Simple AC delay, perhaps filter?

Discussion in 'The Projects Forum' started by AceofAllan, Jun 16, 2011.

  1. AceofAllan

    Thread Starter New Member

    Jun 16, 2011
    7
    0
    Greeting AAC members,

    I'm the EE for a small company that deals with high power resistor products (load banks, heaters, DB systems). Most of what we do is very simple, but today I'm trying to replace a relatively expensive time delay off relay with some cheaper components. Apparently I've been away from some of the more complex circuits for too long, because my initial ideas have failed to do what I want.

    I think all I need is a filter of some sort. In our application, we have a pressure switch feeding a relay that needs to be activated before power reaches the resistive elements. Sometimes the fluctuation causes our DP contactors to chatter and then fail, fusing shut. Ideally, a brief fluctuation (2-5 seconds, doesn't need to be precise) would not actually shut off the contactors.

    The control circuit is all 120v AC power. I thought a simple RC circuit in parallel or series would delay the relay from turning on and off. I've picked a few values to play around with, but nothing has worked the way I want. Could it be that I just don't have a high enough time constant?

    Thanks,
    -AceofAllan
     
  2. #12

    Expert

    Nov 30, 2010
    16,354
    6,851
    Trying to do a delay of 100 to 300 cycles of the power wave seems absurd on its face. I suggest you go backwards to the hydraulic system and add an air column time constant. Otherwise, you will be working in DC to get several seconds of delay.

    It can be done in AC, but not the way you are thinking, and not simple or cheap.

    Of course, this answer assumes a lot that you didn't tell, so I might be all wrong.
     
    AceofAllan likes this.
  3. AceofAllan

    Thread Starter New Member

    Jun 16, 2011
    7
    0
    I appreciate your answer, but can I ask what more information you would need?

    The pressure switch is actually being used to gauge whether we have airflow over our resistor elements to keep them cool. If the air intake becomes blocked, the pressure switch drops out and the load is dumped from our resistors to prevent overheating.

    Is there anything else I can clarify for you?
     
  4. AceofAllan

    Thread Starter New Member

    Jun 16, 2011
    7
    0
    You sparked another idea for me that I'd like to ask about: I could use a bridge rectifier to create a DC voltage, feed that into an RC circuit that eventually activates a transistor of some sort. The transistor then allows AC current to flow through my relay. The rectifier would be fed by the pressure switch, so when pressure drops out, the capacitor in the RC circuit keeps the transistor activated for a few seconds, acting like the filter I want.

    Am I correct in remembering transistors this way? Can I use a DC voltage to activate it, but pass an AC current through it? I think I remember that I have to use two transistors so that the AC could flow both directions.

    I'm prepared to do most of the work myself, but I'd like to know if I'm at least going down the right path here. Will this work?
     
  5. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,395
    1,607
    I think I get the current system: Pressure Switch controls AC to a relay, and you want a small delay between the pressure switch opening and the relay opening.

    The problem is the transistor isn't good for turning on an AC current for the relay.

    Can you change the relay to a DC relay? Or just add an extra DC relay to the chain? That way the pressure switch turns on & off a small DC supply (like a wall-wart) that in turn powers the DC relay; you can add some capacitance there which will give you the delay. Then the DC relay just turns the final relay on & off.

    (BTW, I happen to be one of the world's authority on military grade time delay relays, as I designed the new replacement controller for MIL-PRF 83726, otherwise known as "RELAY, HYBRID, TIME DELAY (ON OPERATE OR RELEASE), CLASS B, TYPE I, HERMETICALLY SEALED, DPDT, 10 AMPERES, ADJUSTABLE TIME DELAY (EXTERNAL RESISTOR), 0.1 TO 500 SECONDS." But they work off 28VDC and I don't have the relay itself, we just make the controller for several relay manufacturers.)
     
    AceofAllan likes this.
  6. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    The answers to the following questions will aid us in providing accurate solutions to your problem.

    What powers your resistor loads, AC or DC?
    What are the contacts rating of your relay?
    What is the relay coil current value?
    What is the maximum current that any load would see?
    What is the maximum voltage supplied to the load resistors?
    Are the load resistors powered from the high side or low side?
     
    Last edited: Jun 16, 2011
  7. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Could you use a commercially available AC-powered time delay relay?
     
  8. AceofAllan

    Thread Starter New Member

    Jun 16, 2011
    7
    0
    You are envisioning my system correctly. Regarding the transistor, I was afraid of that. Yes, an option would be to use a DC relay. This is starting to get more complicated than I want and more expensive than the relay I'm trying to replace.

    The resistor loads really have very little to do with the relay I'm trying to control, but they receive AC power, three phase 480 or 240 volts. We use a step down transformer from here to get our 120 V control power, rather than a more expensive DC power supply.

    The relay is rated for 10 amps, but is not carrying anywhere near that. The coil itself pulls no more than a miliamp, I'm sure.

    Our load steps see currents as high as 120 amps (it's a 90kW step at 480V). Again, the steps have very little to do with the relay I'm trying to delay. The relay is what will prevent steps from receiving power if airflow is lost.

    This is what I do use, but my boss believes the part is too expensive and that a simple circuit could be built to accomplish the same thing at a much reduced cost.
     
  9. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Correct me if I'm wrong, guys, but I believe an AC relay coil will work just fine on DC.
     
  10. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Yeah, usually at about 1/2 the AC coil voltage.
     
  11. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Uhh, a 'Reed Relay' coil draws more than that. I doubt that this is correct.
     
  12. #12

    Expert

    Nov 30, 2010
    16,354
    6,851
    Back from work now...sorry about the delay (is that a pun?)
    Referring to post #4

    The "transistor" you are referring to is called a triac. You feed it some DC current and it passes AC. I think your scheme in post #4 should work. Thankfully, capacitors have gotten bigger and cheaper over the years. At worst, you would have to add a comparator of some sort to force the capacitor to charge to several volts before the switching action works, thus getting a reasonable delay time per microfarad. This is because triacs switch on at less than 1.5 volts and their current needs are a bit unpredictable. They just come with worst case guarantees and usually work with a lot less than the maximum.

    I also thought about how to do this mechanically. If you have several psi to work with, you could put a physical restriction (needle valve) in the air line to the pressure switch and add an air chamber that delays the rise in pressure. It might be as simple as a 3 foot diameter loop of plastic tubing. The important part is that the air volume in the new chamber makes the chamber in the pressure switch look small.

    I derive this idea from working with water pumps. Without a springy pressure chamber, they achieve pressure, switch off, stop the motor, and lose enough pressure to switch back on, sometimes oscillating indefinitely at about 1/2 hz.

    The electronic method can be made deadly predictable, maybe not so much the hydraulic method. Can't remember the word for "air hydraulics" :(
    Which do you want to try first?
     
    AceofAllan likes this.
  13. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    How in the world could (design time) man hours be less expensive than a common time delay relay? By the way, I understand (by your answer) that the relay contacts don't carry the load current.
     
  14. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Perhaps pneumatics?
     
    #12 likes this.
  15. AceofAllan

    Thread Starter New Member

    Jun 16, 2011
    7
    0
    I think I'm going to try the triac method (thanks for that, too). I've attached a hand sketch for how I intend to implement this. CA is my coil for airflow reading, and I added the normally open contact to show how it acts as a permissive for the rest of my circuit.

    The way I set this up, the DC voltage out of the bridge should be smoothed out enough by the capacitor. All I have to do is size Ra and Rb such that the voltage delivered to the triac is enough to activate it and not too much to damage it. Then I just have to scale the resistors and size C to get the right timing I want.

    Does this look at least mostly correct?

    I stopped asking these kind of questions a year ago. Somehow he figures my salary as some kind of overhead and so long as I find time to figure this out and still get the other jobs done, the "design time" costs nothing.
     
  16. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    At least two things wrong with your circuit. First off, Triacs are bidirectional devices but in order for it to conduct in both directions the gate must follow the polarity across A1 and A2.

    I didn't closely Analise your circuit so there may be other issues but here's an AAC issue... Our TOS forbids un-isolated mains projects. If you proceed with this here you will have to use an isolation transformer. A step down transformer, to work with a low voltage AC or DC relay, would also be acceptable.
     
    AceofAllan likes this.
  17. #12

    Expert

    Nov 30, 2010
    16,354
    6,851
    The triac will conduct both phases when provided with DC current through the gate to kathode circuit, but you have to arrange your circuit so the triac is the last thing in a series circuit and its k terminal is connected to the same "common" point as the rectifier. (swap the positions of "CA" and the triac.

    I am now fudging because I can't remember if the K terminal is what I'm trying to describe. The connection of the triac that is closest (on the drawing) to the gate connector is what I'm calling "K".
     
    AceofAllan likes this.
  18. AceofAllan

    Thread Starter New Member

    Jun 16, 2011
    7
    0
    Ok, there's no issue here because I do have a step down transformer from the 480 V that is applied to our resistors, that's what supplies the 120 V. There's quite a few more components, but I didn't include the entire control scheme of my product because I didn't think it was necessary.

    Ok, thanks very much. I think I'll try to look up some other circuits that make use of triacs to make sure I've got things in the proper positions.
     
  19. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    I don't think so. It does not have to be a steady state but it must reverse polarity when A1 and A2 do. There is no cathode/Kathode on a Triac.
     
  20. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    This circuit doesn't use a Triac. It's intended to be used with your existing 120VAC relay marked K2 on the print. I've given you almost a 3 second delay but you can shorten it by reducing the values of R1 or C2. Triacs will always have a zero crossing issue because they all have a minimum holding current rating. This circuit doesn't have that issue.
     
Loading...