OPA211 low noise op-amp simulation in Tina

Rahul,

THD is a function of the magnitude of the output voltage and the OPA211 PDS specification of 0.000015% applies only to Vout=3Vrms (see below).

Thus, in the case of applying 50mV signal, you may expect at best THD+Noise not better than 0.0001% (see the graph below)

However, in your particular case, you use large gain resistors and therefore the thermal noise coming from 11k resistors will completely dominate the THD+Noise performance.  The total thermal noise due to these resistors alone may be calculated to be: Vn=sq-rt[4*K*T*R*sq-rt(2)*sq-rt(BW*1.57)]=sq-rt[4*1.38066e-23*300*11000*1.41*sq-rt(22.5e6*1.57)] = ~9uVrms and thus the noise-to-signal ratio will be: 20*log(9e-6/50e-3) = -55dB and that is what dominates your THD+Noise reading.  Hence: use much larger input signal or much smaller resistors.

Thanks for the reply.  I have a few more questions regarding your explanation.  The BW that you used is 22.5MHz even though I am looking at random spikes in Fourier spectrum within a band of 10Hz to 2kHz, much smaller than 22.5MHz. 

In any case, to avoid this issue, I simply changed the circuit to a voltage follower (unity gain), increased the amplitude of input sine wave to 1V, and have an output impedance as 600ohm||1pF.  I still get the same issue, and this time I am attaching an image to explain the problem.  Based on Figure 4 that you attached, for a 1V amplitude (0.7V RMS), Gain=1, RL=600ohm, I expect a distortion no more than -124dB.  I am using an input of 31.25Hz instead of 1kHz on which the curve was based, but don't expect this to make a big difference.  The supply voltage is also +/-18V.

I have a feeling that these are numerical artifacts, but I am not sure.

Rahul,

I have used the 22.5MHz bandwidth for noise calculation because this is the effective bandwidth of OPA211 in gain of 2 (half of the unity-gain bandwidth).  You are correct suspecting that the noise spikes you see while using the TINA signal analyzer are artifacts  of the tool itself.  I ran the same analysis with OPA211 completely removed and I obtained exactly the same results (see below).

I believe the signal analyzer cannot be reliably used at the noise level below -20dB.  Instead you may run Noise Analysis under Analysis tab (see below).

Noise spectral density:

Signal to Noise ratio: