Re: Need help with op-amp/comparator circuit

From: Anthony Fremont (spam_at_anywhere.com)
Date: 01/15/05 

Date: Sat, 15 Jan 2005 21:34:06 GMT

"John Popelish" <jpopelish@rica.net> wrote in message
news:41E97660.36CFEECC@rica.net...
> Anthony Fremont wrote:

> Almost any rail to rail opamp will work much better than the 741.

Alright, I get the hint. ;-)

> The LM358 in not a rail to rail design, but can work with both inputs
> and outputs from the negative rail to 1.5 volts below the positive
> rail, and they are dirt common. This is the dual version of the LM324
> quad. It also has the same input circuit used in your comparator. So
> if you use this, you should change your reference divider from 2.5
> volts to (5V - 1.5V)/2= 1.75 volts, to center the reference voltage in
> the middle of both the input and output range.
> http://cache.national.com/ds/LM/LM158.pdf
>
> > In this particular case I don't want too
> > much gain, just enough hear the ticking, and not all the ambient
noise.
>
> You can easily add a trim pot between your input and feed back
> resistor on the opamp, with the wiper going to the - input to vary the
> gain. That way you get the biggest possible range of the ratio of
> feedback to input resistance for a given pot resistance.

Yes, I can see room for a couple of "tweaking knobs". ;-) Adjustable
gain on the amp stage and a tweak for the voltage divider on the
comparator would be perfect.

> > My current "microphone" is a discarded PC speaker (hey it's all I
had
> > "in stock" ;-).
>
> Nothing wrong with that.

It actually works pretty decently so far, but I could use a bit more
sensitivity now.

> > John Jardine's little RC feedback circuit gives me the exact effect
that
> > I was looking for, but the pulse is still a bit longer than I want.
I'm
> > curious how the 47K resistor he suggested fits into the time
constant
> > calculations. His cap suggestion was a bit large so I lowered it to
> > .1uF and that gives me ok results. NEWSFLASH: I'm now experiencing
an
> > oscillation problem with the 393. I'm fairly sure it was working
fine
> > at first (at least the scope looked right), but now it's self
triggering
> > and I can't seem to make it stop. :-(
>
> You need to move the reference voltage on the other input of the
> comparator ot a slightly different voltage compared to the center line
> of the audio signal, so the comparator can lock up.

You're right, problem appears to be solved (or at least identified).
;-) Aparently, the 47K resistor upsets the voltage coming from the 741
a little bit. On the 741 side, the voltage is still the 2.52 I wanted,
but the 393 side of the resistor seems to be a bit lower. I guess the
input impedance to the comparator isn't quite infinite and led to a
voltage drop accross the 47K resistor. ;-)

I put a 340K (it's what I found first ;-) accross the 10K-to-ground on
the 393's input voltage divider. That stopped the oscillation and it
still triggers, just with a bit less sensitivity. When I touch my
voltmeter probe (or finger) the junction of the divider it starts
self-triggering again. I'm guessing thats off the stray 60Hz AC voltage
present in the air, more than the added capacitance of the probe or
finger. What do you think?

> > The ultimate plan is to feed the cleaned up pulses into a PIC and
time
> > the distance between the beginning edges. The ratio of these two
times
> > will tell me how much out of beat the clock is. I will also be able
to
> > calculate beats per minute etc...

Thanks for all your help, this has been quite educational. Next up,
writing some PIC code.

Relevant Pages

  • Re: LDR detector
    ... Combine the LDR in series with a fixed-value resistor to make a simple ... voltage divider which can act like a voltage source that varies with ... LDR V_Light. ... The point is to have the comparator compare two voltages. ...
    (sci.electronics.design)
  • Re: LDR detector
    ... you mean connect a capasitor to op-amp input and ldr junction? ... Now use a potentiometer to create a programmable voltage reference. ... The point is to have the comparator compare two voltages. ... The comparator needs a pull-up resistor. ...
    (sci.electronics.design)
  • Re: A Quick Question
    ... determine what value actually satisfies the sensor circuit, then pull the proper resistor out of my all-too-large supply of spare parts. ... It probably looks like a voltage source more than it looks ... A voltage divider will have a characteristic impedance, i.e. it can be modeled as a Thevenin source with a resistor: ...
    (rec.motorcycles)
  • Re: op amp output
    ... John Smith wrote: ... the voltage on pin 12 should be less than ... >0.3V with the lowish current draw of the 5k6 pull-up resistor. ... >comparator is off, ...
    (sci.electronics.basics)
  • Re: op amp output
    ... John Smith wrote: ... the voltage on pin 12 should be less than ... >0.3V with the lowish current draw of the 5k6 pull-up resistor. ... >comparator is off, ...
    (sci.electronics.basics)