Hi,
I've designed pulse oximeter circuit. When one LED lights continuously, I got perfect PPG (Photoplethysmogram) signal.


Then I measure PPG while blinking LEDS, like that:


I got stuck. There is no signal or just noise. When I removed low pass filters, I easily got my opamp saturated. I lowered LED current (I have current source for LEDs) and opamp gain, but there were little information in output signal.

I opearte only on AC signal. My circuit is similar to found on web, for. ex digikey, NXP:

Yes, I have low speed ADC (MCP3426). I set 240SPS on 12-bit to measure faster.
I have changed a lot of things, even software delay, but I have difficulties to get two signals from one photodiode. How to people make those pulse oximeters?

 

C4 couples the AC signal to OA2.
I thing it would be better to DC couple the signal and use an adjustable DC offset on OA2.

 

Probably the light signal level between the RED and IR are so vastly different that the AC coupled amp just slams from rail to rail when you switch the sources.

Two ways to fix it:

1) Use a switchable offset voltage to compensate for the difference.

2) Tweak the LED currents to null out the difference.

It might be possible to fix one LED current and servo the other to null out the difference, as long as the time constant was long enough so it doesn't null the signal too.

 

Actually, I have adjustable DC offset on OA2 and R4 using DAC.
Now, how I can help? Set 0V during switching?