This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Separate Ground Planes for Two 656/657 Based TIA Circuits?

Other Parts Discussed in Thread: INA157, OPA656

Hi all.  First  I'd like to thank the previous posters that have answered my questions regarding my TIA circuits. It has truly been helpful and appreciated.  So as I move from breadboards to PCB, my plan is to layout my two circuits (reference and measurement detectors) on separate boards. But I also want to create a larger composite board with both reference and measurement circuits on the same board.  Each circuits will consist of two 656s or 657s feeding the inputs of an INA157 difference amplifier.  So the larger board will consist of a total 4 656/657s and two INA157s. 

We want to keep the isolation between the two circuits high, so my question is does it make sense to create separate ground planes on the bottom layer of the composite board for each circuit and then connect them at one point near the power supply inputs?  And if I were to do this, does it make sense to forgo a ground pour on the top layer and simply via the ground pins down to their respective ground planes.  My concern with removing the top ground pour is losing the isolation that it provides.  Is it possible to still use a split ground plane on the bottom PCB layer, where all the ground pins are connected with vias, and then use a ground pour on the top layer but only connect it to the bottom pours where the two bottom planes meet?


Lastly, with regards to 50 ohm SMA connectors.  Having worked in the microwave world I understand the importance of impedance matching for maximum power transfer.  But my circuits are designed to work in the 200kHz-500kHz range, so impedance matching isn't really a concern.  I calculated that the necessary trace width for the PCB board is approximately 110 mils, which is pretty wide given the pin pitch.  I could miter down the traces as I approach the chip, but I'm not sure if it's really necessary.  I don't think the input impedance of the op amp circuits are 50 ohms anyway.  So given my operation frequency should I really be concerned with impedance matching? 

Best,

Dave

  • Hello David,

      What I would do for the GND planes is have a split GND plane for the reference and measurement circuit which is connected only close to the source. I would do this split GND plane on both the top and bottom layers, however I would put some separation between the two parts of the circuit to minimize cross-talk. Also I would connect the planes at multiple point siwth stitching vias. Also, I would remove the GND plane from below the trace that connects to the inverting pin of the amplifier and also from below the DUT itself to minmize any excess capacitance. Please look at the layout guidelines in the datasheet of the OPA656/657.

    With regards to the 50Ohm traces, as mentioned previously we really want to minimize the capacitance on the inverting pin and having such a thick trace could be an issue.

    Samir

  • Samir,

    That's really helpful.  I really appreciate it.

    Dave