Re: Measuring liquid level with bubbles using an ultrasonic transducer

[Marc_F_Hult <MFHult@xxxxxxxxxxxxxxxxxxxxx>]

On 12 Mar 2007 17:05:25 -0700, overgeo@xxxxxxxxx wrote in message
<1173744325.420066.302610@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>:

>On 1 mar, 04:46, Marc_F_Hult <MFH...@xxxxxxxxxxxxxxxxxxxxx> wrote:
>> On 28 Feb 2007 15:54:05 -0800, over...@xxxxxxxxx wrote in message
>> <1172706845.461279.186...@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>:
>>
>>
>>
>> >Sorry for my late answer, i have been doing other projects. The liquid
>> >is oil and is viscous, the accuracy is not a lot, the height of the
>> >tank is 5 meter and an error of 10 cm is perfectly asumible.
>>
>> >So i can use a differential pressure transducer, with one transducer
>> >that measures the pressure in the head of the tank and another one
>> >that measures the pressure in the bottom of the tank, and with the
>> >difference of both pressures i get the heigh of the liquid with this
>> >formule h = differential pressure/(specific gravity * g)  where g =
>> >acceleration due to gravity.
>>
>> >I have been looking for some sensors, and in this page there are some
>> >of them.http://www.sensotec.com/pdf_catalog06/fp2000.pdf, the fp2000
>> >wet/wet differential, i think will be suitable, what do you think?
>>
>> >To calculate the pressure range, i can use the formule.
>>
>> >max differential pressure =    h_max * (specific gravity * g),  is
>> >this correct?
>>
>> >In addition the idea of the pipe was good, but i think that it will be
>> >easier the differential pressure.
>> >Finally thanks so much for all your answers
>>
>> I see that this is an industrial application, not home, so I recommend
>> industrial components and mounting of the transducer outside the tank.
This
>> will better address the explosion hazard issues the specifics of which
>> depend on the fluid, application and location.
>>
>> - Yes, that is the equation I provided, but see below for converting
>> transducer output voltage directly to meters of fluid.
>>
>> - You said that the upper part of the tank would be under vacuum, thus
the
>> differential transducer.  The 5m of fluid water requires ~=500 millibar
~=
>> 50kP  =~7.25 PSI. The nearest higher value offered by Honeywell/Sensotec
>> seems to be the 10PSI   differential (~= 70kPa) But see below.
>>
>> - My preference is for 0-10vdc output assuming the associated
electronics
>> are nearby.
>>
>> - So with 10 PSI range and 10vdc Full Scale Output (FSO) the output is
>> 1 PSI/volt = 0.701 meters of water /volt
>>
>> (You'll want to check my arithmetic! 14.7 PSI = 33.9 feet of water =
10.3
>> meters of water. )
>>
>> For a fluid with specific gravity of n, the output would be (0.70/n)
>> meters/volt output. For example, for an oil with a sp. gr. of 0.8, the
>> output would be (0.70/0.8)= 0.875 meters oil/volt output.
>>
>> - The wetted materials of the Sensotec/Honeywell transducers
>> (Hastelloy/lnconel C276 and 316SS) are suitable for any organic fluid.
>>
>> - You want differential wet/wet as I specified previously.
>>
>> - My preference would be for 1/4" NPT female fittings on the transducer
>> because this facilitates connection to 1/4"OD x 1/8" ID stainless steel
>> tubing using 1/4" SS-400-6  Swagelock fittings.
>>
>> I'd make a stainless steel flange and mount two SS-400-1-4BT Swagelock
>> bored-through or SS-400-11-4  bulkhead stainless steel fittings on it
and
>> run the tubing through the fittings into the tank. The tube that goes to
>> the bottom should be sealed (pinch it off) at the bottom and a slot cut
in
>> it a known distance up. This is so that you can set the tube to a known
>> depth from the bottom (lower it as far as it will go)and still avoid
having
>> the input clogged with whatever debris may be on the bottom of the tank.
>>
>> Then connect the transducer _outside_ the tank at or below the level of
the
>> intake on the bottom with additional 1/4"x 1/8" tubing and SS-400-6
>> swagelock union. Fill the lower tube with oil by drawing a vacuum on it
>> until it siphons out oil from the bottom of the tank.. My preference
would
>> be to put a Nupro valve in each line near the stainless steel plate ( =
>> bulkhead) so that I could remove the transducers without disturbing the
>> tank.
>>
>> Hope This Helps ... Marc
>> Marc_F_Hult
>> www.ECOntrol.org
>
>Hi thanks and sorry for my late answers, your explanations are
>excelent, thanks so much. I have revised your maths and right. I will
>chose a 10 psi, but the output 4..20ma because the sensor will be far
>(4 or 5 meters). My problems are with the positioning of the
>transducer in the tank, i can't make an idea of how the transducer
>must be placed.  Thanks again, and can i do something for you?


Hi,

This is usenet and the intention is to be useful and helpful, so you don't
get a bill ;-)

What you can do is let the newsgroup know how your project turned out and
any changes or refinements so that exercise is useful to others to (This is
customary 'netiquette which, unfortunately, more often than not is not
followed any more.)

Some comments:

-- 4-5 meters is not very far, so a voltage output would be fine in nearly
all cases. Because you don't need the values more often than about once
every second, it would be easy to filter out or average out any noise there
might be.

-- Read my previous post carefully again with the understanding that what I
suggest is that you put the transducer _outside_ the tank near the bottom.
The transducer is connected to two tubes hat go up to the top of the tank
along the outside, and down into the tank through a flange (or plate) with
the swagelock fittings I described.  One of these tubes goes to the bottom
inside of the tank.   The second tube is open to the vacuum inside near the
top to where there is never fluid.

This approach has several advantages including not having to route
electrical wires into the tank and being able to install and service
without disturbing the tank (if you add valves to the tubes on the outside
as previously mentioned).

-- the 1/4" Outside diameter x 1/8" inside diameter tubing I recommended
has a metric counterpart as do the recommended swagelock fittings
www.swagelock.com  and Nupro valve. This thick-walled tubing typically
comes in straight pieces, rather than in coils as does tubing with thinner
walls. So the thick-walled tubing may be inconvenient to obtain depending
on where you are. Thinner wall would work too with some minor disadvantages
(more dead volume, not as straight and stiff so assuring that the bottom of
the tube is at the bottom of the 5 meter tank may require a weight on the
bottom of the tube, etc).

Hope This Helps ... Marc
Marc_F_Hult
www.ECOntrol.org