you wrote:
>well this will be the 3rd time i try this, my email
>does not seem to be working properly, here goes:
>
>i hope this explins a little better.
>overall problem: track volume of several hoppers of
>different fluids to accurately pour out ounces.
>Accuracy ~ some fraction of an ounce, volume ~2
>gallons
>Density = variable
>
>i am buildig a robotic bartender that makes drinks by
>looking up the mix from an EEPROM and then activates
>the appropriate solenoid valves to pour out the proper
>amounts. I am trying to come up with a way to
>properly pour say 1 oz out of each hopper of fluid.
>The problem is that the velocity flow rate is
>linear(downward slope), not constant. So I need a way
>to track how much is in each hopper to open the valve
>for the correct amount of time, to properly match the
>repective flowrate to volume that changes as the fluid
>is poured out. Someone else mentioned that I need the
>density, what I planned to do for that was measure
>several types of say gin and average them. Anyway, I
>examined several ways to measure the volume via
>capacitance, weight, etc., but the best way both
>functionally and economically, I thought at least, was
>to indirectly measure it by finding the pressure from
>a custom made manometer. see crude drawing below.
> ___ _ = sealed top
> | | |x| x = pressure sensor -> Stamp
> | | | |
> |^^^^^| |^| ^ = fluid level
> | | | |
> | | | |
> | | | |
> | |__| |
> |_ ______|
> | |
> | |
> [noparse][[/noparse]x] [noparse][[/noparse]x]= solenoid valve(gravity feed)
> | |
>there will be at least 12 of these devices.
>
>I believe that when fluid is placed in the large
>hopper , it will flow into the small hopper joined to
>it (3" PVC tube and 1" PVC tube) and because the
>column of air is trapped in the small tube the
>pressure will increase, and therefore I should be able
>to fit a curve and get the equation of a line for
>pressure vs volume regardless of liquid. If it was a
>constant flow I could just use counters for each
>hopper and open the valves for some constant time, but
>I guess that is the ultimate problem at hand. Maybe I
>have made this harder than it really is, ideally a
>small pump would work the best, but I am trying to
>minimize my costs, as most people do. I was hoping to
>get the sensors for a couple 3 bucks maybe, but I
>guess I'm dreaming. Maybe some approximations or
>assumptions I should make,anyway,if anyone wants to
>give input to my problem, I am more than happy to
>listen. And thank you to those that have given input
>already.
>Ross Cooper
>
>__________________________________________________
>Do You Yahoo!?
>Yahoo! Tax Center - online filing with TurboTax
>http://taxes.yahoo.com/
>
>To UNSUBSCRIBE, just send mail to:
> basicstamps-unsubscribe@yahoogroups.com
>from the same email address that you subscribed. Text in the Subject and
>Body of the message will be ignored.
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

Archiver

Here's another tack:

Peristaltic chemical metering pumps may do the trick. They dispense liquids
accurately enough with excellent repeatability, as long as they are kept
primed. In a water purification system I did that creates chlorine dioxide
by injecting a pair of liquids into a water line, I was able to provide s/w
selection of the mix ratio. The only catch is that you'd need one pump per
drink item - which may or may not fit your budget. There are off-the-shelf
pumps that will allow control via proportional control or timed run.

Chris
>
> well this will be the 3rd time i try this, my email
> does not seem to be working properly, here goes:
>
> i hope this explins a little better.
> overall problem: track volume of several hoppers of
> different fluids to accurately pour out ounces.
> Accuracy ~ some fraction of an ounce, volume ~2
> gallons
> Density = variable
>
> i am buildig a robotic bartender that makes drinks by
> looking up the mix from an EEPROM and then activates
> the appropriate solenoid valves to pour out the proper
> amounts. I am trying to come up with a way to
> properly pour say 1 oz out of each hopper of fluid.
> The problem is .....
....if anyone wants to
> give input to my problem, I am more than happy to
> listen. And thank you to those that have given input
> already.
> Ross Cooper
>
>

Archiver

maybe im just lazy but i would have a little trial and
error...rather than doing all this math to calculate densities and
all that, just have a solenoid that opens when +5V hits it and turns
off when the +5V is off. do some tests where from a gravity drop
from a container of vodka might take 3 seconds do dispense 1 ounce,
then just do simple stamp command like:

high 1 'the solenoid for vodka opens, 5 volts turns it on
pause 3000 ' pause for 3 seconds, stays open for 3 seconds
low 1 'solenoid is turned off, vodka stops flowing


that would be one ounce of vodka, 3 lines of code. for lets say
gin there might be a different time needed for one ounce, like 3.5
seconds...change the pause command to PAUSE 3500, trial and error
until you get it right....once you have it you are set forever.

orange juice, kinda slow maybe 5 seconds?

just my thought..... one stamp could control 16 solenoids because of
the 16 outputs, if the stamps 5V 20ma supply isnt enough, just use a
transistor, a tip 29 will be PLENTY, probably overkill.....try a
2N3904 for more light duty.

this in my opinion will be somewhat accurate, although i dont drink
that much so i dont know if people really notice how much vodka they
are getting anyway...especially computer controlled...has to be
somewhat accurate.

cool project by the way....

steve

optimistic@s...








--- In basicstamps@y..., Klutch <klutch66@y...> wrote:
> well this will be the 3rd time i try this, my email
> does not seem to be working properly, here goes:
>
> i hope this explins a little better.
> overall problem: track volume of several hoppers of
> different fluids to accurately pour out ounces.
> Accuracy ~ some fraction of an ounce, volume ~2
> gallons
> Density = variable
>
> i am buildig a robotic bartender that makes drinks by
> looking up the mix from an EEPROM and then activates
> the appropriate solenoid valves to pour out the proper
> amounts. I am trying to come up with a way to
> properly pour say 1 oz out of each hopper of fluid.
> The problem is that the velocity flow rate is
> linear(downward slope), not constant. So I need a way
> to track how much is in each hopper to open the valve
> for the correct amount of time, to properly match the
> repective flowrate to volume that changes as the fluid
> is poured out. Someone else mentioned that I need the
> density, what I planned to do for that was measure
> several types of say gin and average them. Anyway, I
> examined several ways to measure the volume via
> capacitance, weight, etc., but the best way both
> functionally and economically, I thought at least, was
> to indirectly measure it by finding the pressure from
> a custom made manometer. see crude drawing below.
> ___ _ = sealed top
> | | |x| x = pressure sensor -> Stamp
> | | | |
> |^^^^^| |^| ^ = fluid level
> | | | |
> | | | |
> | | | |
> | |__| |
> |_ ______|
> | |
> | |
> [noparse][[/noparse]x] [noparse][[/noparse]x]= solenoid valve(gravity feed)
> | |
> there will be at least 12 of these devices.
>
> I believe that when fluid is placed in the large
> hopper , it will flow into the small hopper joined to
> it (3" PVC tube and 1" PVC tube) and because the
> column of air is trapped in the small tube the
> pressure will increase, and therefore I should be able
> to fit a curve and get the equation of a line for
> pressure vs volume regardless of liquid. If it was a
> constant flow I could just use counters for each
> hopper and open the valves for some constant time, but
> I guess that is the ultimate problem at hand. Maybe I
> have made this harder than it really is, ideally a
> small pump would work the best, but I am trying to
> minimize my costs, as most people do. I was hoping to
> get the sensors for a couple 3 bucks maybe, but I
> guess I'm dreaming. Maybe some approximations or
> assumptions I should make,anyway,if anyone wants to
> give input to my problem, I am more than happy to
> listen. And thank you to those that have given input
> already.
> Ross Cooper
>
> __________________________________________________
> Do You Yahoo!?
> Yahoo! Tax Center - online filing with TurboTax
> http://taxes.yahoo.com/

Archiver

Why not push a known volume of air into the bottle -- ideally it will
displace the same amount of fluid.

---

Original Message

---

> maybe im just lazy but i would have a little trial and
> error...rather than doing all this math to calculate densities and
> all that, just have a solenoid that opens when +5V hits it and turns
> off when the +5V is off. do some tests where from a gravity drop
> from a container of vodka might take 3 seconds do dispense 1 ounce,
> then just do simple stamp command like:
>
> high 1 'the solenoid for vodka opens, 5 volts turns it on
> pause 3000 ' pause for 3 seconds, stays open for 3 seconds
> low 1 'solenoid is turned off, vodka stops flowing