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.

A Scheme to Nullify Con in Analog Mux

Other Parts Discussed in Thread: TS5A2053

Someone tell me what is wrong with a scheme I devised to nullify the Con of an analog mux (single, 6-8 pin part such as TS5A94411 or TS5A2053). 

In my application the capacitance of the analog mux is affecting the frequency response of my signal above 100MHz.  It occurred to me that I might put a 3K resistor in series with the Vcc, Gnd and the control input and then a bypass cap between Vcc-Gnd and Cntr-Gnd.  I would also clear any copper from below the part.

It seems to me that the signal would now see at least 1k ohms in series with the Con and thus Con would not affect my signal.

The isolation spec implies that the channel to channel capacitance is less than 1pf.  The part only draws 1uA of supply current.  I might need to add a 5V zener across Vcc-Gnd to account for effects of charge pumping through the protection diodes but if the bypass cap is large relative to Con I don't think I will.   

I can't find the flaw in this scheme; but, it seems too easy.

  • Hi Paul,

    Can you provide a sample schematic or block diagram for this proposal? We can review your idea using the drawing as a reference.

    Regards,

    ryan

  • The schematic as requested:

  • Hi Paul,

    The design will not have much of an effect on the input capacitance because it only addresses the input-to-ground capacitance.

    In addition, an ESD spike will have a more adverse effect on the part due to an elevated ground.

    -Ryan

  • Ryan,

    You either did not look very carefully at the schematic or did not understand the problem.
    The problem I am attempting to solve is signal frequency roll-off caused by the Con of the switch.  The Con rolls off the signal because at high frequencies it provides a low impedance path to "something" outside of the part.  This "something" must be through a capacitance inside the part  to a pin or possibly to copper underneath the part.  Any capacitance to a high impedance node cannot have an impact on the signal so I made all of the pins high impedance to the outside world.  The added capacitors are to ensure that the pin to pin relationships remain low impedance.
    Essentially I am "floating" the switch at high frequencies.
    In the schematic there is a series resistor on nearly every pin - the ground pin, the power pin and the control pin.  The only pin left is the other signal pin (off) and the crosstalk spec implies less than 1pf from the signal path to that pin.  The added capacitors connect only to other nodes which are also isolated from the outside.  
    Since there is no path for the Con to reach the outside world it cannot affect my signal.
    ESD will not be an issue in this application.