SBASAE3 December 2025 ADS125H18
PRODUCTION DATA
Figure 8-5 illustrates a setup for a common channel-to-channel crosstalk test procedure. Each even numbered input channel (examples: AIN0, AIN2, AIN4, and more) is driven by an individual voltage source. All odd numbered channels (examples: AIN1, AIN3, AIN5, and more) are shorted together to ground.
The test procedure is defined as follows: Apply 3V constant signal to the input channel under test (example: AIN0), and apply the "interference" pattern shown in Figure 8-6 to all remaining even numbered channels (examples: AIN2, AIN4, … AIN14). Measure the ADC output of the channel under test (example: AIN0) to determine transient deviations from a constant output code. Repeat this procedure for every even-numbered channel. Commonly, the variation in output code is desired to be ±1 LSB on the 16 bit level, or smaller.
Figure 8-7 and Figure 8-8 depict the measured results for the crosstalk test using the setup from Figure 8-5.
All inputs except the input channel "under test" undergo large transient voltage steps as described by the test procedure above. The impact on the channel under test, for example AIN0 in Figure 8-7, is visible in the zoomed plot for AIN0 output codes (see secondary y-axis). In all cases, the deviation from the ideal output is approximately 70 codes (LSBs) on the 24 bit level or less, which corresponds to 1LSB on the 18 bit level.
All other channels perform similar to AIN0 when undergoing the same test. As another example and to demonstrate the similarity between channels, the impact on channel AIN6 from crosstalk is also shown as seen in Figure 8-8. All other channels (AIN2, AIN4, AIN8, AIN10, AIN12, AIN14) are tested using the same procedure, and crosstalk in the order of 18 bit level or less is confirmed.