Connect with us

Detect state of switches in series

Discussion in 'Electronic Design' started by BenFranske, Jul 23, 2007.

Scroll to continue with content
  1. BenFranske

    BenFranske Guest

    Context: I'm trying to design a little monitoring circuit that will
    tell me when my air conditioning is interrupted by my utility's "Saver
    Switch" which reduces load during peak times.

    Specifics: Central air conditioners normally receive a signal to turn
    on their compressor by having 24 VAC applied to two terminals on their
    control board by the thermostat (a switch). In my case my utility has
    wired another (remote controllable) switch in series with the
    thermostat which can be turned off during periods of high demand
    preventing the compressor from being run. I would like to monitor the
    state of both the thermostat switch and the utility saver switch
    independently. I would also like to monitor whether 24 VAC is being
    applied to the compressor control board.

    Goal: A circuit which will provide three 5vdc logic outputs one which
    shows whether the thermostat switch is open or closed, one which shows
    whether the utility saver switch is open or closed and one which shows
    whether 24 VAC is being applied to the compressor terminals. If any
    one of the three logic points is low the compressor would not be
    running, if all three are high the compressor should be running.

    Does anyone have some suggestions about how to do this? It seems to me
    the tricky part is measuring whether the switches are open or closed
    as they may or may not have 24 VAC flowing through them.

  2. D from BC

    D from BC Guest

    Some sort of optoisolator circuit I guess.
    D from BC
  3. Donald

    Donald Guest

    Ok, the goal is to understand that the thermostat is asking for the air
    compressor to be truned on, but the utility says no.

    What do you expect to find from this information ??

    If the utility says its over loaded and it need to shed, it will shed.

    Please help me understand what it is you are trying to get done with
    this information.
    Use opto isolators and a PIC.

    Write a little software to monitor the optos and send that information
    out the serial port to a PC to check the time of day and graph its usage.


  4. BenFranske

    BenFranske Guest

    Curiosity mostly, I'd like to be able to look at a LED display and see
    whether the saver switch is overriding the thermostat at any given
    time. Sometime down the road it would be interesting for me to
    interface it with a PC which I can setup to log when and for how long
    the saver switch is activated along with what percentage of the time
    it actually prevents cooling. I also happen to believe the utility
    stretches the truth about how much they activate the switch and would
    be curious to know if that belief is well founded.

    About the optoisolators, I did think about those but how might they be
    interfaced into the circuit? Don't they need power on the "input" side
    to operate? When either of the switches is open their isn't any power
    through it to measure. Another way to look at it is that I need to be
    able to differentiate between an open and closed circuit between two
    points and the complication is that there may or may not be 24 VAC
    flowing through that portion of the circuit at the same time.
  5. D from BC

    D from BC Guest

    Ok...I'll give this a shot but I'm no pro.
    It's half baked...others might fix it up..

    Monitoring a switch when the switch control not accessible.
    Industrial control setting.

    ------------/ ------->I may or may not be flowing
    | |
    | |
    | |
    +---[ ? ]----+

    The mystery box [?] contains an oscillator.
    D from BC
  6. D from BC

    D from BC Guest

    Note: It's a signal injection technigue.
    D from BC
  7. Problem is what you want to monitor. As the switches are in series the
    closest to source of 24 supply controls possibility of detecting the
    state of other components as there is NO signal.
    Two possibilities:
    a) change those switches to multiple contact ones and apply detectic
    circuit to another set of contacts on each switch.
    b) put 24vAC relay on outgoing contact of each switch and scratch your
    head to make sense of the signals received.


  8. Donald

    Donald Guest

    Ok, I guess the LED display must be an improvement you are planning for
    in the future.

    My first pass would be to put the LED side of the opto across the loads
    in question.

    The AC relay on the compressor will tell you its running or not.
    The AC relay in your furnace will tell you its looking to cool.

    If the furnace wants to cool but the compressor is not ON, the utility
    must have said no.

    Any voltage on the load will also be on the opto.
    Approprate diodes and resistors will be required to limit the current in
    the LED side of the opto.

    If you want to monitor the low voltage side, you can place your opto in
    series with the switch and if there is any current flowing thru the
    switch, the opto will also come on.


    The output side of the opto would go to what ever circuit you would want
    to use.

    A PIC can be programmed to ingnore the 60 cycle pulses on the output of
    the opto.

    Then you ccan drive your LED display with the other pins of the PIC. ;-)


    Let us know how you come along.
    I may do the same.
  9. MooseFET

    MooseFET Guest

    Lets start by making a DC supply to run it from using the 24VAC from
    the transformer. Since the current will be small a simple 1/2 wave
    rectifier would do for this.

    Next we want to provide a small leakage path past the first set of
    contacts in the string. Since I don't know how much capacitance there
    is in the wiring, lets assume that a 1 meg resistor will do this. We
    also need to provide a path to ground for the node between the
    switches. A 100K and do this.

    ! 1M
    +--/\/\-----+--B C
    ! ! !
    24VAC --+--/ O------++----/ O---+-- Control
    / 100K

    A-----/\/\---->!----+--------- Led Power
    ----->!---+--- LM339 Power

    B---/\/\--+----!-\ LED
    ! >---/\/\----!<----LedPower
    A---/\/\-+-+---!+/ D1
    E !
    ! >----/\/\----!<-----Led Power
    D---!+/ D2

    ! >----/\/\----!<-----Led power
    E---!+/ D3

    D1 = First contact closed
    D2 = Second contact but not first
    D3 = Both closed
  10. BenFranske

    BenFranske Guest

    Thanks for the ideas so far I'm going to ponder this some more.
    Another idea I had was, and I'm not sure if this is workable or not,
    to superimpose a constant DC signal over the AC through each switch,
    use that for monitoring the status and then filter it back out. My
    concern with that is the capacitors needed to filter the DC back out
    again after each switch might be larger than standard values. Any
    thoughts on that?
  11. John E.

    John E. Guest

    Place diode, resistor, and LED across 24v relay coil of your thermostat (the
    "upstream" relay). When the 'stat says "turn on", the LED lights.

    When the AC goes off, just look at the LED. If it's on, you've been shed.
  12. MooseFET

    MooseFET Guest

    Ooops my sketch missed having a zener on the Vcc of the LM339. It
    protects the LM339 and the LEDs from voltage spikes. The dropping
    resistor is shown as intended.
    This sounds way more complex than my idea and you are messing with the
    actual power path through the switches. How about using 10KHz instead
    of DC?
  13. ehsjr

    ehsjr Guest

    TSTAT Saver +---> to controller
    / / |
    24VAC ---+------+---o o----+----o o---+---[100R]--+
    | | | | |
    [4.7K] [4.7K] | [4.7K] |
    | | | | |
    [D1] [D2] | [D3] |
    | | | | |
    [LED1] +---[LED2]--+---[LED3---+ |
    | |
    24VAC ---+------------------------------------------+

    A glowing LED corresponds with an open switch, or in the case
    of LED1, 24VAC. The LEDs should be bright enough, even running
    at ~ 2mA when both switches are open.

    Relay contacts can gate the 24VAC to the controller, if necessary,
    by replacing the 100R with a relay. It will probably work as is.
    Worst case, the controller will "see" about 6VAC when it should
    "see" 0V if you do not use the relay.

    If you must have the 5 volt logic signals, use AC input optos such
    as PC354NJ0000F in place of the LEDs, and eliminate D1-D3.

  14. Jasen Betts

    Jasen Betts Guest

    put a resistor in series with a diode across each switch that'll
    create a small DC signal when the switch opens, if the diodes face
    opposite directions the sign of the DC will tell you which switch is
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