Output voltage noise spectral density of transistor

[deepa]

I observed output voltage noise spectral density of my opamp i can see a 1/f structure and then it doesn become constant (white noise),but its value is further decreasing with a higher slope.Is this 1/f^2 noise?
shouldn thermal noise domainate this and the spectrum should be white there after.

 

[A.Anand Srinivasan]

Re: transistor's noise

the noise spectrum depends purely on the surroundings and the material and design... so i dont think it is possible to do a analysis on noise without having an in hand experience on it...

 

[barath_87]

Re: transistor's noise

1/f niose is flicker noise which is at the device level ....... correct me if i am wrong ... i think we cannot do anything unless we change the technolgy.....

 

[A.Anand Srinivasan]

Re: transistor's noise

by changing technology i dont think flicker noise can be reduced.... by changing parameters, material, design etc the noise inherent to that particular material or design etc can be changed.....

 

[ieropsaltic]

Re: transistor's noise

Flicker noise's effect can be decreased by increasing both W and L of the MOSFETs while keeping W/L constant (for proper circuit operation). This should shift the corner frequency to lower frequencies. This way you don't have to change technology.
The 1/f^2 may arise due to filtered white noise. A pole after the noise source would cause a 20dB/dec which may appear as 1/f^2 but I'm not sure if this is your case .

 

[krashkealoha]

Re: transistor's noise

In addtion to ieropsaltic's inputs, if you are able to use a pmos transistor instead of a nmos device in your design, this will also reduce your 1/f noise.