Tool/software:
I have been testing the ADS8688 in our circuit and I am noticing some strange behavior that I cannot figure out. I have external RC filters at each input similar to what is shown in Fig. 95 in the datasheet. We completed initial testing involving setting the inputs to known sine waves with +/- voltages at various frequencies from a signal generator. I was able to confirm correct sampling on all 8 channels.
Now I am testing with two types of motion sensors. One of the sensors has an isolated ground, so the positive signal inputs are connected to the AIN_xP pins and the negative inputs are connected to the AIN_xGND pins (the negative input is not connected to AGND or GND of the ADS8688). This works well for all three inputs tested (CH0, CH1 and CH2).
Then I added a 3-axis accelerometer sensor with single-ended outputs where the signals are referenced to AGND or GND of the ADS8688. I connected the positive outputs AIN_xP on CH3, CH4 and CH5. I also connected a temperature output to CH6. The negative outputs (AGND or GND) are connected to the AIN_xGND pins. When I test this configuration, I get correct results on the first three channels, but the remaining channels all indicate the same voltage measured on CH2.
I then tried reversing the two types of sensors, so the single-ended sensor on CH0 through CH3 and the sensor with the isolated ground on CH4, CH5 and CH6. With this configuration, the first 4 channels show correct results and channels 4, 5 and 6 all show the voltage indicated on CH3.
So each type of sensor is working and I confirmed this independently as well. But when I try to mix signals with different ground references (isolated and GND), the ADS8688 is failing to read correctly.
Can anyone please help provide some insight to what is going on here?
Thanks
