Hello all,
I have been searching for a reference or an article on how to calculate the least amount of op amp’s feedback resistors for the benefit of noise. I noticed that there is a specific reference and chart on AD797’s datasheet which recommends “100 ohm”. But how can I chose the resistor value for LM4562 or other ones for example? I would appreciate if you may help me or share a reference on this matter to me.
Yours,
Alan
I have been searching for a reference or an article on how to calculate the least amount of op amp’s feedback resistors for the benefit of noise. I noticed that there is a specific reference and chart on AD797’s datasheet which recommends “100 ohm”. But how can I chose the resistor value for LM4562 or other ones for example? I would appreciate if you may help me or share a reference on this matter to me.
Yours,
Alan
Last edited by a moderator:
You need to find the lowest resistance load the opamp can drive without distortion rising (this might be 2k in many cases, sometimes as low as 500 ohms, check the datasheet).
Then you arrange than the load seen by the opamp's output (due to the feedback resistors and the actual load in parallel) is not less than that minimum.
That's going to give the lowest value resistors in the feedback network and thus the lowest noise contribution.
Since the AD797's voltage noise is very low (0.9nV/√Hz I think), this is crucial for that device (0.9nV/√Hz is the voltage noise of a 50 ohm resistor).
With a more usual opamp with say 4nV/√Hz the equivalent noise resistance is 950 ohms, so the feedback resistors can be larger.
Note that the resistance seen by the opamp's inputs from a resistor divider is the parallel combination of the resistors, so for instance
a 1k:100R divider's equivalent noise resistance is 91 ohms, not 1.1k
Then you arrange than the load seen by the opamp's output (due to the feedback resistors and the actual load in parallel) is not less than that minimum.
That's going to give the lowest value resistors in the feedback network and thus the lowest noise contribution.
Since the AD797's voltage noise is very low (0.9nV/√Hz I think), this is crucial for that device (0.9nV/√Hz is the voltage noise of a 50 ohm resistor).
With a more usual opamp with say 4nV/√Hz the equivalent noise resistance is 950 ohms, so the feedback resistors can be larger.
Note that the resistance seen by the opamp's inputs from a resistor divider is the parallel combination of the resistors, so for instance
a 1k:100R divider's equivalent noise resistance is 91 ohms, not 1.1k
The feedback resistors load the opamp. As long as this loads combined with the other load stays below the max mA the opamp can deliver, you should be fine. This means that at low output voltages, you can have lower feedback resistors than at higher voltages.
Be sure to keep total power dissipation below the max too. This means that the supply voltage should be as low as possible and adequate cooling should be ensured.
Be sure to keep total power dissipation below the max too. This means that the supply voltage should be as low as possible and adequate cooling should be ensured.
Thank you Mark for the reply , it was very helpful, I will try to calculate an op amp based on your explanations and share it here , regards
Alan
Duplicate threads merged.