I'm building a water tank depth monitor using Freescale's 5050dp pressure sensor:

http://www.freescale.com/files/sensors/doc/data_sheet/MPX5050.pdf

I'm using the circuit in the data sheet that uses the capacitors on the input and out.

I'm running a tube from the input port of the sensor from the top of the tank down about 5 ft into the tank. The readings I'm getting have a constant drift downward even with a constant tank level. (I duplicated this behavior with the same circuit with the tube sitting at the bottom of a pool.)

The simple solution would be that there is an air leak in the tube but I would expect an equilibrium to be reached but that isn't the case. There also doesn't appear to be water in the tube. The problem appears to be electronic, not mechanical.

Any advice would be appreciated.

 

If the pressure goes down monotonically indefinitely then you'll eventually conclude that you are in a vacuum. If the pressure goes down for a while, then up for a while and repeats the cycle then that would be expected. Since liquids are incompressible they will transfer atmospheric pressure to a submersed sensor.

 

The readings I'm getting have a constant drift downward...

As Papabravo notes, without knowing now much drift you see for how long, this may not be a problem. You might also be seeing temperature compensation kicking in slowly.