I’ve been building a spectrophotometer the last few months and am confused by the ADC’s behavior regarding grounding. While everything appears to be working properly, including the ADC output, I have noticed that the ADC still outputs values, proportional to the voltage, when the circuit isn’t grounded. Depending on the circuit and voltage being measured, the ungrounded ADC varies by 0-33% compared to grounded ADC.
To investigate, I started looking for a ground source and determined it be the RPI’s clock pin. Whenever I connect the ADC’s clock to the RPI clock pin, the circuit is powered—-to some extent. LEDs will light up to nearly full brightness, op amp starts amplifying signals and the ADC outputs values that correspond to the input with varying deviation from grounded circuit.
My questions are:
Is the clock pin on the RPI tied to ground when not pulled high? Or is my RPI clock pin burned out?
Is the MCP3008 clock pin internally tied to one of the ground pins—-AGND or DGND?
What would be the best way to isolate this grounding effect on my board? I don’t want the entire board receiving power unless I have my board’s negative pin connected to ground. here’s my circuit diagram