I'm using the ADS8328IBPW at VA=5V, VBD=5V, REF+=4.096V, REF-=0V with CFR at the default values, notably Auto Channel Select enabled, Manual Trigger, EOC active low, etc.
I'm getting very poor DC isolation between +IN1 and +IN0, about 1% change in channel 0 or 1 ADC value (more than 600ADC counts) with a full scale change in the other channel's input. I've verified that the REF+, REF-, and channel voltage inputs are stable (except for whichever channel is changing). I'm communicating with the ADS8328 according to Figure 1 in the datasheet (Manual Trigger / Read While Sampling) - from /CONVST pulse to EOC rising edge is about 1.45us, then /CS goes low and my SCLK frequency is about 7.3MHz. The data acquired from SDO looks great - except changing the DC voltage on one channel has the ~1% change on the ADC result of the other channel.
