Current noise of SSM2212 (or MAT12)

Hello,

We have contacted the product owner of this part to address your concern.

Thanks and regards.

Thanks - any insights on this?

I am also interested in the answer to the current noise specification. Given the detail provided typically from ADI parts, it is puzzling that a matched low noise pair like this has not even a typical current density specification. Curves are always nice too.

OK, I  have the current noise figured out. Current noise is directly related to the DC base bias current that flows because the current noise is directly related to shot noise.

The current noise tends to vary much more than voltage noise, but arises from shot noise from current flowing through the base of the transistor The shot noise spectral density can be estimated as:

i_n = √(2*I_B*q)/√Hz

Where

I_B = the bias current of the input transistors in amps

q = the charge of an electron, or 1*10^-19 C

Not that if the design is bias compensated or used in a current feedback configuration, this equation will not work. The problem for a discrete design is that the current density number is going to depend on the chosen bias current and the Beta of the transistors since the leakage current is the bias current is the current that will be flowing through the base. The input bias current of the SSM2212 is specified, but only at 10 uA as 50 nA.

Observing the above equation, you can see that worst case noise is going to occur when the transistor beta is at a minimum. The 50nA comes from the worst-case beta spec of 200 at 10 uA. So in order to estimate the current density, you are going to need to know what current level you are going to be biasing the transistor at in your completed circuit. From the specified betas and the above equation, you can estimate the mid band current density you will see in your application.

As we know, the voltage noise density is also going to vary according to figure 4 in the datasheet, with voltage noise tending to decrease with increased bias current, so it is inversely proportional to current density. Then you will have to choose a balance between these two noise sources to select the optimum bias current.

That of course is going to depend on the assumed source impedance as well as the noise gain of your final circuit configuration. I hope this explanation will help other people.