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

How do I delay current in a circuit

  1. Oct 9, 2012 #1
    Hello all, hope everyone is having a good week so far.

    Anyway I am trying to solve a problem, hopefully you guys can help me with this.
    I made a quick drawing on Paint to try to explain myself, sorry for the poor art.

    So basically in the circuit shown in my drawing, if you hit the switch then lights 1, 2 and 3 will receive 12 Volts DC and light up.
    The effect that I want to achieve is to create a delay on lights 2 and 3. So when you close the switch, light 1 will light up 1st, then after a few milliseconds, light 2 will light up, and finally after a few more milliseconds, light 3 will light up. At this point all 3 lights are on. And once you open the switch all 3 lights will turn off at the same time.

    I want to know if there is a simple way of doing this. The circuit in my drawing is an existing circuit, I would like to modify it to do what I explained above. I really don’t want to involve an Arduino or something of that caliber. Is there a simpler alternative? Maybe some type of component that I can place right before lights 2 and 3 to create that delay I need?

    I was thinking capacitors maybe?? But That might not work since I need all 3 lights to turn off at the same time once the switch it re opened, and if I use capacitors this probably wont happen right?

    Any help would be greatly appreciated!
     

    Attached Files:

  2. jcsd
  3. Oct 9, 2012 #2
    You'd have to calculate your RC time constants yourself. The switch is a 3 pole 2 throw switch readily available for cheap off RS. You have to wire up the "main pole" up NO and the other 2 poles need to be NC, this will discharge the caps ready for the next time you operate the "main pole".

    circuit222222.jpg
     
  4. Oct 9, 2012 #3
    jonlg_uk, Thanks for the quick response and diagram!

    Sorry to be a drag but I did not fully understand everything you said, I have to point out I have not studied EE. I have read a few beginner/intermediate books though.

    RC time constant refers to what capacitor I would use correct? If so, yea I figured if I went this route I would have to figure out what size caps to use.

    Now I did some research on the switch you talked about and it sounds like that would work. But I failed to mentioned that I cannot modify the switch(the one I drew on the diagram). The switch mechanism cannot be changed, what I can do is modify the circuit after the switch, where the lights are. So thats why I'm wondering how I can do this, is there a special relay I can use with the caps or something like that?


    THANK YOU
     
  5. Oct 9, 2012 #4
    Yes you can get electronic Delay On Timer relays these are readily available from rswww. If you go down that route then there is no need for capacitors or resistors like I drew.

    Here is something similar to what you need:

    http://www.ics-timers.com/op4142_2.html
     
  6. Oct 9, 2012 #5
    If you are not familiar with EE, buying after market delay will be the best way to go if they are available. RC is not a good way as the R will drop voltage and rob some of the voltage from the bulb, so your bulb two and three is going to be dimmer. It gets a lot more complicated if you want them to shut off at the same time as you are going to have to discharge the cap. Opening the power switch do not automatically discharge the cap even if you use a diode for discharging, you literally have to switch to ground to discharge the cap. Then another complication arise..........the in rush current through the diode to discharge the cap is going to be high and not only you need a big diode, you need a robust switch to take the high current surge so you won't melt the contact...............

    Why do you want to delay such a small amount of time, I don't think you can see a few mS difference.
     
  7. Oct 9, 2012 #6
    Yeah I read somewhere that it takes a human about 1/10 of a second to react to a stimulus, not sure if this applies to being able to distinguish between stimulus.
     
  8. Oct 9, 2012 #7
    You are right, I'm not sure exactly how long the delay needs to be but it is about 200 to maybe 300 milliseconds.

    The more I think about this it gets more complicated though. I know its possible because I've seen it done. I'll explain what exactly I was trying to accomplish here.

    So basically this is for a car's turn signals. My car has 3 bulbs on each turn signal(left and right). When you activate one of the blinkers(left or right), the 3 corresponding bulbs start flashing. Now I've seen very simple looking kits sold that make the 3 bulbs light up in a sequence rather than all at the same time. The kit looks simple because all it is, is a harness identical to the original harness plus a little black box, of course I don't know what is inside that little black box but I thought it should not be something crazy out of this world.

    Attached is a picture of the harness. It is sort of pricey for me and I thought it was simple enough to build myself but I'm starting to think its not as simple as I thought :frown:

    I wonder if I just have to split the amount of time that my turn signal stays on into 3 and allocate that time into each one of the 3 bulbs. Or I wonder if maybe each cycle(each blink) is not long enough to do this effect so maybe the kit does something more complicated like use capacitors or something to elongate each cycle and give each bulb an equal amount of ON time.

    Like I said I think is a little more complicated than I thought. I will go outside right now and test my turn signals and see how long they stay on.

    Let me know if you guys think this is too much or if you have any suggestions for me to try :confused:

    Also here is the link to a video of the kit i mentioned so you guys can see the effect it creates




    THANK YOU ALL for you time
     
    Last edited by a moderator: Sep 25, 2014
  9. Oct 9, 2012 #8
    So you think I can use this for the application I explained?
     
  10. Oct 9, 2012 #9
    OK, I don't know how much you can follow. I don't know what is the wattage of the light. But this is a simple circuit using a CMOS NOR and an NPN transistor. This is assuming you use DC to power the lamp. I don't know the latest CMOS gate family and I am too lazy to look at data sheet. So I did not specify the part number for you.

    The delay is governed by the RC, so you can have two of this with one having a long RC time constant for bulb 3.

    I choose CMOS gate to get a higher switching threshold voltage for the RC part. TTL is not good.

    Take a look, this is not a finished circuit, just an idea. If you think you can build it and need more help, post back.
     

    Attached Files:

    Last edited: Oct 10, 2012
  11. Oct 9, 2012 #10

    vk6kro

    User Avatar
    Science Advisor

    There used to be a TV show a long time ago and it was called Knight Rider. It was pretty bad, but there was a car in it that had all sorts of lighting effects on it.

    If you do a search on Google:
    https://www.google.com.au/search?nu...2965.2j0j9j0j1.12.0...0.0...1ac.1.VKmtIp6wn6M

    you can find a lot of lighting circuits that give effects of multiple LEDs switching on in turn.
    The circuits can be used to switch high current lamps if the lamps are driven by power transistors.

    However, you probably shouldn't be playing with the electronics of your car unless you know what you are doing.
     
  12. Oct 9, 2012 #11
    Awesome! Yea I've used an NPN transistor before for simple circuit projects. The rest I will have to do some research on but at least I have a direction to take now. Thats mainly what I needed since I was sort of clue less.

    I will post back if I run into trouble

    Thanks for your help!
     
  13. Oct 9, 2012 #12
    If you are more serious about going for it, give more detail about what voltage DC you are using, what is the wattage of the bulb. So we can track and offer more solid suggestion. Right now, it's just talking in the thin air. If the bulb is high wattage and need more current, a darlington NPN or even an N MOSFET is possible. Just more details.
     
  14. Oct 10, 2012 #13
    Sorry, I wasn't thinking, It should be a NOR gate, not NAND. I edited the original post already. Sorry.
     
  15. Oct 10, 2012 #14
    Oh ok got it, thanks for the correction.

    And yes im serious about building this. I am at work right now but when I get home I will check the bulb wattage.

    As far as Voltage, I'm almost certain it uses 12V from the car battery but I will double check this as well

    P.S. Just some background on myself, I am 22 and I'm currently doing my GEs at a community college and am planning on working towards a BS in Electrical with Computer Engineering at CSULA. I have been in a couple robotics clubs and such things so I do have some experience with electronics just not as much as you guys. But I can get around by doing research when I get stuck, so thanks for your help
     
  16. Oct 10, 2012 #15
    Now the design is much clearer. If you are using 12V, look into the old CMOS 4000 series logic family that was popular in the 70s.

    http://www.analog.com/static/imported-files/rarely_asked_questions/4000_Series_Article.pdf

    They work with Vcc up to 15V. They are perfect for you as you don't need an extra supply to power the logic. Keep R3 like 20K or lower and R1 at least 200K. I would use a N MOSFET, then you eliminate R2 all together. The 4000 family has very weak drive, MOSFET will be the best choice.

    Be very careful when handling these logic. In my days, this family have no protection at the input and they are very susceptible to static damage. Always store them in static protection bags or foam. Ground yourself to the earth ground before you even think about handling them. My experience is they never burn right on the spot, they get "injured"!!! They usual die in a few days. I remember every time we probe the board, certain IC will die in a few days. It was to the point when the system failed, we just change one particular chip without even trouble shoot. It was that predictable. The one that was the worst was 4049 inverter. The NOR did not seem to be as bad. We were not careful about grounding ourselves, but if you take all the precaution, they are very reliable.
     
    Last edited: Oct 10, 2012
  17. Oct 10, 2012 #16
    Very useful information!

    Hmm ok now if I'm not mistaken,the "4001: Quad 2-input NOR gate" is the one I need for this application correct?
     
  18. Oct 10, 2012 #17
    Yes. Make sure you ground all the other unused inputs. Tell me the wattage of your bulbs, I might be able to choose a MOSFET for you. I have good experience ordering in DigiKey, they ship FAST!!!! average 3 days!!! They should carry everything you need....I am not advertizing for Digitkey, I have not try other places, I hit the jackpot with the first one, never have any reason to try others.
     
    Last edited: Oct 10, 2012
  19. Oct 10, 2012 #18
    Awesome! I was just trying to figure out where to buy these things, I was looking around on Google. I’ve only bought stuff from SparkFun, RadioShack and a vendor called NerdKits. Never DigiKey But it looks like they have a huge inventory

    Anyway I’m still at work but I just looked up replacement bulbs for my car on the Autozone website and this is what the specs said.

    Bulb Voltage: 12.8
    Bulb Wattage: 28.54

    I am guessing they are the same as the ones I have in my car right now since they are supposed to be replacements but I will double check this as soon as I get home.
     
  20. Oct 10, 2012 #19
    So I will need to repeat the circuit you drew for lights 2 and 3 on both turn signals right?
     
  21. Oct 10, 2012 #20
    Then it should be very easy, 28W means it's draw less than 3A, plenty of MOSFET can do it for you.
    Yes, you need two of the circuits for two different bulbs. You change the R1 and/or C to set different delay time.

    I am waiting for an important call right now, I'll try to look up a MOSFET for you by tomorrow. I take you want the components with leads rather than surface mount?
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: How do I delay current in a circuit
Loading...