Other Parts Discussed in Thread: REF200
Hi,
I am currently dealing with two separate issues with the PCB i have designed using the AFE4404, with schematic very similar to AFE4404EVAL board
1. in order to verify the consistency of the electronics and monitor the drift overtime, i have physically disconnected the PD ( not through software command), and LED source, then set the dac current to setting:1 and gain to Setting 0, while monitoring the ADC output value. I have abserved the ADC voltage to drift from .25275 to 25265 over one day of data collection after running moving average on data. this data is LED1-LED2 value since my application only requires one LED, so im using LED1 as the main light source and LED2 to read ambient value for subtraction purposes. any suggestions on what might be causing this drift? the value of the drift is insignificant as long as it doesnt get worst through time. is there anything on the front end that could be causing the drift? for example the BAV99w ESD diodes on the front end of EVAL board design, could they create a current leakage situation that inject current in the photo diode node? Im thinking since the input is fully differential, even if the the diodes above have a relatively high reverse bias current they would cancel each other out.
2. when i connect the LED back in the circuit ( while keeping Photo diode physically disconnected still) i see a jump in my ADC voltage output from .25265 to 25285. again the amount is small but i dont understand how the LED current finds its way back to the IN+, IN- analog traces as PD is not connected to the circuit. on possibility is the high dV/dt voltage on the TX trace that could capacitively get coupled to such traces, but the rise in adc voltage seems to be DC and not noise.
3. transimpedance amplifier circuitry usually requires guard ring implementation to deal with current leakage issues in discrete designs, however i cant find any such recommendation on AFE4404 layout documentation. please advise
Regards.
Ramtin
