Multiple sources, highest level detector

Discussion in 'The Projects Forum' started by SteveMorgan, May 14, 2012.

  1. SteveMorgan

    Thread Starter New Member

    May 14, 2012
    4
    0
    I have a need to be able to detect the highest signal from a series of signals, in as simple a way as possible, for a personal project
    As an example, picture say 8 mics, mounted facing outwards in a circular pattern and facing 8 different directions. Simple enough to amplify these individually to detect what they are 'hearing', but if I wanted to be able to tell instantaneously as to which one was hearing the loudest noise, is there a simple way to do this? I can only come up with complex logic that seems way over the top for which I am hoping there is instead a simple logical circuit approach, so that an appropriate led or indicator will light up to say eg Mic1 is loudest, or Mic 2 is loudest or Mic's 5 & 6 are equally loudest.
    Any suggestions gratefully received.
    Thanks in advance.

    Regards

    Steve
     
  2. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    Use multiple active peak detect circuits (consisting of an op-amp and a diode) feeding into a common capacitor (with a discharge resistor across the capacitor).

    The outputs of each op-amp should indicate which of the inputs is at the peak.
     
  3. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
  4. SteveMorgan

    Thread Starter New Member

    May 14, 2012
    4
    0
    Thanks MrChips for your response. Are you able to elaborate a little more? I'm not sure how this would show which sensor has the greatest signal compared with its neighbour at any single point in time? Its crucial that I am able to tell which sensor is currently receiving the larger signal at any one time for directional finding purposes, apologies if this was not made clear in my initial explanation.
    Thanks in advance.
     
  5. SteveMorgan

    Thread Starter New Member

    May 14, 2012
    4
    0
    Thanks again for your response MrChips, good spotting, but no I don't even know the chap, although our purposes do seem somewhat analogous, and I notice that one of the responses also appears to be trying to achieve what I am trying to achieve, (Milesguidon) seems extremely close, although I am trying to find a simpler solution, with perhaps just opamps/comparators and some basic logic gates - if this is achievable, since micros' are outside of my remit at present.
     
  6. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    Here is a simple active peak-detect circuit.

    [​IMG]
    When the input voltage is below the voltage on C1, the SENSE output will be at -Vs.
    When the input voltage goes higher than the voltage at C1, SENSE will go to +Vs and will charge C1 to match the INPUT voltage.

    R1 will discharge C1 in between peak voltages.

    For multiple peak-detects, use one C1 and R1, i.e. all peak-detect circuits share the same C1 and R1. The SENSE outputs will allow you to determine which circuit is peaking at that instant.

    You need to establish your time resolution and time constants desired.
     
    SteveMorgan likes this.
  7. SteveMorgan

    Thread Starter New Member

    May 14, 2012
    4
    0
    Many thanks for this MrChips, I'll run with this and see how far I get.

    Am I correct in assuming that only one Sense(or at least the equal strongest Sense )O/P's will go high at a time, whilst the remaining Sense O/P's will remain low?

    Also, as far as C1 and R1 selection, am I correct in assuming that lower values will give a more instantaneous response?

    Are there any typical guideline values you would recommend that I should use for experimentation for R1/C1?

    I assume that I will not be able to use too low a value or I will end up just following the speed of sound around the microphone array!

    Thanks again for your assistance with this.
     
  8. MrChips

    Moderator

    Oct 2, 2009
    12,442
    3,361
    It depends on what you do with the microphone signals.

    Are you going to use the instantaneous peak signals or will you create an envelope detection circuit for each microphone signal before applying to the peak detect circuit? You need to know what time constants you require. This could be anything from milliseconds to seconds.

    There is no simple straight forward answer to this. You will have to experiment with different RC values depending on what you are trying to achieve.
     
Loading...