Chattering Relay

Discussion in 'General Electronics Chat' started by Robert.Adams, Feb 16, 2010.

  1. Robert.Adams

    Thread Starter Active Member

    Feb 16, 2010
    112
    5
    I have two power supplies, a voltage doubler, and a square wave generator.

    I am using the square wave generator to boost the output of one power supply from 50V to ~96V. The other power supply has two outputs and a fixed 5V output as well.

    I have several SPST relays with 5V coils that can handle up to 250VAC. I am trying to make a configuration that chatters an output signal between the ~96V and 12V. I have been having trouble getting relay chatter without getting overvoltage protection shutoff from the power supplies.

    Can anyone help me?
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    If your load is reactive (ie: inductive as in a coil or motor, or capacitive), then that is where your problem lies.

    Post a schematic of how you have things wired up presently.
     
  3. Robert.Adams

    Thread Starter Active Member

    Feb 16, 2010
    112
    5
    I've attached the basic wiring for the circuit. I believe the load is reactive, but I'm not sure.

    I have not included the wiring for the chatter because that is what is tripping me up. I can't find very many resources for inducing relay chatter, most are about reducing relay chatter. I've tried various combinations to get it to chatter and every time I find something that works, my power supplies shut down due to over voltage protection.
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Try something like this:

    [​IMG]

    R1 limits the maximum charge current for C1 to 1/2 Ampere; much more than that and you'll burn up the contacts pretty quickly. The size of C1 will determine your relay 'chatter frequency'. RLY1 can be a SPST NC relay, but not a NO relay. You will have to experiment to determine a good size for C1.

    D1 and D2 take care of the reverse EMF which occurs when current through the relay coils is stopped.

    The contacts of RLY1 provide an intermittent ground for the coils of both RLY1 and RLY2.

    RLY2 needs to have two poles. The output is connected to the COM terminal.
     
    Robert.Adams likes this.
  5. Robert.Adams

    Thread Starter Active Member

    Feb 16, 2010
    112
    5
    I only have access to G6RN relays. The ones I have are SPST-NO. I've decided to use a function generator to do the switching instead of making them chatter. At this point I've hit the problem where I can only get 29V under load (instead of the ~96V).

    This could be a problem where I need to make the switches break the connection before making the next. I've tried adding a cap/diode combination in parallel with the 96V relay coil in order to slow down it's switching speed, but I still have this problem.
     
    Last edited: Feb 24, 2010
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    They are also available in SPDT. The SPST-NO will not work very well, as you cannot guarantee that one will open before the other closes. You need "break before make". With the SPDT relay in the configuration I've shown, it is physically impossible for both supplies to be connected to the load at any given point in time.
    That may be a limitation of your 96v supply current, inductance from wiring, or the contacts of the relay having high resistance by getting burned from continual make/break cycles under load.

    As I've mentioned, the SPDT relay configuration will eliminate the possibility that both supplies are connected at the same time. However, you will still be plagued by "contact bounce", and slow response times. If you want to eliminate those problems, you will need to abandon relays and use a semiconductor solution.
    Relays switch quite slowly compared to semiconductors.

    I don't know what you're actually trying to accomplish, nor what your load is.
     
  7. Robert.Adams

    Thread Starter Active Member

    Feb 16, 2010
    112
    5
    It is actually for testing transient voltages on a module I have for work. I know the SPDT variation exists, but SPST-NO is all they are giving me.

    I am basically trying to switch between operating voltage and almost 100V as quickly as possible. Ideally I'd like 100 uS spikes, but I could settle for longer.

    The relay method is what my mentor (I'm a co-op) advised me to take.

    EDIT: The voltage doubler is controlled through a square wave which allows me to control transistors that charge a capacitor up to the target voltage. I am told it has very little capacity.
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    OK. What is the expected current when the 96v is applied?

    I think you can avoid much of the problem by simply adding in a couple of diodes to prevent the supplies from 'seeing" each other; one in the +12v supply line, one in the +96v supply line.

    If you wanted to do something really "sneaky", you could use a transistor to switch the low side of the relay coil to ground briefly, ignoring the contacts entirely - as a simple boost coil.
     
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Here, try it like this:

    [​IMG]
     
  10. Robert.Adams

    Thread Starter Active Member

    Feb 16, 2010
    112
    5
    I couldn't get the relays to switch at 100 Hz (the datasheet lists a typical release time of 3 ms and an operate time of 6 ms). I bumped it down to 60 Hz, but I still get a max of about 30V. It is more noisy this time, but it doesn't really get much higher than before.

    I think I need to revisit the doubler design and see if I can get a more robust 100V source. Thanks for all the help, I know a ton more about relays now.
     
  11. blueroomelectronics

    AAC Fanatic!

    Jul 22, 2007
    1,758
    98
    What is this device for?
     
  12. Robert.Adams

    Thread Starter Active Member

    Feb 16, 2010
    112
    5
    Testing transient voltages on an automotive electronics module. It is a make-shift EMC test.
     
Loading...