OPAMP response to drops

Hi everybody,

I'm working on something to measure raindrops with a piezo buzzer and arduino (at this point I've posted a ton of times about my project, but it's due in a week so this is the final stretch). Anyhow, I finally got an opamp to work - I'm using the LTC1050 from Linear Technologies.

This is my circuit: http://imgur.com/9EFf9

Anyhow, I'm seeing this weird phenomenon where all of the raindrop signals I record show a slowly increasing voltage that levels off, instead of a sharp peak and rapidly diminishing oscillations. The signals are shown herehttp://imgur.com/sY9ck- the signal on the top is for a sensor without amplification (it has a ton of noisy oscillations but if you look at the first peak it's a nice response), and the sensor on the bottom is amplified. The different color lines represent different drop sizes that I released onto the sensor (red is the biggest, blue the smallest). X axis is time in microseconds, Y axis is... maybe voltage. I don't quite remember.

Does anybody have an idea of why this would be? I know I have a HUGE gain set, but without that I'm not sure that I could see any drops at all. Is it acting like some kind of voltage integrator? The data still looks nice if I just take the max of each drop response over the 5000 microsecond time period, but I want to understand this curve and I have a really poor circuitry background..

thanks!!

My guess is that the capacitor is taking time to charge.

Your capacitor is causing the circuit to be a low pass filter with a cutoff of around 5Hz - its effectively integrating its input.

First: LABEL YOUR COMPONENTS! Else it's hard to talk about specific component instances.

I'd try with a lower capacitor value for the lower-left capacitor, and perhaps a higher value for the upper-right capacitor.

On the other hand, if you're going to read using the integrated ADC, you will need some amount of low-pass filtering/integrating, because it's not a perfect instantaneous sampler. The analog world is messy :-)

Thanks! I'll try changing that capacitor, and sorry for not labeling.

Question - do the capacitors affect the voltage amplification? This is what I really don't understand - I know that the resistors (unlabelled, but commonly labelled R2 and R1) produce a voltage gain of 1+R2/R1, but will changing the capacitors affect that?

do the capacitors affect the voltage amplification?

Yes. Capacitors have different resistance at different frequencies (just like inductors, but the inverse relation.) The feedback capacitor (top right) bypasses the resistor for high-frequency components, leading to higher gain at higher frequencies. In fact, you may want to remove that capacitor entirely as another experiment, or tie it between "-" and "ground" (leading to a low-pass instead of high-pass filter.)

I'd try a resistor from the opamp + terminal to gnd, maybe somewheres between 22K and 100K. You have 2 caps, in essence, with no DC connection on that point.

jwatte:
leading to higher gain at higher frequencies.

Everything I said, was right, except this. Because the cap works as a wire for high frequencies (up to a limit,) gain at HF will be reduced by that cap, and it will act as a low pass, in addition to the bottom-left.

Reduce the that feedback capacitor by a factor of 50 or more and you'll be getting closer to something useful I think.