Hello.
I'm using ADS8344 to digitize signals from the pressure sensors.
After assembly, to assess the quality of the board histogram of codes was obtained:
The resulting histogram has many peaks that occur in odd codes and this remains for all channels.
To confirm that this is not a feature of the signal from the sensors I made resistor divider on the unused channel pulldown R5 and similar 1k wire connected to the chip ref pin.
The effect was similar:
Effect remains at any frequency clocking, in external clocking and internal clocking modes.
Is it true that this histogram indicates the asymmetry of the conversion process - the emergence of the odd and even codes are not equally likely, and thus the data are unreliable and can not be corrected by post-processing? What is the problem?