What is the corner frequency of the input current noise density for the OPA140? The wide band input current noise density for the OPA140 is listed as 0.8 fA/rt-Hz at 1kHz but I work down in the instrumentation band where 1/f noise is the killer. Are there any plots for the input noise current density?
Hassan,
I don't believe that much is known about the possible flicker region of current noise on the OPA140. It is extremely difficult to isolate and measure current noise of a low input bias current FET amplifier. We generally resort to shot noise calculations to estimate current noise of these amplifiers. The numbers that this produces usually rule out current noise as a significant contributor in commonly used circuit configurations.
Can you provide information on your circuit configuration or application? With some details, we can determine whether current noise is likely to be an issue.
Regards, Bruce.
I'm using the OPA2140 in two locations. The first location uses the OPA2140 in a 1/3 Hz unity gain Sallen Key low pass butterworth anti-aliasing filter that drives a 24 bit ADC. The two resistors in series with the non-inverting input of the filter amplifier are both 309 kOhms. I'm concerned about this large resistance interacting with the noise current of the OPA2140. The feedback capacitor is 2.2 uF and the shunt capacitance is 1.1 uF.
The second location of the OPA2140 is in the "I" and "D" amps of a PID of a null balance up stream of the anti-aliasing filter described earlier. The "I" term amp as a series resistance of 100 kOhms and a feedback (integrating) capacitance of 6.8 uF. The "D" term amp has a series resistances of 75 kOhms in series with 0.22 uF and a feedback capacitance of 75 kOhms in parallel with 0.22 uF.
I am also concerned about the noise generated in these two terms but particularly the "D" term since it is notoriously sensitive to noise.
I'm currently getting around 22 out of 24 bits. Should I be quiet and happy?
I think it is very unlikely that current noise will be a significant contributor in any of these cases. Noise in the D term will, of course, tend to emphasize high frequency noise.
You have not indicated the nature of your control system. What are you measuring? The sensor or input signal may be a major source of noise. It may be instructive to measure the output noise of the PID system with sensor input and with shorted input to see how they compare.
You have answered my noise questions but now I have a rail to rail question.
How well will the OPA2140AID drive the input to a delta-sigma ADC through a 299 Ohm-1uF RC filter?
The equivalent input structure of the ADC (LTC2440) is a 500 Ohm series resistor and a 5pF sampling capacitor.
I suspect the rail-to-rail nature of the op-amps output stage will struggle with this small resistive load but I want your knowledgeable input.
--hassan