I need a little help with shielding for ECG leads.
The design I'm working on has 3 leads, about 60mm in length. We use shielded cable for each lead, which all come to a meeting point at the connector. I am planning to add a TVS diode to each lead. The device is battery operated. Another engineer and I have different opinions about how that diode should connect on the anode side and I'm hoping you can help with the optimal design:
Option 1: ECG lead shield is not connected to the main GND directly, TVS diode anode connects to the ECG lead shield, shunting any EMI to the shield.
Pro: Theoretically, EMI would discharge directly to the shield and never affect the main circuit GND, would avoid GND bounce.
Con: Shield voltage could be at a different point from board GND and therefore pull the ECG signals out of the dynamic range of the differential amplifier on the main board, making the device unable to properly capture ECG. Normal dynamic range is +/-300mV so it is quite sensitive.
Option 2: Shield is connected to main board GND directly, TVS diodes would shunt EMI energy to shield AND GND (same potential).
Pro: Shield is at the same potential and the main circuit GND, ECG signal will always be at voltage that is referenced to GND and therefore stay within the dynamic range of the OpAmp.
Con: Any noise coupled to the shield will also be coupled to the GND, could cause GND or signal bounce.
Option 3: Connect the shield to GND via 1Mohm resistor in parallel with a 0.01uF cap.
Pro: Seems like a potential middle-ground solution. The shield is still referenced to GND potential but is not directly connected to GND itself, minimizing transfer of EMI to the main circuit.
Con: Decoupling capacitance has to be chosen carefully (too large could cause GND bounce, too little might allow shield voltage to vary drastically), should could still be at a different potential for a brief period of time.
I am very curious to get some input on this especially from people who may have faced this problem before or have some good understanding of interactions / pitfalls with wire shielding and EMI protection.