79Views13Replies

Author Options:

Help with OP amps Answered

Picture of

I'm making an instructable about this circuit, it's a heartbeat detector, it's supposed to beep when a heartbeat is detected. Something weird is going on with the comparator, first of all, the input signal to the comparator seems to get "stuck" at the level of the comparator, like it is an invisible barrier that doesn't allows the signal to cross it easily, this seems to happen with and without the buzzer.
Secondly, the comparator "rings", it doesn't displays a clean square signal when the threshold is reached, instead, at the end of the square pulse "ringing" appears. Removing the buzzer or speaker solves the ringing problem. 

I tried replacing the buzzer by a 8 ohm speaker, but the problem persists. Then I tried changing the 1k resistor of the transistor by a 10k, but the problem is still the same.
I don't know what could be causing this, bad design? A faulty chip?

In the second picture it can be seen how the signal "bumps" an invisible barrier when reaches the comparator level, in the third the "ringing" can be seen, the fourth and last picture is how the (previous) comparator signal looks when the load is removed, the problem disappears.

Thanks in advance,

Tags:opamp

13 Replies

user
icengBest Answer (author)2015-08-06

You have an impressive electronic skill set.

Definitely the circuit needs a o.1uF through o.22uF ceramic capacitor very close to the Quad op-amp power pins 11 and 4... It would not hurt to add a 22 uF electrolytic capacitor across the input power too.

What happens is the inductive speaker generates slight fluctuation on the DC power causing two leds to change a miniscule light level that gets re-amplified and you have an oscillator.

I presume the IR sensor like LED is in the correct ckt board position ?

Select as Best AnswerUndo Best Answer

user
Victor805 (author)iceng2015-08-06

I tried to add several capacitors but it didn't work well, I added a debounce circuit (100nF and 10k resistor from the output to ground) and it might have done a bit with the ringing, but not much (maybe with a beefed up version). Finally, I changed the resistor by a 1k, now it works almost perfectly, but the sound is a bit faint, I'll try looking for more combinations.

Thanks for the idea.

Select as Best AnswerUndo Best Answer

user
iceng (author)Victor8052015-08-06

BTW that 10 ohm resistor and 4.7 uF capacitors is called a de-coupler circuit.

It becomes a near separate supply for noise making electric component.

Select as Best AnswerUndo Best Answer

user
Victor805 (author)iceng2015-08-06

That makes a lot of sense. After some more testing I've concluded all the problems occur only when a
buzzer or speaker are connected, there might be a bit of coupling. I'll add a capacitor as you said and maybe I'll try to add some sort of debouncing circuit to the output of the comparator to break that loop, if that still doesn't works I'll try decreasing the current though the buzzer changing the 100R by 1k, hopefully that will reduce interferences.

My electronic design skills are still lame, I cheated and copied the first part of the circuit from make:


I just designed the comparator and included the square wave generator with a buzzer.

Select as Best AnswerUndo Best Answer

user
steveastrouk (author)2015-08-07

put 100uf across the leds

Select as Best AnswerUndo Best Answer

user
steveastrouk (author)iceng2015-08-08

no, his circuit shows the illumination coming from two leds fed by the resistor. If he put a cap across them, the rc network would act as a smoothing filter. Or he puts a 317 on the leds to stop feedback.

Select as Best AnswerUndo Best Answer

user
steveastrouk (author)iceng2015-08-08

you can't fool me. That's the same schematic !! And there are no leds in it.....

Select as Best AnswerUndo Best Answer

user
steveastrouk (author)iceng2015-08-10

Stupid system. Then like your first version. 100uF x 0.0001 =0.01 secs, which should be nicely filtered at the rate that comparator was ringing

Select as Best AnswerUndo Best Answer

user
iceng (author)steveastrouk2015-08-10

Never understand why my pics get snipped.

Select as Best AnswerUndo Best Answer

user
Victor805 (author)2015-08-06

Here's the circuit board, clearances are not really respected in some places, but I haven't found any errors here. (The unconnected traces are jumpers going on top of the board)

Select as Best AnswerUndo Best Answer