Connect with us

Tone detector?

Discussion in 'Electronic Design' started by Jem Berkes, Jan 21, 2004.

Scroll to continue with content
  1. Jem Berkes

    Jem Berkes Guest

    For a custom application, I'm trying to detect a special audio tone (I'm
    thinking of using DTMF, two-tune custom signal) that originates from a
    computer sound card.

    What's the easiest way to detect this? These are relatively low frequencies
    (within hearing range). Seems to me I could just have two bandpass filters
    and a trigger, but thought I'd ask in case there's an easier way.
  2. John Fields

    John Fields Guest

  3. Hi,

    Here's a cheap and cheerful way -

    Cheers - Joe
  4. Jem Berkes

    Jem Berkes Guest

  5. Jem Berkes

    Jem Berkes Guest

    What's the easiest way to detect this? These are relatively low
    Well I've got NE567N on order... in the meantime, I thought I'd try
    building a tone detector on a breadboard from a bandpass circuit based
    around a single op-amp (LM358)

    It started out ok; I did a manual analysis of the transfer function as
    derived from the feedback circuit, pulled the schematic out of Sedra &
    Smith. Used gnuplot to estimate my center frequencies, built it and the
    filter worked pretty well. I was pretty proud of my math there :)

    But then I added on my post-processing; an RC smoother feeding a
    comparator so I could get a digital "tone present" result. Things started
    to get ugly... the output from the comparator is quite unstable, so I'll
    need to throw on a Schmitt trigger. But even worse, all my frequencies
    are slipping!

    It seems that as I add on circuitry I can't get the tone detector to
    respond to any decently high frequency tone. When it's designed to detect
    a 400 Hz tone it's responding to a 250 Hz tone, and it almost seems to
    max out at around 300 Hz no matter how I push the component values.

    Is this due to my breadboard and capacitive couplings? I don't do much
    high frequency analog stuff, so I could appreciate any insights as to why
    I'm hitting this wall. Would making this on a PCB help much?
  6. I read in that Jem Berkes <>
    Probably not.
    High frequency? You said DTMF originally, which is mid-band audio.
    Please tell us what frequency/ies you are using.
    Unlikely. Your RC network is perhaps loading the output of the tone
    decoder excessively. Post your schematic with component values on
    alt.binaries.schematics.electronics and tell us here that you have done
  7. Jem Berkes

    Jem Berkes Guest

    Is this due to my breadboard and capacitive couplings?
    I'm aiming for the range 100Hz to 1KHz
    I have now posted my calculations, bandpass circuit, and additional
    detector there. That also describes the frequency variations I'm seeing
    (some cutoff, but I'm unsure what's causing it). Thanks!
  8. Nemo

    Nemo Guest

    Oddly enough, I had just discovered the marvellously cheap '567 for my
    own work. Trouble is, when I tried to order it, I found it was obsolete.
    Further research shows, that Philips and natSemi have stopped making it.
    NatSemi have a new lower power CMOS device (LMC567) which is bout 3x the
    price. Anyone know if the old NE567 is still being made by anyone?

    I was interested to see it went obsolete in these 2 manufacturers
    (couldn't spot any others) around 1996 - I guess there was no money in
    it - but also there seems to have been a rise of codecs and other more
    sophisticated chips, as DTMF chips went obsolete (?) or became
    integrated into ASICs, etc.

    Anyhow, I would be most grateful if anyone can point me at a new source
    for the things. I've been burnt by obsolescence a lot lately and want a
    multisourced part which will still be around in 7 years' time (normal
    life of our products).

    Thank you,

    These are surplus, of course.

  10. Si Ballenger

    Si Ballenger Guest

  11. Nemo

    Nemo Guest

    Thank you for the leads chaps.

    However, I need a manufacturer or two - not people who have them in
    stock - so I can be sure the device will still be available in 7 years
    in quantities of 5,000 a year. My company can't afford to guess ow many
    we'll need in that period and bulk buy the lot 8)

    NatSemi do an LMC567, which is more expensive but available. I still
    feel I need a second source to use it. Slightly different emphasis to
    hobby electronics.

    Incidentally, my first LMC567 arrived yesterday, I'll let you know if I
    get it working!
  12. Jem Berkes

    Jem Berkes Guest

    Post your schematic with component values on
    I was going to reply to your posts on schematics but it looks like my ISP
    doesn't keep messages in binary groups for very long! Sorry about that.

    Anyway I understand the issue now with the capacitive load on the op amp
    and ways to buffer it adequately; however this is going to add so many
    components to my circuit that I'm just going with a different solution to
    my original problem all together.

    Specifically, as an alternative to in-band DTMF control signaling I'm going
    to use an entirely different interface for control anyway that's more
    digital in nature. The total components required will be very low in
    comparison, so it's worthwhile.
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day