Decoupling Capacitors/Resistors

mearvk

Hello.

Ignoring my possible lack of correct terminology my question is about reading circuit diagrams in terms of understanding what they do, instead of just understanding how they are to be built.

Have a look here:

http://embedded-lab.com/blog/?p=1671

If you look in the middle of this article you see the following image:

http://embedded-lab.com/blog/wp-cont...lAmplifier.jpg

Now what I see is that what is happening is the diodes on the left drive the signal which is amplified across the two op amps and read by the microcontroller.

My actual question is:

1. Between the diode circuit on the left and the input to the op amp you have 1 uF cap and a 68k resistor. What purpose does the resistor serve? What purpose does the capacitor serve?

I've seen decoupling capacitors which, I believe, are used to quiet a noisy current before it is input to something else. The inline capacitors, I was told, sort of work to clarify a noisy current. I'm not sure what a decoupling resistor is for however I imagine that some percentage of the current will cross the 68 k resistor and so maybe it just acts as a current limiter?

Thanks.

sophiecentaur

The series C is an AC 'coupling' capacitor and not a 'decoupling' capacitor.

The 68k resistor is there to define a time constant for the high pass (RC) filter (DC blocking) which the 1uF capacitor is there for. (T=0.068s) The IC amplifier will have a very high input impedance (many MΩ) and, without the 68k resistor, the time constant would be many seconds.

jim hardy

Oops I see sophie answered you. So this post is unnecessary but i'll leave it.

If I read it right -

A line or two above the diagram the article says: The 68 k resistor in combination with the capacitor makes a ~2.3 hz high pass filter before input to amplifier.

The feedback network makes the amplifier into a low pass, again ~ 2.3 hz,
so you have a 2.3 hz bandpass filter that's not overly sharp. That's the frequency range of interest for heartbeats? 2.3 X 60 = ~140 bpm....

dlgoff

Since the OP's title uses the term "decoupling" capacitor, I thought seeing them decoupling noise from a pcb IC supply trace in a picture might be useful. Compliments of http://bilgisayarlarim.com/VTech/Laser_3000/

mearvk

Thanks Sophie and Jim.

It seems that he implemented a high-pass filter but, based on the article, he meant to implement a low-pass filter:

"The signal conditioning circuit consists of two identical active low pass filters with a cut-off frequency of about 2.5 Hz. This means the maximum measurable heart rate is about 150 bpm."

http://en.wikipedia.org/wiki/High-pass_filter

mearvk

Thanks but I'm not at the point where I can look at that picture and see anything other than noise in terms of the inferable system logic.

dlgoff

Actually it's fairly simple. The capacitors are across the integrated circuit's power supply "+" and "-" traces. See what happens to the noise? The closer to the chip the better they decouple (they're also know as bypass capacitors).




GIF file complements of http://www.williamson-labs.com/480_anim.htm

mearvk

That's a nice animation. The little green chip at the very end is a miniature capacitor?

dlgoff

Yep. :)

also notice the extra circuit copper forming a "ground plane" which can also decouple.

jim hardy

He did not mention in his article that the 1μf-68K comprise a high pass filter. He simply described them as blocking DC.
See figure 1 in the wiki link you referenced. It explains that nuance.
http://en.wikipedia.org/wiki/High-pass_filter

So he has a passive high pass filter followed by an active low pass filter, the term would be "Cascaded".
Observe he arranged his op-amp low-pass filter as a follower.
Follower has high input impedance, unlike inverting arrangement.
He did that to present high impedance to the 68 k resistor.
He doubtless picked 68K for a reason. 1K or 1 meg with 1 μf would also have blocked DC, but 68K gives him a corner frequency in his range of interest.
If he allowed the 68K to be paralleled with a low impedance inverting opamp instead of the follower, he'd have changed his RC time constant and shifted his hi-pass frequency.

Glad to see you're studying these things.
Schematics are a language and study is how one becomes fluent in the nuances.
"Science is but language well arranged" - lavoisier