# How to calculate the input-refered noise for two OTA in parallel?

1. ## How to calculate the input-refered noise for two OTA in parallel?

Hello

I'm a bit unsure on how to calculate the input-refered noise for two OTA in parallel.

Lets say that both OTA have a input referred noise level of Vn = 60 nV/sqrt(Hz) individually.

I have read that if I regard two noise sources connected in series as uncorrelated i can add them.

If i connect the two OTAs in parallel and refer the noise to the input, will the noise become bigger or smaller? •

2. ## Re: How to calculate the input-refered noise for two OTA in parallel?

Output noises are uncorrerated.
So input reference noise is 3dB down.

Gm is twice.
On the other hand, Iout_noise is by sqrt(2).

1 members found this post helpful. •

3. ## Re: How to calculate the input-refered noise for two OTA in parallel?

So according to theory the resulting input refered will be reduced by half to 30 nV/sqrt(Hz)?

The reason why i'm asking is that im simulating the individual stages to be 58 nV/sqrt(Hz) and 67 nV/sqrt(Hz) individually.
And when im simulating them in parallel i get 56 nV/sqrt(Hz).

Does that sound reasonable? •

4. ## Re: How to calculate the input-refered noise for two OTA in parallel? Originally Posted by nidare So according to theory the resulting input refered will be reduced by half to 30 nV/sqrt(Hz)?
Wrong.
It has to be 42.42, if both gain is same and operation points are not changed. Originally Posted by nidare The reason why i'm asking is that im simulating the individual stages to be 58 nV/sqrt(Hz) and 67 nV/sqrt(Hz) individually.
And when im simulating them in parallel i get 56 nV/sqrt(Hz).
Does that sound reasonable?
You have to take gain difference into consideration.
And operation points are changed by connection.

1 members found this post helpful. 5. ## Re: How to calculate the input-refered noise for two OTA in parallel?

I see, I understand it now. --[[ ]]--