AVR-Chat

I've lost the original post about this application, but we've had to use 
level sensors here a good bit.

Capacitive sensors will be very temperature sensitive.  This is not just 
a function of capacitor plate physical issues, but the vapor in the vessel.

Vapor in the vessel will also have an affect on ultrasonic and 
radio/radar type level sensors so the vessel temperature is going to be 
an issue.

A float switch will be subject to issues of fluid condition.  If there's 
a chance the fluid may have oil, pond scum, or other types of scuzz in 
it, there's a possibility that this stuff will build up on the float and 
eventually cause it to fail.  It will also be subject to problems if 
there's too much turbulence.  The float can get damaged.

A pair of pressure switches can also be used in a window comparator, but 
with a bubbler configuration.  Run a tube down in the vessel to the 
desired depths.  Run some low pressure air in the tubes.  Set the top 
level switch to come on when the fluid level reaches that high, perhaps 
to turn off a feed pump or turn off a purge pump.  Set the lower switch 
to operate when the level goes below its tube end, perhaps to turn on a 
fill pump.  So with this configuration a pump would come on until the 
level inside the vessel reaches the tube for the top pressure switch 
then turn off when the switch operates.  The pump will stay off until 
the fluid level goes below the end of the lower switch's tube, and then 
turn on again when it does.  The the cycle repeats.  These switches can 
be had for about 15 bucks each, with teflon diaphragms.  The pressure 
source could be something as simple as a 10 dollar aquarium aerator pump.

For a one-off application, is the original poster spending 500 bucks to 
find a 50 cent solution to a 75 dollar problem?  If we put a value on 
time, it could very well be much more economical to buy a 100 dollar 
temperature compensated pressure transducer mounted in the side of the 
vessel toward the bottom.  That would easily buy a stainless steel 
wetted parts transducer.  It wouldn't be accurate to 0.0025% of reading, 
but it would be plenty accurate for production purposes.


REB



David Kelly wrote:
> On Wed, May 13, 2009 at 09:47:39AM -0700, Bruce Parham wrote:
>   
>> Sorry, my earlier response was incomplete, just the germ of an idea.
>>
>> Since the inverted standpipe is initially filled with air, when the
>> liquid level rises, the gas will be compressed allowing liquid to
>> enter the tube which will reduce the pressure head. The actual
>> pressure measured will depend on the total volume of the tube and the
>> amount displace by liquid but will always be less than the pressure at
>> the bottom of the tube.
>>
>> -- The gas compression correction is left as an exercise for the
>> student. --
>>     
>
> All to avoid purchase of a sensor which won't be damaged by the liquid?
>
> The compressed air in the inverted standpipe will eventually dissolve
> into the liquid. This proposed solution will not work in a static
> situation, must be cyclical allowing the air in the standpipe be
> refreshed periodically.
>


[Non-text portions of this message have been removed]