You are putting this into a concrete plant?  You should really use some sort of
industrial solution.  A number of companies make non contact and contact
distance sensors specifically for liquid level measurement.  The one I am
thinking of now for your application is a non contact laser liquid level sensor,
which you simply point down on the liquid.  Many of these can be used in
corrosive environments (glass, stainless steel, etc) as well.

As far as the water content of sand, the only thing I can think of would be to
take a certian volume of the sand, and weigh it.  The same volume will be
heavier the more damp it is.  I don't think the usual method of electrodes
spaced so far apart will be very accurate (nevermind having to replace them
every month, and recalibrate them daily due to wear) Most dampness/humidity
sensors require contact with the substance you are measuring.  Perhaps you can
tell us more about the properties of wet sand.  Does it reflect a lightsource
differently depending on its dampness?

-Adam

Hugo Jorge M|ller wrote:
>
> Hello
>
> I am beginning to work in the final project of my career of Electronic
> Engineering and at the moment, we have decided to make the following work
> (with a friend): To automate a plant of Concrete.
>
> The idea is to use a PC for the general control of the plant and to arm a
> network (Possibly RS485) with terminals that will be slaves working with
> PICs, to complete different tasks, as the following ones:
>
> 1) To control the pneumatic servos and read proximity sensores and other.
>
> 2) To Sense the weight of the materials with cells of loads, indicate the
> value measured in a  7 segments display and to send the measures througt the
> network when they are requested or periodically. This partly would be
> resolved, but I accept ideas and suggestions in this respect.
>
> 3) To Sense the level of water in the main dosificador tank. The tank should
> be filled until certain level and then it is pumped to the truck concrete
> mixer. I have think in sense the height of the column of liquid with a
> sensor of pressure Motorola of the series MPX, I also thought of the
> ultrasonic solution  but I don't have idea that resolution i can achieve.
> The tank in question has 1100 Liters and a height of about 2 meters. aprox.,
> it is cylindrical. I need a precision of about 5 liters. (or better).
> Ideas? Comments? Other viable methods?
> On the other hand I should dose some aditives, special liquids that
> unfortunately they are corrosive and at the moment it is filled a small tank
> of about 26 liters by means of a key, for graveness, then they closes this
> key (superior) and two keys open up, one in the inferior part that allows
> the exit of the preservative toward the MIXER and the other one in the
> superior part that injects compressed air at 10 Kg/cm^2 to impel to the
> preservative and to pump it without having contact with the same one.
> The problem is to measure the level or the quantity of this aditive without
> touch it and to be able to pump it too to the mixer.
> I thought of measuring it with sensor MPX, but the problem is when opening
> the key of compressed air, the sensor doesn't support 10 Kg/cm^2 (although
> it could measure in differential form, but I suppose that the transitory one
> provoked, it would destroy to the silicon sensor).
> Here it is where I request them help that mensuration methods advise me?
>
> Somebody had worked with sensor MPX with PICs ?  What kind of interface/ ADC
> use ?
>
> 4) Another topic not less difficult of solving it is to be able to measure
> the relative humidity that there is in the sand and in the concrete once
> blended. Here the only method that I am happened (electronic) it is between
> two by means of the mensuration of conductivity electrodes of a material
> that is not oxidized, located inside the material to measure, at a certain
> distance. That other methods or variantes suggests me?
>
> As for the net RS485, has somebody used them with PICs? That protocols
> suggest? Some link with information?
>
> Good, I won't extend more in the detail, I wait their collaboration with
> ideas...
>
> Sorry if there are errors in the text, but my language is Spanish  :-)
>
> Thank you from already.
>
> Hugo J. M|ller
> H.J.M. Hardware & Software Diseqos Electrsnicos
> San Nicolas 683
> Tel-Fax : 54-(0)-343-424-5953
> (3100) Parana (Entre Rmos) Argentina
> Email : hmuller [ANTISPAM-QUITAR-ESTO] @arnet.com.ar
> Web Site: http://www.pagina.de/hjm
> UIN (ICQ) : 38.605.074