Need a circuit that notifies me when a signal is stabile for a few seconds.

Discussion in 'General Electronics Chat' started by Deangreen, Sep 10, 2015.

  1. Deangreen

    Thread Starter New Member

    Sep 10, 2015
    4
    0
    Does anyone have any ideas on a simple circuit that activates an LED when my signal (0-6 volts) has stabilized. I would like to be able to set a window time such that if the signal doesn't change within 1-5 seconds, then the LED will light up. I would appreciate any ideas, especially simple and cheap ones.
    Thanks!
     
  2. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,046
    3,813
    Define non-stable:

    Define stable:
     
  3. Deangreen

    Thread Starter New Member

    Sep 10, 2015
    4
    0
    A signal that changes maybe less than about 0.05V/sec. But would like to be able to change the delay at which a comparator compares a signal with the delayed signal.
     
  4. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    Show an example signal with the instability you want to filter.
     
  5. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,538
    1,251
    So you want an indication when the rate of change is below some dV/dt, or when the long term DC value has stabilized (at any value) even through there are small amplitude changes still happening, or what? It sounds like the decision criteria for the LED has two parameters, and any simple comparator circuit handles just one at a time.

    ak
     
  6. Deangreen

    Thread Starter New Member

    Sep 10, 2015
    4
    0
    Seconds Volts
    0 3.00
    1 3.25
    2 3.50
    3 3.60
    4 3.63
    5 3.65
    6 3.68
    7 3.70
    8 3.71
    9 3.71
    10 3.72

    Note that the time it takes to stabilize varies. Students waste a lot of time watching a pH meter slowly rising, but there is rate where it starts to slow down and the small changes seen in the pH is not significant. I assume the rate can be adjusted with a pot and capacitor. For example 0.01 pH units per 5 seconds may be slow enough to get a decent pH. I read some diy using 741s but the circuits were for over and under voltages set with a reference signal.
     
  7. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    I think you're making invalid assumptions. A circuit to do what you want will not be simple because it involves sampling, timing, and comparing. If you're handy with microcontrollers, you can program the parameters you want.

    In the example data above, the "circuit" would have gave a "stable" indication after the 4th second, but the voltage continued to change after that. Until you have a better idea of what you need (not want), no one will be able to help you.
     
  8. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,538
    1,251
    The word "stabilized" is ambiguous from a signal processing point of view.

    1. Do you want an indication when the rate of change is below some dV/dt value, no matter what the DC value is?
    2. Do you want an indication when the rate of change is below some dV/dt value, even if the DC value still is increasing or decreasing slowly?
    3. Do you want an indication when the peak-to-peak signal variation has fallen below a specified error band, no matter how quickly it is varying within the band?
    4. Do you want an indication when the signal has remained within a fixed detection window greater than a minimum amount of time?

    etc.

    ak
     
    atferrari likes this.
  9. Robin Mitchell

    Well-Known Member

    Oct 25, 2009
    732
    200
    Differentiator connected to a monostable which then triggers a Flip Flop

    No change = differentiator output = 0
    This triggers a 555 monostable (negative edge going).
    Maybe when the differentiators output goes high it can reset the monostable
    Monostable has finished triggers set pin on Flip Flop
     
    atferrari likes this.
  10. crutschow

    Expert

    Mar 14, 2008
    13,013
    3,233
    You can use a differentiator circuit to indicate a voltage rate-of-change. (Opps, Mitch beat me to it).
    For example a differentiator with a time-constant of 2 seconds will give a 100mV DC output for an input change of 0.05V/s.
    Thus when the differentiator output drops below 100mV a comparator could light an LED.
     
  11. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,538
    1,251
    Since there is nothing in the OP about a monotonic input, the differentiator would need to be followed with either an active full wave rectifier and a single comparator, or a window comparator.

    ak
     
    atferrari likes this.
  12. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,046
    3,813

    Simply tell the students to record the value when the hundredths digit changes less than 1 unit per second. Any filtering (capacitor/resistor) combination will essentially slow the rate of change and students may actually wait longer.

    You can also cover the 100ths digit and have them record 2 significant figures. Close enough for a high school or freshman college lab.
     
  13. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    The "problem" is that students waste too much time observing their experiment. If that time is wasted, what would you call the copious amounts of time students spend pursuing less fruitful endeavors?
     
  14. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,046
    3,813
    The problem is that there is generally not enough laboratory equipment in the biology labs and Chemistry Labs in high schools and colleges (engineering programs consume too much of the capital equipment budget). Anyhow, there is likely a time issue and there is no chance for everyone to complete the experiments if too many kids dick around with the meter waiting for their last digit to stabilize. Those are likely the future engineering students!
     
  15. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    You're able to read between the lines much better than I... The OP was talking about students wasting time. I thought s/he was trying to prevent them from wasting time on meaningless observation of experiments, not tying up lab and equipment time.

    If indeed a school doesn't have enough laboratory space or equipment, that's their problem. They either need to design their experiments accordingly or increase their teaching capacity. With what they charge for tuition, you'd think they could offer students their money's worth.
     
  16. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,046
    3,813
    Ok, then stop responding to this thread. Let it be THEIR problem (whoever THEIR might be). For now, this instructor is trying to solve a problem that he thinks needs to be fixed. And, he is looking for a low cost/no cost solution to one of his many problems. It really does not need to cause you any worry, he wasn't implying that it was your problem to solve or that he specifically needed your time. I've taught high school chemistry and biology as well as college and graduate level chemistry courses. The budgets are thin and no matter what you say, you will always vote no on the tax increase to pay a bigger part of tuition, you will always want your alma mater to have a good football team (and a multi-million dollar coach). So, thank you. Your help has been completely helpful to this point but I think this person can solve the rest without your valuable input - "that's their problem". That kind of feedback can be said of every person asking for help on this site. I truly do not understand why you would be here offering help with that attitude.
     
  17. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    A data slicer (look it up) followed by a fast monostable.
     
  18. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    What gives you the right to say such things about someone you know absolutely nothing about? For your information, I don't care for any college sports.
    I was actually trying to help the OP find a solution. You sidetracked the thread with your dubious ability to interpret what the OP wanted without needing to ask clarifying questions and telling the OP that s/he should change the way the experiment is done.

    For the record, I would have suggested that the OP use a microcontroller to sample the voltage and use whatever criteria s/he wanted to signal the end state. If you had taken the time to actually read all of the posts, s/he was first talking about a 6V settling voltage; but gave less than 4V in the example data. Furthermore, most of the clarifying questions have not been answered.

    And where did you get the notion that the OP is a "teacher"? Could just as easily be a student who doesn't want to waste their time doing a lab exercise. I'm certain that you aren't a gopher.:confused:
     
    Last edited: Sep 10, 2015
  19. Deangreen

    Thread Starter New Member

    Sep 10, 2015
    4
    0
    Thanks for the suggestions and comments??. The data slicer seems closer to what I need. The "stable signal indicator" would be very useful for other projects where a signal slowly rises...... with the signal/time slope shallowing out, but never reaching 0. So I was thinking about something that compares the signal every X seconds. I was thinking of something that remembers a signal (capacitors, diodes?) and compares it to the signal x seconds later. I'm very new at electronics (that is why I'm on this site for helpful suggestions) and do p/t science demos for HS. Since I'm a bit older and my brains signal rate is much slower, please be patient.
     
  20. Robin Mitchell

    Well-Known Member

    Oct 25, 2009
    732
    200
    Sample and hold circuit can be used to "remember the voltage". Have a second sample and hold circuit to take a reading then feed these two signals into a comparator. You can have an amplifier connected to the sample and hold of the original measurement to account for drop in voltage across the capacitor due to natural discharge. So if you know that the capacitors voltage will drop by 0.1V then have an adding amplifier that adds 0.1V to that signal before comparing the voltage with the current sample.

    Then maybe you could send this signal through a differentiator so that if the comparators output is changing then you know that the input is still changing. One the signal becomes stable a monostable is triggered.
     
    Deangreen, absf and atferrari like this.
Loading...