I want to detect the output current of a photo diode over a wide range; from 0.1 mA to 10 nA and below in industrial enviroment

To do this, I use the ADL5303 in following circuit:

The photo diode is connecting with a coax cable with shield on GND.

My big problem is the very bad EMI performace against HF-interfereces, because the signal to INPT-pin is very sensitve.

What can I do to to increase the rebustnes against HF-interferences?

  • 0
    •  Analog Employees 
    on Jul 28, 2021 2:52 AM

    Hello george64,

    It's difficult to give specific guidance here, because each design is unique. All we can do is offer general suggestions. First, try to determine the frequency range of EMI that causes the most problem.

    Perhaps a low-pass LC or RC filter on INPT can help. For this type of solution, make sure there is no large filter capacitor directly at the INPT pin to ground, because excess capacitance from INPT to ground will cause oscillation. If the filter design ends with a shunt capacitor, use some series resistance to isolate the capacitor from the INPT pin. 

    You might try temporarily removing the coax cable, placing the PD directly at the log converter IC, on the PCB. If this fixes the problem, it means the coax is picking up the HF and coupling it in. In this case, a different coax or coax termination circuit would be in order. For example, consider improving the shield ground at both ends of the coax. 

    If removing the coax experiment above doesn't remove the EMI problem, it means the ADL5303 circuit on PCB is susceptible to HF radiation. In this case, the PCB layout will likely need to have the INPT trace shortened, or the INPT trace will need a shield. It could also mean an improved PCB layout is needed, for example multilayer PCB with solid ground plane. 

    A good reference is the Henry Ott book, "Noise Reduction Techniques in Electronic Systems".   -Bruce H.