Post Go back to editing

ADG888: Decoupling audio ground from power GND

Dear all,

the following schematic is for processing signals from/to different audio sources.

It uses a couple of ADG888 for the audio switching function, and a simple debouncer (MAX6816) to have a better control.

The question: is it better to keep the audio ground line separated from the main power supply GND ? Or is it mandatory?

Thank you for your idea!

M

audioswitch2.pdf
  • Hello Massimo,

    From a signal perspective, the voltages should not exceed the Absolute Maximum ratings of the part:

    • Analog Inputs, Digital Inputs :−0.3 V to VDD + 0.3 V

    Regards,

    Sean

  • Thank you Sean.

    I am puzzled about  AC signals, I mean symmetric waves around the "zero" level.

    Could a negative input be applied to the ADG888 as reported in the schematic diagram?

    Please note the single +5V power supply.

    Regards

    Max

  • Hello Massimo,

    The ADG888 is a single supply part and cannot take a signal that is below its lowest supply rail, which is GND. Please see my previous reply about the Absolute Maximum ratings. The input pins are protected from ESD using diodes to the rails, when a signal goes beyond the rails, this can turn on these diodes and potentially cause damage.

    Regards,

    Sean

  • Hi Sean,

    many thanks for having driven my attention on this point. Everything is clear now.

    Considering that I want continue to use ADG888 for the project (USB for single power supply requirements are mandatory), can you help me on correcting the schematic in order to allow to bipolar signals to be processed correctly?

    I guess that at least a level shifting to Vcc/2 (= 2.5V) are necessary for each input, is it not?

    Can you suggest a single-chip solution based on AD?

    Many thanks for your help, it is really appreciated!

    Regards,

    Max

  • Hello Massimo,

    You will need to level shift the signals to pass them through the ADG888. Depending on the signal range will determine how much shifting you will require.

    We do have the ADG4613 which has negative signal passing capability. This can be configured to act like the ADG888 by tying D1 and D2 together and IN1 and IN2 together. Similarly, D3, D4 and IN3, IN4 will create the second SPDT.

    Regards,

    Sean

  • Hello Sean,

    thank you for your answer.

    Can you indicate to me an easy application note if available, or a schematic diagram using the ADG4613 supplied by a single +5V for an input of a signal +/- 1.5Vpp ?

  • Hello Massimo,

    There is not a particular way to connect the ADG4613 to allow the negative voltage signal. It will be the same as normal single supply connection:

    • VDD should be connected to your positive rail.
    • VSS and GND should be connected to your ground reference.

    You can see the typical on resistance profile of the switch channel using a single supply at 5V and the signal range range sweeping between -5.5V and 5V in Figure 5 of the datasheet, under the Typical Performance Characteristics.

    Regards,

    Sean

  • Hello Sean,

    thank you for your answer, you're really a lovable person.

    So, an ADG4613 supplied with single rail +5V, can it accept bipolar signals through S1/D1, i.e. +/- 1.5V, is it not ??

    Is it easy to implement the following switching function by using ADG4613 ?

    Sincerely,

    max

  • Hello Massimo,

    The ADG4613 can be configured as a Dual SPDT by tying the IN1 and IN2 pins to the same digital input. Tying D1 and D2 together will then allow you to have the output of D1+D2 = S1 or S2.

    Similarly, you can do the same for IN3, IN4 and D3, D4.

    Regards,


    Sean