Continue to Site

# Average of signals

#### Curios123

##### Newbie
Hi! I understand that signals that have slow variatons in time cand be approximated with their average value but I don't understand how the system will see them. For example, why we cand say that a signal that has 0 average over time will be seen at the system output as 0?

The averaging time obviously must be larger than the signal period.

Can you give an example of the type of signal you are referring to?

Hi,

and do you talk about analog averaging or digital?

Klaus

The averaging time obviously must be larger than the signal period.

Can you give an example of the type of signal you are referring to?
Yeah, sure. Let's say that an offset voltage is read with a T period. During one period +Voff is read, during the 2nd period -Voff is read. If this T is let's say 1u and the average is considered during 1ms, can we say that the system will have a 0 output offset?

A rephrase of my question will be: on what grounds we make this approximation based on average value of a signal?

Hi,

and do you talk about analog averaging or digital?

Klaus
Analog

Hi,

* low pass filtering? What technique, filter order..
* or true averaging over a dedicated period of time. (equal weighting over a known period of time .. using integrator and analog switches..)

Klaus

You’re asking a question in a vacuum. The average value of the AC coming out of your wall socket is zero. Would you stick your tongue in the socket because it’s “zero“ volts?

You need some context.

You’re asking a question in a vacuum. The average value of the AC coming out of your wall socket is zero. Would you stick your tongue in the socket because it’s “zero“ volts?

You need some context.
Actually, that’s not totally correct. Lets say it AC through a transformer.

Yeah, sure. Let's say that an offset voltage is read with a T period. During one period +Voff is read, during the 2nd period -Voff is read. If this T is let's say 1u and the average is considered during 1ms, can we say that the system will have a 0 output offset?

A rephrase of my question will be: on what grounds we make this approximation based on average value of a signal?
There is not one single answer here. Length of signal is critical for measuring dc offset.
dc offset will show on the average amplitude in time domain or as power in dc bin in frequency domain.
Imagine single tone say at 10MHz. Time domain could easily mislead depending on length of signal. Frequency domain (dc bin) is expected to be zero if there is no leakage i.e. you are observing complete cycles of 10MHz.
If in doubt I will observe average of successive lengths and see when does it settle.

Hi,

They spend their time for you ... no "please", no "thank you" ...

Klaus