Circuit Nodal Voltage

I have a voltage divider feeding into an op-amp voltage follower buffer circuit. I'm trying to calculate the voltage level at the non-inverting input of the op-amp vs. the source voltage. The possible supply voltage is roughly in the range [0V, 5V]. First I tried using a simple voltage divider, so:

$$V_A = R_2 / (R_1 + R_2) * V_S$$

However, when I tried simulating this circuit in TINA-TI, the nodal voltages don't match those calculated.

I thought that was because the notion that there is no current flowing through the op-amp input wasn't truly non-zero, so I tried compensating for the op-amp input resistance and capacitance. This time, the output voltage should be:

$$V_A = (R_2 || R_3) / (R_1 + R_2||R_3) * V_S$$

While this equation does match TINA-TI's simulation of the no op-amp circuit, it still doesn't match the op-amp circuit.

What else am I missing?

edit:

further information: I'm using the LMH6612 op-amp.

Suppose I want to target an output voltage of 2.5V. According to my first equation the required source voltage is ~2.82V. By recursively updating my simulation I calculated that the actual required source voltage is ~3.47V. Using the second circuit the required source voltage is closer, but only marginally at ~2.87V.

• How far off is the simulator from the analytical solution? Did you account for the op-amp offset current. At 0.01 uA, this would cause about 1 mV difference. – The Photon Aug 18 '12 at 17:00
• I updated my post with a test scenario. The difference is on the order of ~0.5V. – helloworld922 Aug 18 '12 at 17:08
• Is this happening at DC? – s3c Aug 18 '12 at 18:06 