AD5940: Crosstalk and µA-level transient current on unselected switch matrix channels during FCA measurements

Hi,

We are using the AD5940 for Fast Chronoamperometry/FCA measurements with GUI-selectable switch matrix routing. We are observing a crosstalk/fake current issue between switch paths even when the selected measurement channel is not physically connected.

For example, when the sensor/electrode is physically connected to the SE0 path, and the GUI/firmware selects the DE0 path, we still observe a transient current response in the µA range. Ideally, since DE0 is not connected to the sensor, we expected the measured current to be close to zero or only in the nA leakage range.

We verified that the GUI is sending the expected switch values:

SE0 selection:
N = 0x00000100
T = 0x00000010

DE0 selection:
N = 0x00000020
T = 0x00000020

The issue is also seen between AIN2 and AIN3, where an unconnected AIN channel can still show a fake FCA-like transient. The waveform looks like a charge-injection or capacitive-coupling transient rather than true sensor current.

Our current understanding is that CE0 and RE0 may still be active while the selected WE path is changed, so the driven sensor connected to SE0 may be coupling into the unselected DE0/AIN path through the switch matrix, board parasitics, or HSTIA input path.

Could you please confirm the recommended switch matrix configuration for this use case?

Specifically, we would like your guidance on:

1.For FCA/chronoamperometry, what should be the correct D/P/N/T switch configuration when selecting SE0, DE0, AIN0-AIN3, and AFE3 as the working electrode/current input?

2.When testing channel isolation, should D and P switches also be opened or routed to unused AFE pins, instead of keeping CE0/RE0 active?

3.Is it expected that an unconnected DE0/AIN input can show µA-level transient current if CE0/RE0 are still driving a connected cell on SE0?

4.What is the recommended sequence when switching channels?

       - Open all switches

       - Power down AFE blocks

       - Reset HSDAC to midscale

       - Wait for discharge/settling

        - Apply new switch matrix

        - Start ADC/HSTIA measurement

5.Are there any specific precautions for minimizing crosstalk between SE0/DE0 and AIN0-AIN3, AFE3_LOAD in AD5940 switch matrix measurements?

Please let us know if this behavior is expected due to internal switch capacitance/charge injection, or if there is a preferred routing/discharge sequence we should implement in firmware.

Thank you