I am designing a light scattering sensor then needs ambient light rejection.
I have a 500Hz PWM modulated LED source that shines into a sample volume.
A collimated photodiode picks up the scattered modulated light, as well as higher frequency noise and near-DC ambient light (sunlight).
I started with the AD630 "lock-in amplifier" design from the datasheet (i.e. fig 14, without the output filter). The input is the transimpedance photodiode amp output (pin 1 and 16). I added the input PWM signal (0-5V, injected into either pin 9 (with 10 gnd) or vice versa). When I use the PWM signal on pin 10, and pin 9 is GND, then the output square wave and the input square wave appear to be in phase.
Anyways I was curious about the output. The output square wave swings both positive and negative, although the input from the photodiode only swings between GND and 5V (i'm using an AD8628 for the transimpendance photodiode amp).
Anyways, i'm essentially trying to get a DC voltage which quantifies the difference between the photodiode response between a LED ON and OFF state. I can imagine performing this digitally (i.e. with a fast ADC and some simple code), however i'm trying to use the AD630 to do this.
Have I the wrong product here? The output square wave's amplitude (both +ve and -ve) grows with increased ambient light, however i'm expecting that if the demodulator is doing its job, then the 1khz signal it (AC or otherwise) to be invariant of the DC ambient light conditions.
I am looking for guidance here - is there something i've missed? The design in the datasheet mentions a Low pass filter, however i don't trust the LPF will solve the problem of the AD630's demodulated output amplitude being sensitive to the ambient light.
Any advice would be greatly appreciated.