Original Question: Noise analysis by Tarzan
Hi there,
I just started using ADISIMPE recently and I have a question about noise analysis.
I tried to check output/input noise of the amp circuit.
But the result of the graph is pretty wired.
The output noise is much smaller than input noise…..
I changed the parts from AD8606 to AD8601 and I got a result as I expected.
Please help me find out.
Why AD8606 case does not work properly.
I attached the file and graph.
Best regards,
Tarzan
Verified Answer: RE: Noise analysis by JGunao
Hi Tarzan,
Kindly try using a dual supply configuration in your simulation, a +/-2.5V voltage supply will do. Below is the screenshot of my simulation.
If you have any more concerns please me know.
Regards,
Jay
Question: RE: Noise analysis by Tarzan
Hi,
If there is a person whno knows ADISIM well, please give me a hand.
Please let me know if my question is not clear.
Thanks in advance.
Best regards,
Tarzan
Question: RE: Noise analysis by JGunao
Hi Tarzan,
I'm glad that I helped you in your simulation. I recommended you to use dual supply configuration because most of our spice models are tested and modeled to work at dual supplies. Since you used a single supply of 5V, it is just the same of using +/-2.5V supply. In addition to that, this will also be applicable to other simulation analysis and not just for noise analysis.
I hope this helps.
Regards,
Jay
Question: RE: Noise analysis by Tarzan
Hi Jay,
I'm sorry that I could not reply quickly.
And I really appriciate your support.
By the way, I sitll wonder why this kind of thing happens.
Is there anything we should to consider when we try the noize analysis of OMPAMP.
Best ,
Tarzan
Question: RE: Noise analysis by ScottH
Hi Tarzan,
I would like to point out one thing. For a +5V supply, the output range of the AD8606 is specified as typically 20mV to 4.98V. We model the output saturation in our spice models. In the configuration that you posted, the expected output voltage would be about 0V. The output can only get down to about 20mV, so it will be saturated. Then when the simetrix simulator applies a small-signal ac voltage at the input, the output doesn't move or moves very little because it is clamped to 20mV, so it follows that the output noise would be much smaller than the input noise, just as you observed.
Jay's solution solved this because the negative supply brought the expected 0V output within the output range. The other thing you can do, if you need to use the +5V supply, is add a bias source to raise the output voltage so the output is not saturated, as I did with V3 in the following picture:
Best regards,
Scott