Searching \ for '[EE]: Hopefully, someone can answer this one!' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=hopefully+someone
Search entire site for: 'Hopefully, someone can answer this one!'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Hopefully, someone can answer this one!'
2002\03\03@045236 by Kevin Maciunas

flavicon
picon face
Dear all,
       Herewith the problem.  I want to sense the level(*) of water in a
sealed aluminium tank.  The tank can have a hole or two bored into the
top, but not the sides.  Access into the tank is basically impossible.
The whole shebang is (rather nicely) welded shut.

       (*) In actual fact, I don't really want the level - I want the volume
of water.  The tank is mounted on a fire-appliance and at present we
have four float switches on the back of the rectangular tank.  Very
spiffy when the appliance is parked nose up on a steep hillside!  This
can get steep enough that it shows 4/4 when actually there is about 1/4
of a tank full.

       With a few depth probes and some simple math, a PIC should be able to
give a quick at-a-glance water remaining (and probably water use per
unit time, etc etc).  Any ideas anyone?  I apologise in advance if this
is a real simple "taught in every Mech/EE course" but I'm a Computer
Scientist, not an EE type :-)

/Kevin
--
Kevin J. Maciunas              Net: spam_OUTkevinTakeThisOuTspamcs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@050934 by Kathy Quinlan

flavicon
face
----- Original Message -----
From: "Kevin Maciunas" <.....kevinKILLspamspam@spam@CS.ADELAIDE.EDU.AU>
Subject: [EE]: Hopefully, someone can answer this one!


> Dear all,
>         Herewith the problem.  I want to sense the level(*) of water in a
> sealed aluminium tank.  The tank can have a hole or two bored into the
> top, but not the sides.  Access into the tank is basically impossible.
> The whole shebang is (rather nicely) welded shut.

Ok I take it you know what the tank holds ? and *IF* they are like the WA
tanks and operators ;o) they open the hole on top (inspection hatch) dump
the fill hose in and just fill till it over flows ?

If so, just build a restable PIC flow controller IE measure the water out
and subtract that from the water in :o)

I do not know any accurate (within 10%) methods of determining tank volume
cheaply (the expensive way is to sit it on a large load cell and measure the
weight (water does not vary much in weight per L )) So I think flow out is
the easiest way </me shrugs>

Regards,

Kat.

____________________________________________________________________________
/"\   ASCII Ribbon Campaign  |        K.A.Q. Electronics
\ / - NO HTML/RTF in e-mail | Software and Electronic Engineering
X  - NO Word docs in e-mail  |      Perth Western Australia
/ \                                            |        Ph +61 419 923 731
____________________________________________________________________________

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@052215 by Kevin Maciunas

flavicon
picon face
On Sun, 2002-03-03 at 20:37, Kathy Quinlan wrote:
> ----- Original Message -----
> From: "Kevin Maciunas" <kevinspamKILLspamCS.ADELAIDE.EDU.AU>
> Subject: [EE]: Hopefully, someone can answer this one!
>
>
> > Dear all,
> >         Herewith the problem.  I want to sense the level(*) of water in a
> > sealed aluminium tank.  The tank can have a hole or two bored into the
> > top, but not the sides.  Access into the tank is basically impossible.
> > The whole shebang is (rather nicely) welded shut.
>
> Ok I take it you know what the tank holds ? and *IF* they are like the WA
> tanks and operators ;o) they open the hole on top (inspection hatch) dump
> the fill hose in and just fill till it over flows ?
>

Yes, and that's what we do too.  The touble is, you can work for quite a
while on 3,000litres of water.  So people tend to forget..

> If so, just build a restable PIC flow controller IE measure the water out
> and subtract that from the water in :o)
>

That would imply flow measurement on the outfall of the tank.  Thought
of that already - problem is, the pump recirculates water to keep the
volute case cool.  The volume recirculated is variable.  We also pump
from other sources (and maybe put some of THAT into the tank too..).

> I do not know any accurate (within 10%) methods of determining tank volume
> cheaply (the expensive way is to sit it on a large load cell and measure the
> weight (water does not vary much in weight per L )) So I think flow out is
> the easiest way </me shrugs>
>

Weighing the tank would, of course, work.  I was hoping for a simple and
cheap solution :-).  Hey, I can always wish!

/Kevin
--
Kevin J. Maciunas              Net: .....kevinKILLspamspam.....cs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@055820 by Kathy Quinlan

flavicon
face
----- Original Message -----
From: "Kevin Maciunas" <EraseMEkevinspam_OUTspamTakeThisOuTCS.ADELAIDE.EDU.AU>
Subject: Re: [EE]: Hopefully, someone can answer this one!

> Weighing the tank would, of course, work.  I was hoping for a simple and
> cheap solution :-).  Hey, I can always wish!

I just had a brain wave :o) (Coke must be kicking in ;o)

When in hospital they measure the volume of urine in the bladder with
ultrasonics (and it is quite accurate considering the bladder is not a
regular shape and its shape depends on its volume)

do a google and see what you can find on "ultrasonic measurement" also try
it with "volume" or "capacity" etc :o)

Regards,

Kat.

____________________________________________________________________________
/"\   ASCII Ribbon Campaign  |        K.A.Q. Electronics
\ / - NO HTML/RTF in e-mail | Software and Electronic Engineering
X  - NO Word docs in e-mail  |      Perth Western Australia
/ \                                            |        Ph +61 419 923 731
____________________________________________________________________________

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@061202 by Christoph Graf

flavicon
face
Kat

What about measuring the distance top of tank - water level ???
I don'r know how big the tank is, but I could imagine infrared
or ultrasonic.

Regards Chris

{Original Message removed}

2002\03\03@062242 by sambuddy

flavicon
face
Ok,
       I've thought about it deeply for 2-3 seconds and I'm making the following assumption, it's a water tank on the back of a fire truck. What I would look into is
putting a preasure sensor on the tank, I would then feed the exhaust from the truck engine into the tank, taking note of the trucks engine capacity and RPM therefore
knowing how much exhaust is being pumped into the tank. I would do this until the tank reached a certain preasure (it wouldn't have to be much). If you know how
much exhaust it took to get the tank to a certain preasure while it was empty you should be able to work it out from there. e.g. 10000 cubic feet of air to reach the
preasure while it was empty, if you pump in 5000 cubic feet, then it should be half full. To do the measurement again, release the preasure and start again,( it should
be a fairly quick process if you want it done quicker, rev the engine up).
Hope this helps

Regards
Stuart O'Reilly


3/03/02 8:53:10 PM, Kevin Maciunas <kevinspamspam_OUTCS.ADELAIDE.EDU.AU> wrote:

{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@063731 by Kevin Maciunas

flavicon
picon face
On Sun, 2002-03-03 at 21:36, Christoph Graf wrote:
> Kat
>
> What about measuring the distance top of tank - water level ???
> I don'r know how big the tank is, but I could imagine infrared
> or ultrasonic.
>
That was my naïve thought initially - just measure down from the top to
the water level.  The tank is about 1200mm high (four feet).  Any sensor
would need to deal with being very wet, of course.  The water is
*supposed* to be reasonable, but might contain high levels of dissolved
salts.  It might also contain a layer of "foam".  Quite a lot of foam,
in fact.  On many occasions, highly persistent foam can be seen coming
out of the overhead fill for the tank :-)  Ultrasonics was the only
thing I could think of that would cope with all this, and I just don't
know enough about the signal you get reflected back...

/Kevin
-- Kevin J. Maciunas              Net: KILLspamkevinKILLspamspamcs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD
--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@070301 by Jinx

face picon face
Can you add two clear plastic tubes to the outside of the
tank, on opposite sides ? Finding the average of the two
levels would tell you the true level. The tricky part is finding
the levels

Polystyrene ball with small magnet inside to detect with
Hall effect or reed switches

Use the ball to break a beam. Either a loooong line
of emitters/detectors or perhaps even a small motor
on a track that could scoot up/down so as to use just
one tx/rx optical pair

Tube sealed at the top so you could measure the air
pressure

Inflate a rubber sac, attached at the bottom of the tank,
and measure its expansion

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@081718 by Stuart Meier

flavicon
picon face
Hmm

1. Measure water pressure at the base of the tank (hydraulic pressure
sensor)
2. Measure the angle of the tank base using two accelerometers mounted
at 90 degree to each other
3. Look up table to correct pressure for slope of tank.

Dare I say Voila? Of course , if your tank is very irregular in cross
section, lookup could be preety big table.

Stuart
{Original Message removed}

2002\03\03@082752 by Sergio Masci

picon face
What about using the acoustic properties of the air space inside the tank?
What if you generate a varying tone in the air space and listen to see which
is the loudest frequency you get back. Kind of like blowing air across the
top of a bottle, the note you get depends on the air space inside the
bottle. Maybe you could measure the power output from a speaker inside the
air space and when the output is greatest this would correspond to the
loudest frequency. Then it would be a simple mater of calibrating the device
against known quantities of water.

Regards
Sergio

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@083631 by Russell McMahon

face
flavicon
face
>         Herewith the problem.  I want to sense the level(*) of water in a
> sealed aluminium tank.  The tank can have a hole or two bored into the
> top, but not the sides.  Access into the tank is basically impossible.
> The whole shebang is (rather nicely) welded shut.

   See 2 & 3 below first

A few quick thoughts:

1    - Using any of the several possible methods suggested, measure level at
front and rear of tank.
This gives the mean level. In extreme cases (low level and steep angle) the
water does not reach the :wall at the upper end and this method fails. At
this stage the tank is probably excessively empty. (Multiple air tubes as
below would allow this situation to be detected and a true result
calculated.

2    - As above but one sensor in tank centre which gives mean depth of tank
at any sensible tilt and fullness. .

3    - ***  BEST THOUGHT I THINK ===>   Someone mentioned presure in air
filled sealed tubes (Jinx?) .
This is used for level detection in washing machines with great success.
Insert a single tube in the middle of the tank. Seal at top end. Tube
reaches not quite to the bottom in the centre. When fully empty, tube is
filled with air. As tank start s to fill the air is trapped and as level
rises the pressure in tube increases. A pressure sensor at top end allows
"gead" to be measured. I suspect this method would allow results in the 1%
to 5% range which would be ample for your application.

- Consider strain gauges/load cells to "weigh" tank. This would not be
especially hard and would be extremely effective.
Mounting gauges on existing chassis /mounting members may be possible.
Calibration by empirical means (fill it in 5% steps and watch what happns
:-) !

- Audio resonance of tank air volime MAY work - ever listened to a tank
filling - pitch changes. May be badly spoofed by foam.

- Accelerate through gears - measure time for 0-50kph or whatever. Calibrate
time against load !!! :-)
OR brake from say 50 kph to standstill and calibrate.

- Some form of central sensor (see some of Jinx's ideas) could help as
above. Very likely a long capacitive fully sealed probe in tank centre would
suffice.A long linear "float" valve in tank centre may also be doable

- Knocking on sides to establish level mechanically is probably viable.
Could automate this :-) - would be amusing.

       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@085701 by Bob Ammerman

picon face
----- Original Message -----
From: "Christoph Graf" <RemoveMEcgrafTakeThisOuTspamNONLIMIT.CH>
To: <spamBeGonePICLISTspamBeGonespamMITVMA.MIT.EDU>
Sent: Sunday, March 03, 2002 3:06 AM
Subject: Re: [EE]: Hopefully, someone can answer this one!


> Kat
>
> What about measuring the distance top of tank - water level ???
> I don'r know how big the tank is, but I could imagine infrared
> or ultrasonic.
>
> Regards Chris

To handle tilt, you would have to measure at three points around the
perimeter of the tank. These three points would defined the plane which is
the surface of the H2O.

You can now compute your volume appropriately with a sneaky bit of geometry.

Bob Ammerman
RAm Systems

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@085908 by michael brown

flavicon
face
> Hmm
>
> 1. Measure water pressure at the base of the tank (hydraulic pressure
> sensor)
> 2. Measure the angle of the tank base using two accelerometers mounted
> at 90 degree to each other
> 3. Look up table to correct pressure for slope of tank.
>
> Dare I say Voila? Of course , if your tank is very irregular in cross
> section, lookup could be preety big table.
>
> Stuart

Why couldn't you just use more than one pressure sensor and average their
readings?  Let's say 9 sensors arranged in 3 rows of 3.  Since water
pressure goes up linearly with respect to volume, couldn't the average
reading of multiple sensors be treated the same as the reading from a single
sensor.  I'm sure things may be a little more complex due to the bottom of
the tank not being flat, but why wouldn't this work?

michael brown

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@113459 by dpharris

picon face
Hi-
Is there any possibility you dould weigh the tank?  Possibly strain guages at the
mounting points?  Obviously volume=weight.
David

Kevin Maciunas wrote:

{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@114527 by Victor Faria

flavicon
face
OK,I'll take a shot!
how about 2 stainless steel rods from the top.
in parallel the top of rods connected to a led or????
when you hit water led lights up.
regards
victor Faria
{Original Message removed}

2002\03\03@114735 by Pic Dude

flavicon
face
Not necessarily -- if you measure from the center of the tank,
that should be a automatic avreage of the level.  1 ultrasonic should
be all that's needed, even when tilted.  However, if tilt is enough
such that the sides of the top plane of water isn't at the side walls
anymore (such as one side touching the base of the tank), then
this is no longer valid.


{Original Message removed}

2002\03\03@115402 by Mahmood Elnasser

flavicon
face
The best way I came across was to use IR transmitter and receiver at the
top of the tank, the level of IR reflected from the tank depends on the
height of the water, u can do fancy stuff like taking the signal on the
power line using power line modems, this way only 2 wires go to the unit
at the tank.
Mahmood

{Original Message removed}

2002\03\03@120647 by Olin Lathrop

face picon face
> 1. Measure water pressure at the base of the tank (hydraulic pressure
> sensor)
> 2. Measure the angle of the tank base using two accelerometers mounted
> at 90 degree to each other
> 3. Look up table to correct pressure for slope of tank.

No, it's really simpler than that.  You want the water pressure right at the
tank outflow.  If the truck is parked on a hill, the extra water in the
corner of the tank is useless since it can't be used.  The raw pressure
(assuming top of tank is open to the air) at the outflow is exactly what you
want in all cases, hill or not, because it goes to zero when all *useable*
water is drained.  The tank capacity is effectively diminished when parked
on a hill.


********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, RemoveMEolinspamTakeThisOuTembedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@122341 by Chris Loiacono

flavicon
face
Simple one.

You say the top of the tank is accessible, right? Put three holes (OK, you
did say one or two, & I am asking for 3)in through the top, triangular
pattern, put flange fittings at each. Insert rigid tubing from these flanges
to reach near the bottom of the tank. Fit the tops of these tubes with
pressure sensors, Motorola MPX's would be cheap and easy to read. Do the
math or make a lookup table to get the mean level.

Chris

> {Original Message removed}

2002\03\03@124051 by Bob Ammerman

picon face
A single sensor in the center will indeed work if the tank is symmetrical
perpendicular to any possible axis of tilt. The 3-sensor solution will work
for any shape tank.

Bob Ammerman
RAm Systems

{Original Message removed}

2002\03\03@133747 by Olin Lathrop

face picon face
> A single sensor in the center will indeed work if the tank is symmetrical
> perpendicular to any possible axis of tilt. The 3-sensor solution will
work
> for any shape tank.

I agree with your geometry, but if you're fighting a fire you want to know
how "full" the tank is with respect to when no more water comes out of your
hose.  You don't care about the extra water trapped in the corner of the
tank because the truck is parked on a hill.

You guys are making this waaay to complicated.


********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, olinEraseMEspam.....embedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@151839 by Russell McMahon

face
flavicon
face
> No, it's really simpler than that.  You want the water pressure right at
the
> tank outflow.  If the truck is parked on a hill, the extra water in the
> corner of the tank is useless since it can't be used.  The raw pressure
> (assuming top of tank is open to the air) at the outflow is exactly what
you
> want in all cases, hill or not, because it goes to zero when all *useable*
> water is drained.  The tank capacity is effectively diminished when parked
> on a hill.

This indeed gets you the tank's "effective head"  which would be a measure
of the useful gravity flow rate that will come out of the tank.However, it
gives a very variable impression of the total remaining volume in the tank
(depending on where the exit pipe is). Say for example the exit was at
bottom front and the truck is parked nose down on a steep hill - the exit
head will be high even when the tank is rather empty.  You can though pump
most of it out. Park on the opposite slope and the apparent fill is lower
than true but you can at least move the truck to get at the rest of it.

I think that as long as the tank is relatively symmetric in each of forward
(X)  and sideways (Y) dimensions then a single centre mounted head sensor *
giving the average head will give the best indicator of true volume
remaining. Combining this with X and Y tilt sensors (say a dual axis
accelerometer) would allow usable volume before the exit draws air to be
calculated. (The technologically obsessed could even make a device to
predict how much more water could be got out by reparking the truck bu the
best advice there is probably to park with the drainndownhill :-) ).
(* - Such as the air tube with pressure sensor previously mentioned or some
other).



           Russell McMahon

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@171637 by M. Adam Davis

flavicon
face
I think of all the ideas, I like these two the best:

1. Audio at top.  Put a speaker through one hole, and a microphone
through another.  Sweep a range of frequencies - The one which gives the
greatest output is the resonant frequency of the air cavity.  A small
look up table will do the rest.

2. Three pressure sensors at the bottom of the tank.  (or even in the
sides very near the bottom).  Good part about this is if a sensor fails
you've still got two left and can at least give a good idea of what's left.

-Adam

Kevin Maciunas wrote:

{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@174708 by Pic Dude

flavicon
face
Trying to remember some physics here, but if pressure were
measured, wouldn't the amount of minerals, salt, etc change the
density of the water and cause inaccurate readings?  Unless of
course there were a "reset" switch that was triggered everytime
the tank was filled to the very top.



{Original Message removed}

2002\03\03@174915 by Russell McMahon

face
flavicon
face
> I think of all the ideas, I like these two the best:
>
> 1. Audio at top.  Put a speaker through one hole, and a microphone
> through another.  Sweep a range of frequencies - The one which gives the
> greatest output is the resonant frequency of the air cavity.  A small
> look up table will do the rest.

What will a full head of foam on top of water do?


       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@192645 by Kevin J. Maciunas

flavicon
picon face
On Sun, 2002-03-03 at 22:32, Jinx wrote:
> Can you add two clear plastic tubes to the outside of the
> tank, on opposite sides ? Finding the average of the two
> levels would tell you the true level. The tricky part is finding
> the levels
>
> Polystyrene ball with small magnet inside to detect with
> Hall effect or reed switches
>
> Use the ball to break a beam. Either a loooong line
> of emitters/detectors or perhaps even a small motor
> on a track that could scoot up/down so as to use just
> one tx/rx optical pair
>
> Tube sealed at the top so you could measure the air
> pressure
>
> Inflate a rubber sac, attached at the bottom of the tank,
> and measure its expansion
>

That's all fine, EXCEPT that I don't want to drill holes in the side of
the tank.  Think of "fail-safe" in this context :-)

/Kevin
--
Kevin J. Maciunas              Net: RemoveMEkevinEraseMEspamEraseMEcs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@192853 by Kevin J. Maciunas

flavicon
picon face
On Sun, 2002-03-03 at 23:46, Stuart Meier wrote:
> Hmm
>
> 1. Measure water pressure at the base of the tank (hydraulic pressure
> sensor)
> 2. Measure the angle of the tank base using two accelerometers mounted
> at 90 degree to each other
> 3. Look up table to correct pressure for slope of tank.
>
> Dare I say Voila? Of course , if your tank is very irregular in cross
> section, lookup could be preety big table.
>

Yup, that'd work.  My return question is: what kind of pressure
transducer can I get that can be inserted from the top of the tank to do
this?  The tank is a nice rectangular prism - perfectly regular!

/Kevin
--
Kevin J. Maciunas              Net: RemoveMEkevinspam_OUTspamKILLspamcs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@201255 by Jinx

face picon face
> That's all fine, EXCEPT that I don't want to drill holes
> in the side of the tank.  Think of "fail-safe" in this context :-)

OK then, use an S-bend. Starts at the bottom of the
tank, over the top, down to the bottom of the tank on
the outside and then back up to the height (or more)
of the tank

nyah nyah ;-)

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@210732 by Olin Lathrop

face picon face
> Trying to remember some physics here, but if pressure were
> measured, wouldn't the amount of minerals, salt, etc change the
> density of the water and cause inaccurate readings?  Unless of
> course there were a "reset" switch that was triggered everytime
> the tank was filled to the very top.

Geesh, guys.  You're not trying to fill a rocket with the exact right amount
of propellant.  I'm not a fire fighter, but I would expect they want to know
the tank is "kinda full", "getting low", and "needs to be refilled soon".
You also aren't going to park a fire truck on a 45 degree hill.  10 degrees
is a very steep hill (steeper than 6:1 grade).  Even at that angle, the
horizontal cross section of the tank is essentially the same as when the
truck is level.  The outflow pressure is therefore proportional to the
remaining water volume unless you are already well past the "needs to be
filled NOW" level.  This might be an interesting engineering challange, but
let's not loose sight of the real problem.


********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, RemoveMEolinTakeThisOuTspamspamembedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@212922 by Russell McMahon

face
flavicon
face
> Trying to remember some physics here, but if pressure were
> measured, wouldn't the amount of minerals, salt, etc change the
> density of the water and cause inaccurate readings?  Unless of
> course there were a "reset" switch that was triggered everytime
> the tank was filled to the very top.


Yes - but I would not expect that the effect would not be a major one for
any normal concentrations of additives (eg foam).

       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@220331 by Kevin J. Maciunas

flavicon
picon face
On Mon, 2002-03-04 at 05:06, Olin Lathrop wrote:
> > A single sensor in the center will indeed work if the tank is symmetrical
> > perpendicular to any possible axis of tilt. The 3-sensor solution will
> work
> > for any shape tank.
>
> I agree with your geometry, but if you're fighting a fire you want to know
> how "full" the tank is with respect to when no more water comes out of your
> hose.  You don't care about the extra water trapped in the corner of the
> tank because the truck is parked on a hill.
>
> You guys are making this waaay to complicated.
>
>

This is correct - accessible H2O counts!  The problem that kind-of needs
solving is the one in which the operator believes there is far more H2O
than there really is.  Other desireable properties are simplicity and
robustness [some people do not take good care of appliances, sadly].

So far, my favourite solutions are those from Jinx & Russell - a tube
with entrained air and a simple pressure transducer on the top.  I can
pretty clearly see how these could work and be retro-fitted.  And
they're pretty simple.

To all Piclisters: I'm impressed :-)

/Kevin
--
Kevin J. Maciunas              Net: EraseMEkevinspamspamspamBeGonecs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@220958 by Kevin J. Maciunas

flavicon
picon face
On Mon, 2002-03-04 at 11:41, Jinx wrote:
> > That's all fine, EXCEPT that I don't want to drill holes
> > in the side of the tank.  Think of "fail-safe" in this context :-)
>
> OK then, use an S-bend. Starts at the bottom of the
> tank, over the top, down to the bottom of the tank on
> the outside and then back up to the height (or more)
> of the tank
>
> nyah nyah ;-)
>

Got me.  I refuse to argue with Jinx, I know I'll lose!  I could kick
myself for not thinking of that....

For you Jinx : http://www.kleinbottle.com/ - perhaps you could buy one
of their hats, to keep warm on the cold NZ nights!

/Kevin
--
Kevin J. Maciunas              Net: RemoveMEkevinKILLspamspamcs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@221204 by Kevin J. Maciunas

flavicon
picon face
On Mon, 2002-03-04 at 10:10, Olin Lathrop wrote:
> > Trying to remember some physics here, but if pressure were
> > measured, wouldn't the amount of minerals, salt, etc change the
> > density of the water and cause inaccurate readings?  Unless of
> > course there were a "reset" switch that was triggered everytime
> > the tank was filled to the very top.
>
> Geesh, guys.  You're not trying to fill a rocket with the exact right amount
> of propellant.  I'm not a fire fighter, but I would expect they want to know
> the tank is "kinda full", "getting low", and "needs to be refilled soon".
> You also aren't going to park a fire truck on a 45 degree hill.  10 degrees
> is a very steep hill (steeper than 6:1 grade).  Even at that angle, the
> horizontal cross section of the tank is essentially the same as when the
> truck is level.  The outflow pressure is therefore proportional to the
> remaining water volume unless you are already well past the "needs to be
> filled NOW" level.  This might be an interesting engineering challange, but
> let's not loose sight of the real problem.
>
>

Again, right on the money, except that the steepest hill we've been on
is 25 degrees.  That is seriously steep.  The resolution of measurement
would fall into the "about 100-200litre steps" range.  Something like
that'd be nice.

/Kevin
--
Kevin J. Maciunas              Net: kevinSTOPspamspamspam_OUTcs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@224900 by Jinx

face picon face
> use an S-bend

I meant to add if you use an external clear tube, the firefighters
can use those wonderful processors known as eyes. The only
proviso is that once the tube is filled it would have to be re-filled
if the tank runs dry, but as they're filling the tank it should be no
work to give the tube a squirt too

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@225418 by M. Adam Davis

flavicon
face
I considered that, but I was concerned about the actually pressure of
the outlet pipe.  The pump is incredibly strong, and I would wonder
wether the pump was taking water from the tank (and therefore the tank
providing the pressure) or whether the pump was literally sucking the
water from the tank.  If it was sucking the water from the tank, then
the pressure right at the pipe inlet would be measured very low (as the
pump tries to suck the water from the measuring tool).  So I thought
this would be somewhat inaccurate when the pump is in operation.

-Adam

Olin Lathrop wrote:

{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@225421 by Robert Rolf

picon face
If the tube is connected to the tank, top and bottom, no refiling is required.
One could also have a tiny air hole on the top of a bottom fill tube to
vent it and get away with a single bottom connection.
"Sight tubes" are a common fixture on farm tanks.

Jinx wrote:
>
> > use an S-bend
>
> I meant to add if you use an external clear tube, the firefighters
> can use those wonderful processors known as eyes. The only
> proviso is that once the tube is filled it would have to be re-filled
> if the tank runs dry, but as they're filling the tank it should be no
> work to give the tube a squirt too

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\03@225613 by M. Adam Davis

flavicon
face
Good point...

Three pressure sensors at the bottom (or at the top with tubes) will
essentially approximate a weight sensor, and is probably the most
reliable way to gat an accurate reading...

-Adam

Russell McMahon wrote:

{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\04@003448 by Kevin J. Maciunas

flavicon
picon face
On Mon, 2002-03-04 at 14:16, Jinx wrote:
> > use an S-bend
>
> I meant to add if you use an external clear tube, the firefighters
> can use those wonderful processors known as eyes. The only
> proviso is that once the tube is filled it would have to be re-filled
> if the tank runs dry, but as they're filling the tank it should be no
> work to give the tube a squirt too
>

Just to complete the documentation loop, existing appliances have a
clear pipe running top to bottom (or vica versa :-) ).  This is totally
useless - the vinyl (I think) tubing goes opaque brown, the level can't
be seen and at night it's a real challenge.  Some versions have a rigid
plastic pipe with an orange float - these seem to go brown due to algal
growth.

I never thought such a straight forward question would engender such a
storm of neural activity!

/Kevin
--
Kevin J. Maciunas              Net: KILLspamkevinspamBeGonespamcs.adelaide.edu.au
Dept. of Computer Science      Ph : +61 8 8303 5845
University of Adelaide         Fax: +61 8 8303 4366
Adelaide 5005 SOUTH AUSTRALIA  Web: http://www.cs.adelaide.edu.au/~kevin
Fingerprint = 7E5A A0C2 22BC 5993 17F2 93CE B1FD DEC6 D0C0 50CD

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@020125 by cdb

flavicon
face
<HTML><HEAD>
<BASEFONT FACE="Arial" SIZE="2" COLOR="#000000">
</HEAD>
<BODY>
<div>How about one or two balanced bridges connecte to a comparator one of the arms of the bridge being a PTC resistor encase in something water proof. The PTC will alter its resistance depending on the temperature of the water, which will also depend on the volume of water. Not too sure about accuracy on a hot day though.</div>
<div>&nbsp;</div>
<div>I like the washing machine idea.</div>
<div>&nbsp;</div>
<div>&nbsp;</div>
<div>colin </div>
<div><FONT FACE="Arial" SIZE=2>-- </FONT></div>
<div><FONT FACE="Arial" SIZE=2>cdb, EraseMEbodgy1spamEraseMEoptusnet.com.au on 04/03/2002</FONT></div>
</body></html>

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@020333 by Russell McMahon

face
flavicon
face
> Just to complete the documentation loop, existing appliances have a
> clear pipe running top to bottom (or vica versa :-) ).  This is totally
> useless - the vinyl (I think) tubing goes opaque brown, the level can't
> be seen and at night it's a real challenge.  Some versions have a rigid
> plastic pipe with an orange float - these seem to go brown due to algal
> growth.

More inane potential solutions are thus suggested -

My blocking the top of this existing tube and placing a pressure sensor in
it you get your head sensor for that tube.
By mounting an ADXLxxx dual accelerometer (or something similar) to the
appliance you get your body X/Y tilt angles.
Then it's simple trig to get the volume. Stops working if the tube draws air
at the bottom (tilt away from tube inlet) or if the tank fills to the top at
the tube due to tilt. So a centre mounted air tube is still probably best.
It too can suck air at very low levels and highish angles but by that stage
you know you are dangerously low.
Or just use the head sensor with a roughish correction chart for body tilt.
A REAL mechanical engineer could produce a linkagebased device that moved a
meter scale to adjust the pointer for body tilt     :-)          :-)
:-)


You MIGHT get away with mounting conductive insulated strips down either
side of the sense tube and attempting to read the capacitance (very simple
to do if the results are totally wrecked by adjacent body etc).

Float a flashing doo-da in the sight tube to make the fluid point MUCH more
visible.
This could eg have a xenon strobe tube that only flashed every minute or so
at relatively low intensity so the internal battery lasted for mega-yonks.
Even a flashing high intensity LED may do and would be VERY easy to try. At
10 second flash rate and 0.1s flash a 12 mA pulse would get you a year on a
1000 mAH battery. Some modern LEDs are VERY bright at this current (10's of
candela ! )

Float a ferrous cored ball in the tube and detect it. How? Maybe a coil
fairly loosely wound down the outside with drive signal and a sense signal
picked up in another coil at the point where the ball floats. Maye the coils
could even be antiphase wound and at 90 degrees to each other (by winding
each at 45 degree pitch to the tube with the ball providing coupling. I'd
have to think for a while to see if that would really work.

Maybe a tame cockroach or tadpole ....



       Russell McMahon


{Original Message removed}

2002\03\04@030306 by Jinx

face picon face
> brown due to algal growth

(a) water snails - hmmm, perhaps not

(b) small brightly coloured tropical algae-eating reef fish
    - but there's the danger of the tank looking like a pair
    of 1970s disco pimp shoes

(c) small algaecide tablet (as got from a pet shop) that
    can sit in the tube for several weeks killin' time

(d) two arms, each with an elbow joint, thin enough to pass
     through holes in the top. I'm assuming that the tank is
     higher than it is wide, which is why I suggest two arms

Imagine your arm hanging straight down. The top of the
tank is at your armpit. Lucky tank eh ?

Each has a float (where your hand is). When the tank is
empty, one float sits on the bottom of the tank. The length of
the lower part of the arm (the forearm) is such that it won't
touch the side of the tank when it rises to the horizontal. The
other arm is similar and moves in the upper portion of the
tank that is past the level of the lower arm when horizontal.

Either the (elbow) joint has an optical angular measurement
device, or the forearm could house several mercury tilt switches.
Those tilt switches can be set at angles to trip when the forearm
raises to a certain height, so you can determine the angle of it

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@033248 by Robert Rolf

picon face
"Kevin J. Maciunas" wrote:
>
> On Mon, 2002-03-04 at 14:16, Jinx wrote:
> > > use an S-bend
> >
> > I meant to add if you use an external clear tube, the firefighters
> > can use those wonderful processors known as eyes. The only
> > proviso is that once the tube is filled it would have to be re-filled
> > if the tank runs dry, but as they're filling the tank it should be no
> > work to give the tube a squirt too
> >
>
> Just to complete the documentation loop, existing appliances have a
> clear pipe running top to bottom (or vica versa :-) ).  This is totally
> useless - the vinyl (I think) tubing goes opaque brown, the level can't
> be seen and at night it's a real challenge.  Some versions have a rigid

Hummm. How about using some of the truck's lighting circuits
to illuminate the tube? With a tube at either end the human brain
could easily figure out how full the tank was, even when tilted.

> plastic pipe with an orange float - these seem to go brown due to algal
> growth.
>
> I never thought such a straight forward question would engender such a
> storm of neural activity!

If it were so straight forward, would he have needed to ask the list <G>??

R

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@034739 by Lee Jones

flavicon
face
>> [water level in fire truck tank]

Couple possible solitions that don't involve electronics
at all (I know, heretical).

A dip stick?  Person sticks it into the tank, pulls it out,
and reads the water level remaining.

Mechanical indicator dial on a twisted, rectangular rod.
Rod has a 1/2 to 3/4 twist over it's ~4 foot length.  Rod
goes through a bearing in the tank cap.  Tank cap also has
2 fixed round wires going down into the tank.  Far end of 2
wire provide bearing for end of twisted rod.  Float has 2
round holes (for round wire rods) and rectangular hole (for
centered rectangular rod).  As float rises and falls, it
slides along and rotates rectangular rod and actuates needle
indicating water level.

                                               Lee Jones

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@050850 by Alan B. Pearce

face picon face
>Weighing the tank would, of course, work.  I was hoping for a simple and
>cheap solution :-).  Hey, I can always wish!

Could you use the height of water in the tank given by the existing sensors,
and make an inclinometer from an ADXL202, and calculate the amount of water
that way. It will give you an error on the slope you describe though, but
could be shown as ">1/4 tank" under those conditions.

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@074411 by NDuckworth

flavicon
face
How much water does the tank hold?

Would the difference between a full and empty tank change the compression of
the trucks suspension significantly? If so, you could measure the displacement
of each spring to correct for the attitude of the truck.

Just a thought.

Nigel


On Sunday, March 03, 2002 9:53 AM, Kevin Maciunas
[SMTP:@spam@kevin@spam@spamspam_OUTCS.ADELAIDE.EDU.AU] wrote:
{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@083657 by Francisco Ares

flavicon
face
Did anyone though about a pressure sensor?  A tube closed in one end
that is put to the tank with its opening down, so the air pressure
inside of this tube increases as the water level rises.

There are some pressure sensors from Honeywell, for example
(http://content.honeywell.com/sensing/prodinfo/pressure/) but s search
on the net will show many options.

A diferential type of sensor would be required, the high pressure inlet
connected to the upper part of this tube and the lower pressure inlet
left opened to the inner pressure of the tank (which I suppose is
atmospheric pressure)

Some sensors are very small and they give a 0 ~ 5 V signal range.

Then a A/D capable PIC could do the reading and the math.

Best  regards
Francisco


Kevin Maciunas wrote:

{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@092551 by Thomas McGahee

flavicon
face
Several possible approaches come to mind:

1) Since you mentioned that clear tubes become discolored due to algae
growth, use a sealed approach. A flexible bag (something akin to the old
water bottles) is attached to a rubber hose. The other end of the rubber
hose
forms part of an "S" curve and attaches to a clear plastic tube.
The bag goes down at the bottom of the tank. It will need to be kept
in place by a mount or a weight, or a strong magnet.

The plastic column is initially raised and *COLORED* water is poured
into the top of the column until the bag and rubber hose are filled
such that the colored water shows in the bottom of the clear column.

The column is then lowered and mounted on the side of the tank.
Excess colored water can be removed by inserting a thin plastic
hose like that used for aquariums and sucking out as much water
as needed. (Note that the top of the column should be higher than the
maximum water level in the tank)

The top of the clear column should be capped off with something
that will seal the column but also allow the water in the column to
rise and fall without much air compression. A *LARGE* balloon or
beachball or even a modified inner tube could be used.

Now when the level of water in the tank changes, so will the level
of colored water in the clear column. Because the column is a closed
system you will not have algae contamination.

Multiple systems could be used to measure water height at different
parts of the storage tank. This would help when the tank was
at a steep angle.

The larger the bag at the bottom of the tank, the more accurate
the results.

2) Do pretty much the same thing, but instead of a clear plastic tube
have a pressure transducer attached to the rubber hose. This keeps the
transducer outside the tank.

3) Measure the volume of *AIR* in the tank. The more air there is, the
less water there is.

To measure the air volume use a waterproof microphone and a waterproof
speaker. These should be mounted just inside the tank, where they
will remain reasonably dry.

Sweep a frequency into the speaker and use the microphone to detect
the resonant frequency of the system. The amplitude will peak at
the resonant frequency, which will probably be at quite a low frequency.
You should be able to map out resonant frequency versus air volume
and then relate that directly to amount of water left in tank.
Be aware that the speaker and microphone will each have their
own "resonant" frequency, but these should be much higher than
the resonant frequency of the tank system.

4) Same as method 3, but mount microphone and loudspeaker
to OUTSIDE of tank. In this case you are vibrating the tank
itself, not just the air, and you would need higher sound levels.

5) Have a mechanism that on command "bangs" the side of the tank.
Use a PIC to analyse the resultant sound and determine amount of
water left in tank. All of these methods require calibration
before you can get accurate results.

Fr. Thomas McGahee

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@120328 by Chris Loiacono

flavicon
face
Free program management:

I think two things are repeatedly being overlooked: 1) the entire system
requirement approach and 2)simplicity, which includes not re-inventing the
wheel, and making sure the customer is happy with the result on day one, and
ten years later.

The dry air pressure sensor in-a-tube approach (see earlier message below)
is a standard, accepted practice that is often employed in low-cost liquid
level sensing apps for several reasons, among these are: 1) completely dry
sensor and wiring 2) minimal if any compromise to the original tank
structure and seal.

If the tank will always be used on a reasonable grade, and  the resolution
and accuracy required are not very high, a single ceter-mounted sensor would
suffice. I have a fixed verticle tank 10 ft. away from me right now that has
had this same system working in it daily for the past 16 years with zero
leaks and zero maintenance. It was done before conditioned sensor outputs
and integrated A/D. Think of how much simpler it should be today!!
New wheel shapes not required.

If I were the Engineering Project Manager on this one, I would have added
these requirements to the customer's because reducing the number of holes
and seals in a liquid-containing vessel means increased reliability and less
maintenance. Top-mounted tubes require one hole for the tube, sensor access,
& all. Sensing dry air at the top of a tube means almost no sensor
maintenance over the life of the tool. Integrated sensors with conditioned
outputs are available that make programming straightforward.

So, less holes and fittings, less leaks now & later, also fast and easy to
assemble, less parts cost. Standard, off-ther-shelf fitting and tube. The
tank level can be read when in use or at rest. High reliability of dry
hardware. No mechanical parts to wear or break. No concern with algae or
reasonable contamination.

Callme arrogant if you like, but....C'mon, guys...This should already be
set-up and running for under $100.00 cost -

As usual, the serious effort needs to go into the user interface. How will
the user be informed of the tank level, and what will the system do if the
tank is low or empty?

Chris


> {Original Message removed}

2002\03\04@144954 by Peter L. Peres

picon face
Imho you need at least three sensors to read the height regardless of the
angle. This may be more complicated than you think (capacitive linear
sensors may work - do you have a budget ;-). Imho just use a flowmeter on
the water outlet and a float switch for 'full' to initialize the mass flow
measurement. It will not account for evaporation and such.

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@160404 by Eoin Ross

flavicon
face
The issue with a flow meter may well be cost ... even small bore ones that only go up to 10 gallons per minute reach into the US$200 range - this is going to be on a fire truck in Australia - and from what the original poster has said I am gathering it is a 4WD vehicle that will see some severe conditions, not your typical urban fire-truck application.

My initial thought was capacitive sensors as in aircraft fuel tanks too - the other suggestion that has been presented of using the resonant frequency (IMHO) has merit in a "noiseless" environment - this application will see a lot of ambient noise and vibration from 3-8 HP single cylinder motors driving pumps to possibly a large diesel engine from the truck driving the pumps - that makes me tend to think that the speaker idea may be too unreliable.

The pressure sensor sealing the top of an air filled tube seems to me to be the easiest and most reliable solution. 15 PSI differential pressure sensor  with one port to the (filtered) atmosphere would be the sensor to use (maybe more PSI advisable).

>>> TakeThisOuTplp.....spamTakeThisOuTACTCOM.CO.IL 03/04/02 02:48PM >>>
Imho you need at least three sensors to read the height regardless of the
angle. This may be more complicated than you think (capacitive linear
sensors may work - do you have a budget ;-). Imho just use a flowmeter on
the water outlet and a float switch for 'full' to initialize the mass flow
measurement. It will not account for evaporation and such.

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@170913 by Chris Loiacono

flavicon
face
In regard to>

> The pressure sensor sealing the top of an air filled tube
> seems to me to be the easiest and most reliable solution. 15
> PSI differential pressure sensor  with one port to the
> (filtered) atmosphere would be the sensor to use (maybe more
> PSI advisable).
>
So, a PSIG sensor type will work, which is basically a differential sensor
with the ambient pressure at the second port. Also, it takes a lot of
pressure on a gas to increase the pressure by 15 psi. Your sensor range
selection  depends upon the height of your tank and the tube diameter.

As the level rises in the tank, the air wil be compressed, right....  What
happens when the fluid is rapidly pumped out? Well, next consider that many
positive pressure sensors do not tolerate negative pressure, so if the pump
is capable of emptying the tank fluid pretty quickly, meaning more quickly
than return air can enter the tank, you will need to provide a tank top vent
to allow equalizing without delay. Without this your sensor substrate can be
damaged by the suction and you will get inaccurate output that would be
difficult to diagnose when debugging the final set-up.

Chris

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@181529 by Dave Mumert

flavicon
face
Hi All

I hate to be propose a non-electronic system but Rochester makes a sender
unit rated for 48" deep tanks and off-road use.  Not exotic but probably the
easiest.
http://www.rochestergauges.com/Pages/PDFs/8340_8343.pdf

Dave Mumert

{Original Message removed}

2002\03\04@183932 by Eoin Ross

flavicon
face
I was thinking about atmospheric pressure - which won't be an issue with the differential/gauge sensors (looked at measuring atmospheric pressure before for a future project with absolute pressure sensors)
I would imagine the tank has a pretty good vent at the top - otherwise the pump would become woefully inefficient against the negative head (Unless it is also capable of sucking out of creeks/lakes -  and by the sound of it we are dealing with a rural/bush fire-truck which makes that highly likely)
BTW - I am not the original poster so I haven't seen the tank :)

The equation for pressure vs head height would be ...

The size of the tube shouldn't affect the pressure - only the height of the liquid.

As pressure = force/area and force is directly proportional to the height (which "won't change" for this example)

If we increase the area by making our tube larger - force will increase - but the area will increase in exactly the same proportions, therefore pressure will remain the same.

We get to a pretty easy equation for what range the sensor will see - 1 gram/cm2 per cm of head. (assume maximum head is the longest diagonal in the tank - although if the truck got that far over I think they'll be worried about more than the water level in the tank ! )

I bet the poor guy in Aussie never thought there'd be this much debate/info thrown at him LOL.

Eoin

>>> TakeThisOuTchrisKILLspamspamspamMAIL2ASI.COM 03/04/02 05:07PM >>>
In regard to>

> The pressure sensor sealing the top of an air filled tube
> seems to me to be the easiest and most reliable solution. 15
> PSI differential pressure sensor  with one port to the
> (filtered) atmosphere would be the sensor to use (maybe more
> PSI advisable).
>
So, a PSIG sensor type will work, which is basically a differential sensor
with the ambient pressure at the second port. Also, it takes a lot of
pressure on a gas to increase the pressure by 15 psi. Your sensor range
selection  depends upon the height of your tank and the tube diameter.

As the level rises in the tank, the air wil be compressed, right....  What
happens when the fluid is rapidly pumped out? Well, next consider that many
positive pressure sensors do not tolerate negative pressure, so if the pump
is capable of emptying the tank fluid pretty quickly, meaning more quickly
than return air can enter the tank, you will need to provide a tank top vent
to allow equalizing without delay. Without this your sensor substrate can be
damaged by the suction and you will get inaccurate output that would be
difficult to diagnose when debugging the final set-up.

Chris

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@205504 by 859-1?Q?Alexandre_Guimar=E3es?=

flavicon
face
Hi,

> The pressure sensor sealing the top of an air filled tube seems to me to
be the easiest and most >reliable solution. 15 PSI differential pressure
sensor  with one port to the (filtered) atmosphere would >be the sensor to
use (maybe more PSI advisable).

   You can even look at cheap pressostats made for washing machines ! Some
of them have 3 outputs and are incredibly reliable. I never could figure out
how that could work reliably but it does in fact. I have never seen one of
those fail and I always have the first look on the broken washing machines
of all my family :-) The pressure sensor will give you more details but if
you just need 3 or 4 levels of water just put together one or two of these
washing machines pressostats and you will have something that is more
reliable than a single sensor. I do not think vibration will be a problem
either. I had an old machine that kept "walking" around all the time during
the washing cycle. My fellow countrymen will probably remember the first
washing machines from "enxuta" ! It was a great piece of equipament but had
this terrible noise, shaking and "walking" problem. The level sensor never
gave me any trouble on that one either and it lasted many years.

Best regards,
Alexandre Guimaraes

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@205521 by Russell McMahon

face
flavicon
face
> I hate to be propose a non-electronic system but Rochester makes a sender
> unit rated for 48" deep tanks and off-road use.  Not exotic but probably
the
> easiest.
> http://www.rochestergauges.com/Pages/PDFs/8340_8343.pdf
>
> Dave Mumert




Not rugged enough !!! :-)
It only has about a 3 hour lifetime at maximum rated vibration.
How's that going to stand up to use on a real Ocker bush fire truck?


       RM :-)


(Read the vibration specs carefully to see what I mean ...)


>
> {Original Message removed}

2002\03\04@214456 by 859-1?Q?Alexandre_Guimar=E3es?=

flavicon
face
Hi,

> > I hate to be propose a non-electronic system but Rochester makes a
sender
{Quote hidden}

   There is another problem with it !! The variable resistor is usually not
sealed from the fuel. It will not give nice results with the water jumping
all over it.

Best regards,
Alexandre Guimaraes

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@225730 by Chris Loiacono

flavicon
face
Now I have a new way to drive the people in my local appliance parts store
nuts! They can never figure out why I am often there looking through the
various heaters, switches, and sensors hanging on the wall. They really hate
when I ask them if they can get a sensor like the one I saw in a
thig-a-ma-jig...

he he :)

BTW, this sounds like a good idea. Are they really inexpensive?
when I last checked the cost of a low pressure switch for my Sears water
softener tank, I quickly decided it wasn't worth fixing with the original
type part...

Chris

>     You can even look at cheap pressostats made for washing
> machines ! Some
> of them have 3 outputs and are incredibly reliable.
>

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\04@235213 by 859-1?Q?Alexandre_Guimar=E3es?=

flavicon
face
Hi, Chris

> BTW, this sounds like a good idea. Are they really inexpensive?
> when I last checked the cost of a low pressure switch for my Sears water
> softener tank, I quickly decided it wasn't worth fixing with the original
> type part...

   Down here in Brazil I remember to get an adjustable one with 2 outputs
for less than us$ 9 ! I consider that quite inexpensive for all the
mechanics that are inside of it, much cheaper than a pressure sensor and all
the electronics that has to go with it. By the way, even cheaper are the
solenoid valves to control the water flow, aprox. us$ 2 each ! I can never
understand how they manage to manufacture that consumer stuff so cheaply.

Best regards,
Alexandre Guimaraes

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\05@005008 by Chris Loiacono

flavicon
face
Alex:
Wow! i put a new solenoid valve in my refrigerator's ice maker a while ago -
I think I paid about $27.00 US for it. The pressure switch for my water
softener was over $50. So, should I get appliance parts from Brazil? Will I
be able to carry them on board my flight?

It sounds like the US really has cornered the market on greed!
Really, though - I will look into the cost of these things here. Thanks for
the idea.

Chris


> {Original Message removed}

2002\03\05@021102 by mooseman

flavicon
face
deep pocket theory.

you charge whatever the market will bear.

lots of factors, cost of living, average annual salary.... blah, blah.

not saying i like it either, but that seems to be the way of things.

moose.

ps/ about the onboard thing.... not sure. i recently flew from vancouver to
calgary and back. on my trip back to vancouver, when going through security,
i had a little 2" long keychain style cresent wrench on my keychain.....  it
was not allowed to go with me. i asked if they thought i would be able to
remove a wing mid-flight.... they have no sense of haha.
:-(

On March  4, 2002 09:03 pm, you wrote:
{Quote hidden}

> > {Original Message removed}

2002\03\05@033016 by Alan B. Pearce

face picon face
>deep pocket theory.

>you charge whatever the market will bear.

>lots of factors, cost of living, average annual salary.... blah, blah.

Probably find the parts are made in Brazil, and shipped to USA

:)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\03\05@111501 by Mike Mullen

picon face
Okay, why is everyone trying to measure the pressure at the top of the sight
glass, via the air column?  Take the damn glass off, plug the top hole, and
put a pressure sensor in the bottom hole.   2 feet/psi is probably accurate
enough as a conversion factor. Screw the tilt correction, it ain't in the
sight glass, so you don't need it either.

Mike Mullen

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\03\05@120835 by Chris Loiacono

flavicon
face
All have been good solutions. The best choice depends on the the user's
(payer's) perspective as regards cost, ease of use, reliability and
serviceability.

Each solution offered has merit. Each offers unique features and overcomes a
different potential limitation. Only the people that will use it can shed
further insight in this, I believe...

The really super thing aboutthis is that it shows that this list can
accomplish nearly anything in the world of control design.
Now, is there anyone out there thathas venture capital connections?

(slightly half-serious chuckle)

Chris

> {Original Message removed}

2002\03\05@174143 by Dwayne Reid

flavicon
face
At 12:04 PM 3/5/02 -0500, Chris Loiacono wrote:

>Now, is there anyone out there that has venture capital connections?
>
>(slightly half-serious chuckle)

Warning: biased link below.  But check out the Banff Venture Form.

<http://www.icetalliance.org/events/default.html>

I say 'biased' because I a member of the ICET Alliance board.  But all who
have attended have had very high praise for the event - especially those
who used the event to arrange their funding!

dwayne


Dwayne Reid   <.....dwaynerspamRemoveMEplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax

Celebrating 18 years of Engineering Innovation (1984 - 2002)
 .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-.   .-
    `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'   `-'
Do NOT send unsolicited commercial email to this email address.
This message neither grants consent to receive unsolicited
commercial email nor is intended to solicit commercial email.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\03\06@125154 by Peter L. Peres

picon face
>deep pocket theory.
>
>you charge whatever the market will bear.
>
>lots of factors, cost of living, average annual salary.... blah, blah.
>
>not saying i like it either, but that seems to be the way of things.

In a normally working free market economy the fact that a price is X
automatically means that there is probably a way to get it for Y where Y <
X, iff you look hard enough and spend the extra effort.

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservspamspamBeGonemitvma.mit.edu with SET PICList DIGEST in the body


2002\03\06@130209 by Martin Peach

flavicon
face
> >deep pocket theory.
> >
> >you charge whatever the market will bear.
> >
> >lots of factors, cost of living, average annual salary.... blah, blah.
> >
> >not saying i like it either, but that seems to be the way of things.
>
> In a normally working free market economy the fact that a price is X
> automatically means that there is probably a way to get it for Y where Y <
> X, iff you look hard enough and spend the extra effort.

In a normally working free market economy the price will go down to meet the
real cost of production, which will itself decline as the production process
is made more efficient. In the real world such things as tariffs, cartels
and intellectual property rights impede this process.
/\/\/\/*=Martin

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spamBeGonelistserv@spam@spamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body


2002\03\06@132045 by mooseman

flavicon
face
indeed.
really fabulous deals like NAFTA (north american free trade agreement)
what i want to know is free trade for who???

anyways, i had better not go off on that tangent because it will rapidly
deteriorate into a serious rant. ;-D

moose.

On March  6, 2002 09:59 am, you wrote:
{Quote hidden}

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservEraseMEspammitvma.mit.edu with SET PICList DIGEST in the body


2002\03\06@141212 by Russell McMahon

face
flavicon
face
> In a normally working free market economy the price will go down to meet
the
> real cost of production, which will itself decline as the production
process
> is made more efficient. In the real world such things as tariffs, cartels
> and intellectual property rights impede this process.

Yes, but - Intellectual property rights also have real cost of productsion
and production proces components.
Only so many geniuses will work for food alone (or less) and only so many
parents will continuw to raise geniuses to live out their lives in 3rd world
mental sweat shop conditions.


       RM

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservEraseMEspamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body


2002\03\07@042131 by Alan B. Pearce

face picon face
> > >deep pocket theory.
> > >
> > >you charge whatever the market will bear.
> > >
> > >lots of factors, cost of living, average annual salary.... blah, blah.
> > >
> > >not saying i like it either, but that seems to be the way of things.
> >
> > In a normally working free market economy the fact that a price is X
> > automatically means that there is probably a way to get it for Y where Y
> > < X, iff you look hard enough and spend the extra effort.
>
> In a normally working free market economy the price will go down to meet
> the real cost of production, which will itself decline as the production
> process is made more efficient. In the real world such things as tariffs,
> cartels and intellectual property rights impede this process.

Well it is usually competition that drives down prices. The price stays high
until there is a competitive market. Witness what happened to Intel CPU
prices when other manufacturers started making pin compatible processors.

--
http://www.piclist.com hint: To leave the PICList
@spam@piclist-unsubscribe-requestRemoveMEspamEraseMEmitvma.mit.edu


2002\03\08@150747 by John

flavicon
face
Hello Kevin & PIC.ers,

Pointer    1.
If the tank is rectangular, then it'll follow that a depth measurement taken
in the `plan-centre' of the tank will not vary wrt vehicle attitude.

Pointer    2.
Depth measurement has been thrashed out on the list a lot in the past.
Archives should help.

pulled out of me own grab-bag, the following with compliments to the
original originators  : 0)

..
Right now the most attractive proposition looks like a weighted plastic tube
running back to a fish-tank air pump and pressure transducer. One pump can
be used for more than one tube, with one-way valves. The really big
advantage of this is that you can run the tubing a long distance from the
tank to a convenient place for the electronics with no complicated
mechanical bits to construct.
..

Date: Fri, 16 Jul 1999 17:13:24 -0700
From: Gordon Couger <EraseMEgcougerspam@spam@rfdata.net <@spam@gcougerspam_OUTspam.....rfdata.net>> ()
Subject: water level contoler

John,

The simplest capacitate water level sensor I ever made was a piece of
TV twin lead and a NE602 mixer chip. I used the twinlead in the frequency
as a capacitor in the oscillator and counted the frequency which is
inversely proportional to the depth of the water in the vessel.

It appears to me that the water effectively progressively shorts out the
twin lead making it appear shorter. It does not work as a capacitor with
the water increasing the capacitance because the frequency go up with the
water level not down.

Good luck
Gordon

Gordon Couger spamBeGonegcougerEraseMEspamcouger.com <gcougerspamBeGonespamcouger.com>
624 Cheyenne
Stillwater, OK 74075-1411
405 624-2855 GMT -6:00 http://www.couger.com/gcouger
<http://www.couger.com/gcouger>

;From: "john maniraj" <RemoveMEmercyjohn@spam@spamspamBeGonehotmail.com <.....mercyjohn@spam@spamEraseMEhotmail.com>>
()
;Subject: water level sensing

;;Hello Friends,

; I am trying to design a water level controller using the 89C51
;microcontroller.

; Could anybody give me an idea on the water sensing part?

; Electronic magazines suggest developing a square wave and using a
;capacitor to block the DC to the sensor,(to avoid electrolysis) but are we
;generating a pure AC? (since no negative supply is used anywhere).

;Any help will be greatly appriciated.

;John.



{Quote hidden}

       best regards,   John


e-mail from the desk of John Sanderson, JS Controls.
Snailmail:          PO Box 1887, Boksburg 1460, Rep. of South Africa.
Tel/fax:            Johannesburg  893 4154
Cellphone no:   082 741 6275
email:                .....jsandSTOPspamspam@spam@pixie.co.za
Manufacturer & purveyor of laboratory force testing apparatus, and related
products and services.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\03\09@071845 by Peter L. Peres

picon face
> NE602 + twinlead to measure water level

The frequency output of that device will not vary linearly with water
level as the NE602 uses a LC tank oscillator and that gives f ~=
1/sqrt(C). A RC oscillator can be used instead for a linear indication, or
TDR using much higher frequency, both giving linear f / C relationship.

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\03\11@052016 by Alan B. Pearce

face picon face
>The simplest capacitate water level sensor I ever made was a
>piece of TV twin lead and a NE602 mixer chip. I used the
>twinlead in the frequency as a capacitor in the oscillator
>and counted the frequency which is inversely proportional to
>the depth of the water in the vessel.

>It appears to me that the water effectively progressively shorts
>out the twin lead making it appear shorter. It does not work as
>a capacitor with the water increasing the capacitance because the
>frequency go up with the water level not down.

It sounds to me as though the twinlead is being used as a transmission line,
and the surface of the water acts as a moving ground plane. I would reckon
that the length of twinlead from the oscillator to the water surface is
probably a half wavelength at the oscillation frequency, rather than the
normal 1/4 wavelength, as the water will try and make a low impedance point,
being an effective groundplane, rather than a high impedance point.

It may be possible to tune the frequency range of the oscillator using some
L or C at the oscillator end of the twinlead. A transmission line looks like
an L or C depending which side of the resonant frequency it is operating, so
by doing the right sort of things it could be possible to pull the frequency
closer to where the length of twinlead is a quarter wavelength, which will
mean the oscillator is running at a lower frequency.

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservEraseMEspam@spam@mitvma.mit.edu with SET PICList DIGEST in the body


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...