For a new project I want to use AD8676 as low noise amplifier. The problem is that trying to simulate the circuit with pspice i found an unexpected behavoiur for the spice model. The problem is that even used as a buffer with 0 feedback resistor the AD8676 shows an output noise of more then 20 nv/rtHz @ 1kHz that is a little bit less than 10 times the datasheet declared voltage noise density @ 1KHz.
Is it a pspice model issue or is there something that i have not considered?
I have simulated the noise performance of AD8676 and the results were similar to your data. I have forwarded this to the Applications Engineer of AD8676 and he confirmed that SPICE model has an error after he simulated the part himself. He also conducted an actual bench test and his results matches the datasheet.
Thank you for catching this bug on the model. We are currently resolving this and will get back to you ASAP.