Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Signal monitor

  1. Jul 24, 2008 #1

    wolram

    User Avatar
    Gold Member

    I am looking for a cheap device to check sensor function, the sensor is detecting at 120
    rpm and should give a 24v output.
    If there is a circuit i could build my self that would be great.

    Thanks.
     
  2. jcsd
  3. Jul 24, 2008 #2

    berkeman

    User Avatar

    Staff: Mentor

    Are you saying that the output is a 120*60 = 7200Hz sine wave at 24Vrms? Or does it put out a pulse train of 24V peak pulses at 7200Hz? Do you have a DVM with a simple frequency counter in it (my Fluke handheld has a low-frequency counter capability)?
     
  4. Jul 24, 2008 #3

    wolram

    User Avatar
    Gold Member

    Hi berkerman, it is a train of 24v peak pulses, i do not want to leave an expensive meter coupled to this circuit for fear of damage, environmental and human, i need to look at this sensor to rule it out as the cause for a glitch a glich that has gone by the time a meter is set up.
     
  5. Jul 24, 2008 #4

    berkeman

    User Avatar

    Staff: Mentor

    Okay, got it. You want to leave the detecting circuit in-place, not just use it for spot checks.

    What would be considered not working? What are the tolerances on an okay pulse train, in terms of amplitude and frequency? At what point do you want the detecting circuit to call the output no-good? Are you planning on battery powering this detector thing, or do you have wall power available? Or do you want it self-powered on the 24V pulse train, and it just keeps an LED lit if the sensor is putting out a pulse train within tolerances?
     
  6. Jul 24, 2008 #5

    wolram

    User Avatar
    Gold Member

    As long as the sensor is puting out a 18 to 24v signal and not missing any the sensor would be deemed okay, a battery powered device would be ideal.
     
  7. Jul 24, 2008 #6

    berkeman

    User Avatar

    Staff: Mentor

    How about a simple audio output? You could cut off the bottom parts of the pulses and drive a small speaker with the 7.2kHz audio pulses. The ear might be able to hear if there are any consistently missing pulses or whatever the various failure mechanisms are. Don't know if having a continuous audio output would be too obnoxious, though. Are you looking more for an LED indicator?

    Also, how many pulses have to be missing before you light a trouble light? Or is it the frequency changing that you want to alarm? This is all do-able, I'm just trying to get an understanding of how complex it will need to be.
     
  8. Jul 24, 2008 #7

    wolram

    User Avatar
    Gold Member

    I think an LED would be best, say it lights if any pulse is missed, this semsor forms a sort of (and) circuit, the logic and hard wear have been in use for years but now it will have a train of (misses) one day and be okay for may be a week.
     
  9. Jul 24, 2008 #8

    berkeman

    User Avatar

    Staff: Mentor

    Okay, then the easiest way would be to use a timer to time the period between pulses, and if that timer goes above some set count, it trips the circuit that lights the LED.

    So the first part of the circuit would decrease the amplitude of the signal down to logic levels, and make a trigger signal from each pulse which was 18V to 24V in amplitude. That would be fed into the reset input of a free-running counter, so that the counter never reached the alarm count (which it would if a single pulse were missing, or some number of pulses in a row, depending on how you want to configure it). The alarm count would trigger a flip-flop, which would light the LED. The whole circuit can be reset with a single pushbutton, to reset the counter and turn off the LED.

    Does that give you enough to build it, or do you want some more details on how you can make it? I'd start with a 74HC4060 counter IC, with maybe a 5MHz crystal on it, to give you a nominal count of about 5MHz/7.2kHz = 695, and you could use the 1024 count output to trigger the alarm LED flip-flop. Divide down your input pulses with a resistor divider, and run that into a comparator circuit that looks for the 18V minimum equivalent voltage level. The comparator output would feed the reset of the 74HC4060. The whole thing should be able to run off of a 9V battery for a fairly long time (use a low-current LED....).
     
  10. Jul 24, 2008 #9

    wolram

    User Avatar
    Gold Member

    I would much appreciate some detail i am not at a level where i can design a circuit,
    building will not be a problem though.
    Thanks berkerman.
     
  11. Jul 24, 2008 #10

    berkeman

    User Avatar

    Staff: Mentor

    I'll sketch something when I get a chance, wolram. It might be the weekend though. I'll also think about simpler circuits that might do the job -- less accurate, and might require turning a knob to tune the sensing frequency or something. Any chance we can draw power from the pulse train? We wouldn't need much. What's the output impedance of the sensor?
     
  12. Jul 24, 2008 #11

    wolram

    User Avatar
    Gold Member

    The sensor can drive a load, from memory it will power a 1 watt solonoid coil.
    Thanks for the help berkerman.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Signal monitor
  1. Condition monitoring (Replies: 2)

  2. Battery Monitor (Replies: 1)

  3. Generator Monitoring: (Replies: 2)

  4. Energy Monitor (Replies: 3)

Loading...