You can use a common pressure transducer as an altimeter if you employ a special technique to make the transducer respond linearly. If you were to simply amplify the output of a pressure gauge's bridge (V_B in Fig 1) by a gain of 32.7 and an offset of 3.2V, such a circuit would produce an output of 0V at sea level and 1V at 10,000 ft. However, this simple design would read 38.5 mV (385 ft), which is too high at 5000 ft. The nonlinear relationship between atmospheric pressure and altitude causes the error, which peaks at 3.85% of full scale.

You can make the gauge respond linearly to changes in altitude by modulating the gauge's excitation current. Fig 1 shows a circuit that increases the bridge's excitation as output voltage increases.

To respond linearly to altitude, the excitation current must change approximately 4% throughout its calibrated range for each 1% of midscale nonlinearity. For end-point calibration at sea level and at 10,000 ft, the excitation current must change 16.6%.

V₁ controls the current source that IC₂ and IC₃ form in the figure. The source provides 1.5-mA bridge excitation at sea level from the 100-µA current source flowing through the parallel combination of R₁ and R₂. Excitation increases to 1.75 µA at 10,000 ft as the increasing output voltage affects V₁ via R₂. The offsetting voltage, V₂, is adjustable to correct for bridge offset and to compensate for barometric-pressure variations. The linearity-correction circuit reduces error to a small fraction of that inherent in the gauge's pressure-vs-altitude relationship.

You can apply this linearization technique to any sensor whose output is ratiometric with an excitation signal. The correction introduces a nearly ideal second-order term to the transfer function that can correct the simple bowed response common to many transducers. You can modify the circuit for sensors that have a negative error at midscale so that the excitation decreases as output increases. EDN BBS /DI_SIG #1377