Dear all,
for my master thesis project I am designing a photodiode readout circuit. I use the TLC2274 opamp as a transimpedance amplifier with a gain of 680kOhm and the photodiode is in zero bias. The power is supplied by USB, so 0 and 5 volts. This is because I use an Arduino for readout and I want to keep my device portable. For simulations, I use PSpice and the TLC2274 model from TI.
In simulations the performance near the negative rail of the opamp is bad. If the amount of light is very small, the current from the photodiode will be very small. This also means the output voltage of the TLC2274 will be near the negative rail, i.e. 0 volts. In simulations, below roughly 130 mV the output is distorted and hangs at 130 mV.
Because the datasheet of the TLC2274 says it is designed to work from rail to rail, I decided to test it in real life. And guess what? It works fine near the rail! I've successfully measured output voltages down to 5 mV.
My question is: since the TLC2274's PSpice model is not rail-to-rail, can I trust the TLC2274 in my scenario, i.e. on near negative rail usage? Furthermore, can I use the PSpice model for bandwith simulations, or will it be inaccurate?
Thanks in advance,
Stephen