The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> integrated noise
https://designers-guide.org/forum/YaBB.pl?num=1238552542

Message started by LeonQQ on Mar 31^st, 2009, 7:22pm

Title: integrated noise
Post by LeonQQ on Mar 31^st, 2009, 7:22pm

Hi all, Suppose I have a one-pole simple amp with the pole at 10kHz, my signal is at 1kHz. To calculate the input integrated noise, I integrated the noise density upto 10kHz, but what should be the low limit of the integration? The lower it is the more flicker noise contribution will be. How low the integration limit is proper? I am confused.

Your inputs will be greatly appreciated.

Leon

Title: Re: integrated noise
Post by thechopper on Mar 31^st, 2009, 8:11pm

Theoretically the low corner frequency should be related to the amount of time you will be "observing" your signal. That will determine the amount of 1/f noise power that will affect your signal.
For example, integrating down to 1Hz should account for the noise power affecting your 1KHz signal during 1sec.

This is an approximating but works as a rule of thumb.

Regards
Tosei

Title: Re: integrated noise
Post by raja.cedt on Mar 31^st, 2009, 9:46pm

hi tosie,
i am also looking for the answer for this question,but i didn't understand your ans clearly,so can you please explain clearly.

Thanks,
rajasekhar.

Title: Re: integrated noise
Post by thechopper on Apr 1^st, 2009, 7:32pm

Hi rajasekhar,

What I meant is that the lower corner frequecy of interest should be dictated somehow by the timeframe used for measuring your signal.

In theory flickler noise could be very large if you "wait enough time". Or in other words is unbounded as t -> ∞.

So the question you should ask yourself is: how many cycles (assuming a sinuosidal signal) of my lowest frequency signal I will wait for measuring the amplifier's noise?
That number of cycles (measured in sec) should give you an idea of what the min frequency for integrating your noise should be.

Hope this is clearer
Tosei

Title: Re: integrated noise
Post by vivkr on Apr 1^st, 2009, 11:59pm

LeonQQ wrote on Mar 31^st, 2009, 7:22pm:

Hi there,

This topic has been threshed out several times in the past on this very forum and in great detail. Please search the past posts and you will find more discussion. Incidentally, you will run into a problem if you are going to use the pole frequency as the upper limit for noise calculation, since this will account for only some of the white noise.

For the lower end, I would say that this should in some way depend on the memory of your system. All systems are lossy and will "forget" information collected in the past. Also remember that 1/f noise power per decade is fixed. So going from 10 Hz to 1 Hz will add as much noise as going from 1 billion years to 10 billion years. It may be sufficient to ignore 1/f noise at very low frequencies if it does not add much to your total noise number.

Regards,

Vivek