Hi Benny,

For some reason I was feeling quite at home reading your response !

In my case I mostly use the (static) pressure transducer as an input for data logging - as you rightly said +/- 0.5 Bar in a 250 Bar span is just fine for most applications and that is the reason I get away with the setup as shown by me. This is OK for plots in the time domain - you get a very decent plot particulary when a moving average of about 3 to 5 samples is used.

However in a new application, I need to plot the pressure values in the frequency domain with the data captured by a dynamic pressure transducer. The signal is conditioned by a matched charge amplifier and the frequency bandwidth lies in the region of 50 to 1500 Hz. The pressure variations that are captured by the transducer ( being dynamic ) are only in the range of 0-20 Bar and hence the worry on noise. ( This is a cold start analysis of a hydraulic power steering when starting a vehicle after it is exposed to -10 deg. C. Pressure peaks happen for about 10 seconds till the very cold trapped fluid circulates out of the steering gear / actuator. )

Based on the suggestions given already in this thread I think I will go in for a charge amplifier with three port isolation. And another option would be to look at digital transmission as opposed to analog transmission ( as per Erik ).

Raghu