ADBMS2950BCCSZ datasheet(43/97 Pages) AD

43 / 97 page

Data Sheet

ADBMS2950B

Rev. 0 | Page 43 of 97

18-bit. However, because the peak-to-peak noise is significantly

larger than 1 LSB, an accumulation based on multiple Ix

conversions, nevertheless, yield an effective resolution that is

higher than 18-bit (dithering). The random noise of I1ADC is

uncorrelated to that of I2ADC. The random noise is also

uncorrelated between consecutive 1 ms IxADC conversions.

Figure 32. IxADC Precision Depending on Number of Averaged Conversions

at Zero-Input

Figure 33. IxADC Noise Distribution of the 1 ms Conversion Results

For larger input signals, the accuracy is also determined by the

gain error, and the noise becomes a relatively smaller error

term. Figure 34 shows the expected IxADC output for a 10 mV

input signal, ±0.1% gain error limits, and the ADC output

including noise.

While averaging still improves the peak-to-peak reading

somewhat, the TME is dominated by the ±0.1% gain error term.

Averaging with up to 32 conversions can be done through the

on-chip accumulators configured by the ACCI setting.

Averaging of more conversions must be done in the host.

Figure 34. IxADC Precision Depending on Number of Averaged Conversions

at 10 mV Input Signal

In general, the IxADC TME can be decomposed in a precision

and an accuracy term. Using the electrical specifications, the

TME can be calculated as a function of the input signal with the

following equation:

TME = offset + signal × (gain error + nonlinearity) + noise.

The noise can be improved by averaging, while the accuracy is

determined by the ADC specifications. Figure 35 shows TME

vs. input signal for four accumulation settings: ACCN = 1,

ACCN = 8, ACCN = 32, and an average of 64, which must be

computed in the host controller, as the on-chip accumulator can

only be set to a maximum of ACCN = 32. This requires the host

to calculate the average over two consecutive IxACC results.

This calculation again uses a 3.5σ peak noise estimate, which

means that 99.98% of all conversion have a lower noise value

than the curves shown in Figure 35. As the total measurement

time decreases or increases, the curve approaches the limits set

by the offset plus noise or gain plus nonlinearity error

specifications.

Figure 35. IxADC TME vs. Input Voltage

ADBMS2950BCCSZ Datasheet(PDF) 43 Page - Analog Devices

ADBMS2950BCCSZ Datasheet(HTML) 43 Page - Analog Devices

Html Pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97

Datasheet Download

Link URL