Author Options:

A Bit of Circuitry Advice! Answered

Hey all! Me again :D. 

So I was messing around with my electronics and decided to make a pulse sensor, but I ran into a couple of problems along the way!

My project is based off of this, http://makezine.com/projects/ir-pulse-sensor/, but I didn't have a LM324, so I was trying a workaround. 

So grabbing my infrared light detector, and new multimeter, hooked my IR sensor up between 5V and and A0 and a 10k resistor from A0 to GND, just like with a light sensor, but obviously, it wasn't sensitive enough, so I kicked it up notch. Since I didn't have the specific amplifier that MAKE used I decided to use what I had, which was a LM358. After visiting the datasheet and finding a schematic, I came up with this circuit (image below). Now it's not sensitive enough yet to measure pulses.

The circuit from MAKE makes use of the the 2 of 4 internal op amps in the LM324, and I thought I could make the same circuit on my LM358, since it had two op amps inside of it. So basically, replacing the LM324 int the circuit with a LM358. 

The image below is a circuit I made up using just one of the LM358's op amps. Tonight I will try to use both just like in the MAKE circuit. 

If you guys could just give me a few pointers and set me on the right track that would be great, or if I even need two amplifiers! Why can't I just make the first one more powerful? 

And the capacitor between the - and the output varies. I'm still trying out different caps. The 0.1uf is too much, so I'm experimenting. 



If you already have a working circuit and substitute different op amps you should at least stick to the layout.

For example the photo sensor is meant to pull the voltage to ground for a good signal, you keep a steady signal by reversing the sensor and resistor.

Also you have no transistor to amplify the signal from the op amp.

Stick to the schematics first, then we can see how it works and where it needs improvements ;)

Well I just wanted to say thanks for helping me. I'm so used to cheating in electronics by using my arduino and pre-made sensor circuits! Now I'm actually trying to get my hands dirty, so bear with me. . .

First off, I had no clue that the sensor position had any effect on the circuit! I didn't even think about it.

Secondly, I thought since I was already amplifying the voltage that I didn't need the transistor, since I was already getting a decent 2v output from my op amp. But wrong again I was, because the transistor is to amplify the current rather than the voltage? And the op amp simply amplifies voltage and not the current?

Because the output from my sensor/resistor voltage divider is giving me a nice 0.2 v when I have my finger up against it, and after the op amp with a gain of 10 I have 2v. So it's obviously amplifying voltage, but not really amplifying the current much?

And the transistor is used to amplify the sensor's current I guess?

Thanks so much!


The transistor seems to be an emitter follower, basically a buffer between the opamp and the arduino. The opamp by itself could supply enough current to drive an AVR I/O pin.

The opamp is set up to smooth and filter the sensor input. It's possible the transistor also adds something to that signal processing, although as designed it wouldn't do much to change the output signal.

The overall circuit is pretty specific. For instance, the first opamp looks like a non-inverting integrator, the second op amp a non-inverting differentiator. The two in series would filter the input in a designed fashion. You won't achieve the same results with a single op amp.

Hey thanks a lot! I've learned quite a bit :D. I built the circuit using an the LM358, but the capacitors seem to limit speed that it can measure too much. If I run around and come back to my sensor it can't read anything until my heart rate is normal again.

There appears to be a low-pass RC filter on the IR sensor, with a corner freq around 2 Hz. But a faster heartbeat should register, no problem. It might be related to the overall analog signal processing, if you're not using two op amps...(or using them differently).

I would stick pretty close to the Make schematic, if possible. It looks like a very "focused" op am application, as is. That's a really clean output "trace."

Otherwise, maybe start from scratch. I'm sure the Arduino could do some signal processing as well...

The transistor is in place to provide a decent 5V signal that is less dependent on the opamp-not really that necessary if you get a decent signal but for most circuits it is prefered to have 5V signals.

Since the input on your arduino does not need real current it is really all about the 5V signal.

You can also check the output nicely with a LED, add a resistor suitable for 5V and check if it lights up properly.

If it does the circuit should be fine for the Arduino.


The 68K resistor is now 100K.