MCP3919 datasheet(61/88 Pages) MICROCHIP

61 / 88 page

 2014 Microchip Technology Inc.

DS20005347A-page 61

MCP3919

At low-signal amplitude values (typically 1000:1

dynamic range and higher), the crosstalk between

channels, mainly caused by the PCB, becomes a

significant part of the perturbation as the measurement

error increases. The 1-point measurement error curves

in Figure 2-5 have been performed with a full-scale

sine wave on all the inputs that are not measured,

which means that these channels induce a maximum

amount of crosstalk on the measurement error curve.

In order to avoid such behavior, a 2-point calibration

can be put in place in the calculation section.

This 2-point calibration can be a simple linear

interpolation between two calibration points (one at

high amplitudes, one at low amplitudes at each end of

the dynamic range) and helps to significantly lower the

effect of crosstalk between channels. A 2-point

calibration is very effective in maintaining the

measurement error close to zero on the whole dynamic

range since the nonlinearity and distortion of the

MCP3919 is very low. Figure 2-6 shows the

measurement error curves obtained with the same

ADC data taken for Figure 2-5 but where a 2-point

calibration has been applied. The difference is

significant only at the low end of the dynamic range

where all the perturbing factors are a bigger part of the

ADC output signals. These curves show extremely tight

measurement error across the full dynamic range

(here, typically 10,000:1) which is required in

high-accuracy class meters.

MCP3919 Datasheet(PDF) 61 Page - Microchip Technology