Delay on, Delay off

Discussion in 'Electronics Resources' started by JohnSD, Aug 12, 2011.

  1. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    Guys,

    I am very new to building circuits and I am attempting to build a delay on/delay on circuit. I am operating a solenoid that will draw up to 1.4 amps at 12 volts. I want the curcuit to:

    -Delay power to the solenoid for an adjustable time from a 1 to 60 seconds with an accuracy of +-1 second once a switch is thrown.
    -Continue powering the solenoid for an adjustable amount of time 1 to 60 seconds once the switch is placed to the off position with the same accuracy.
    -Be inexpensive, sturdy and reliable.

    I have attached my first cut. I bread boarded this circuit and it did not do what I needed it to do. Any thoughts? I am not wedded the 555 for this project.
     
  2. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    First of all connect the RESET input of the 555 to the power supply's +. Otherwise it's disabled.

    Is the first delay supposed to be given by the combination of RC at the base of the trigger input transistor? What's the voltage at the upper pole of the 2k pot?

    Even if it worked with the ON delay transistor I doubt it will be as reliable as the 555 timer (and won't give you exactly the same delay). You might consider using only 555's.
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Well, the 555 timer is only accurate to within ±2.25% initially, and drifts more over temperature, so if you need 60 minutes within 1 minute, that's 1.66% accuracy. The 555 isn't going to meet your specifications. You'll need to either go to a microcontroller that will stay better than 1% accuracy, or use a crystal oscillator - the latter will be much more accurate.
     
  4. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    I got this from an (old) app note from Philips:
    "The 555 and its complement, the 556 Dual Timer, exhibit a typical
    initial timing accuracy of 1% with a 50ppm/C timing drift with
    temperature. To operate the timer as a one-shot, only two external"

    I didn't check datasheets of all manufacturers. Which manufacturer are you refering to?


    Btw, I think he meant seconds.;) Shouldn't matter for the accuracy percentage though...
     
  5. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    The accuracy needed sounds like an app for a microcontroller.

    There are tons out there, from PICAxe to Basic Stamp, with Arduino in the middle for costs, all are < $60 to get started. You would need an external driver transistor for the solenoid, but the uC would give the ability for an LCD display for both delays once triggered.

    Otherwise, 555 will be close, but when going beyond a few seconds, you end up using electrolytic capacitors, which vary widely in tolerance based on temperature and age, so the setting would need to be tweaked now and again, rather than "set and forget".

    --ETA: You could also use digital logic at a clock speed of 32.768 kHz crystal feeding a counter, this would give the accuracy with a lower cost and a few more parts, however.
     
  6. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Smallest part count and highest accuracy and the adjustable feature is really only achieved with a microcontroller.

    1.
    How do you want to adjust the delays? Potentiometer, up/down switches? If you want high accuracy u also need to know what exactly you adjusted. How do you want this feedback? Using a 7-segment display? However if you use DIP-switches to program the timer you wouldn't need this visual feedback...

    2.
    Do the two delays have to be adjusted independently?

    3.
    What's the minimum adjustable step? 0.1s, 1s...?
     
    Last edited: Aug 13, 2011
  7. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    National Semiconductor's datasheet for the LM555.
    Link to datasheet: http://www.national.com/ds/LM/LM555.pdf

    Not all 555 timers are equal. ;)

    Accuracy will also depend quite a bit on the stability of the capacitor used. They can vary quite a bit over temperature, let alone initial values.
     
  8. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    After my bread board attempt failed I did look at some examples of 555 in use and looked at the data sheet again and noticed it should be wired high if the reset feature is not being used. Thanks for confirming that that that was wired incorrectly.

    Yes, the first delay is supposed to be given by the combination of the RC at the base of the trigger input transistor.

    12 Volts to the upper pole of the 2k pot.

    I have no problem using two 555s or a 556. I am brand new to this if there is a better way to skin the cat I'm in.
     
  9. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    2.25% should be good enough. I was looking for +-1 second out of 60 which would be 2/60=3.33%.

    +-2.25 (4.5%) which would work out to +- 1.35 seconds will get me close enough.
     
  10. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    If it is simple and cost effective I am in.
     
  11. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    Adjusting the timing with a potentiometer will work fine. I will probably use rough calcs to get me in the ballpark for timing and then calibrate the timing using trial and error. I do not need visual feedback.

    I do need to be able to adjust the delay on and the delay off independently.

    2 second intervals is the minimum.
     
  12. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    This is a good point. I had not considered the difference in accuracy between manufacturers.

    I also noticed the NE555 is only rated from 0-70C while the SE555 is good from -55 to +125C.
     
  13. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    See the attached picture. There are missing the correct monostable values, the output solenoid driver transistor and maybe one can simplify the circuit.

    If nobody did it until tomorrow I can maybe continue.;)

    Tell us if that's what you wanted.

    The output will be ON-delayed with one 555 (or half of a 556) and OFF-delayed with another one.
     
  14. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    This looks like what I need. Thanks!:D I could just use pots with a wide range to test it and then get more specific once I find out the resistance range that will give me the values I need right?:confused:
     
  15. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    No. I didn't put the right values, because otherwise the simulation gets very slow. You can calculate it. I'll do it later this day.
     
  16. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Ok that should do it.

    I recommend using a MOSFET instead of the darlington. The TIP121 has a quite big collector-emitter saturation voltage. If you use a Mosfet change the base/(then gate) resistor to a smaller value, let's say 100R for example.

    The MOSFET has to be rated for the power supply voltage/solenoid rating.

    The timer can be adjusted between +- 1.8s and 62,3s. As stated before the poor temperature stability of electrolytic capacitors may be a problem for you.
     
    Last edited: Aug 14, 2011
  17. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    Interesting point concerning the saturation voltage. I actually bread boarded my initial circuit again, except this time I wired pin 4 high like it should have been in the first place. The circuit actually functioned with a delay on and a delay off, however, the LED I had on the output was dim. I checked the voltage and the output for the TIP121 was only a couple volts.

    I think I will separate out the RC timer section and work on that. The values I had didn't work well. The range for the pot was very narrow to adjust the delay on and the MPSA14 output voltage was also too low. Do you think the MPSA14 is a good choice for the RC timer portion?
     
  18. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    I don't think a simple RC-transistor solution will give you a reliable timer like the one you wanted. First you said you could use 2x555s, now you want to use your original circuit?
     
  19. JohnSD

    Thread Starter New Member

    Aug 12, 2011
    9
    0
    I didn't have the other circuits so I just started playing with the components I had. I have never played with an RC-transitor timer so it is interesting to see how it works. Even if I don't use the RC-transitor I would like to know its capabilites and limitations. Is the problem with the RC-transitor consistency?

    The parts count with 2 555s is high. I could reduce it significantly by using a 556 right?
     
  20. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    1x556 = 2x555. You save one IC, all other components are still needed.
    Problem with a simple RC/transistor combination is that it has to turn on very fast in order to generate a pulse to trigger the 555. The simple RC/transistor circuit you have doesn't turn on equally fast, because you are creating the delay adjusting the base resistor value to change the time needed to charge the capacitor... Best would be a schmitt trigger circuit, which involves more components.

    Or you can try to use a relatively big capacitor and put a zener diode in series with the base and try to work on the steeper charging slope of the Capacitor at the beginning, not at the end. (to turn the transistor ON quickly)

    The smallest part count and highest reliability/accuracy you'll have with a microcontroller. They cost almost nothing, but you need to know how to program.
     
    Last edited: Aug 15, 2011
Loading...