how do we find the period?

[circuit975]

I salute the entire community on my behalf.

when I was working on square wave on signals and systems, I thought that I sometimes confused the period of the signs.

i find that the period of such a sign is 4. but I'm not sure if I did it by heart or really knowing how it's done.

how do we find the period of a graph of this type in the most accurate and scientific way?

so are we looking at starting from one place and ending from another, what exactly are we looking at?

by the way, I can't help saying: I made the drawing myself , as an example , to Decipher what happened.

in short :
what is the logic of finding a period?
what is the logic of this? if anyone knows, the community will honor him.


Thank you very much!

 

[danadak]

Period is basically the time from the start of a cycle to its end, so in your pic
the rise of the waveform to the next cycle rise, or fall/fall. For analog waveform
from the ref point in amplitude on one waveform to the same amplitude point
in the next waveform.

In general for continuous waveforms.

 

[Externet]

To find the period you need a periodic waveform to measure/test/read. And it is 1 divided by frequency.

 

[AnalogKid]

If the waveform is a constant frequency, then the period is the inverse of that frequency. The period is the time between two identical points in the waveform, one cycle apart. For example, the time from one rising edge (such as at -5) to the next rising edge (at -1) is the period.

ak

 

[Tony Stewart]

The plot scale is inconsistent in time (e.g. -7 to -5) but the numbers from each polar edge are equal.

But if you had a repeating binary data pattern that was 1T,2T (for +,-) the cycle would be 3T = 1/f although the pulse is no longer square so there will be even harmonics included that are otherwise null on symmetrical square pulses. That is how a spectrum analyzer would see it. For more detail read about Baud rate vs Bit rate.

 

[circuit975]

The plot scale is inconsistent in time (e.g. -7 to -5) but the numbers from each polar edge are equal.

But if you had a repeating binary data pattern that was 1T,2T (for +,-) the cycle would be 3T = 1/f although the pulse is no longer square so there will be even harmonics included that are otherwise null on symmetrical square pulses. That is how a spectrum analyzer would see it. For more detail read about Baud rate vs Bit rate.

If the waveform is a constant frequency, then the period is the inverse of that frequency. The period is the time between two identical points in the waveform, one cycle apart. For example, the time from one rising edge (such as at -5) to the next rising edge (at -1) is the period.

ak

To find the period you need a periodic waveform to measure/test/read. And it is 1 divided by frequency.

Period is basically the time from the start of a cycle to its end, so in your pic
the rise of the waveform to the next cycle rise, or fall/fall. For analog waveform
from the ref point in amplitude on one waveform to the same amplitude point
in the next waveform.

In general for continuous waveforms.

dear dear friends, thank you for your answers, but first of all I need to understand how the period is found on the graph. after that, I think I'll understand what you're saying.

actually, what I'm trying to figure out is in the simplest sense, how are we going to look at the graph to find the period of this graph? so December. - Until December 5th. -1 ?

if it's between -5 and -1, why?

 

[danadak]

Lets look at cycle from 1 to 5. So 1 is ref point, which has a point in time,
with a sign (direction waveform moving in amplitude), and find the next
point in time with same sign. Then period = t(5) - t(1).

Note this is not a continuous wave, its duty cycle changing with time, so
period, normally for continuous wave, is qualified by noting thats the period
for a specific cycle, not the whole wave train. So period, as normally used,
not quite as useful. And of course f = 1/t, hence frequency also not static
either. Hence for non continuous wave signals period and frequency lose
usefulness, meaning, except on a cycle basis.

 

[Tony Stewart]

For periodic signals, the period of 1 cycle , T is simply one complete cycle using any starting point.

- such as noon Dec 1 to noon Dec 5 = 4 daily cycles.

- or from rising edge to the next rising edge

 

[jpanhalt]

As for reducing by 1, years ago (maybe someone here) described it simply as counting fence posts or spaces. That is, one space has 2 fence posts.

 

[Ian Rogers]

From your diagram

 

[AnalogKid]

if it's between -5 and -1, why?

Because that is the length of one cycle. If a waveform or chart has a repeating pattern, and each pattern is the same, then the period is the time from the beginning to the end of one pattern. That is the agreed upon definition, and has been in place for way over a century. In electronics, a voltage or current pattern is called a cycle, and the measurement unit is time. Based on post #6, it looks like your time unit is days, and one cycle spans 4 days.

Even though your chart is poorly drawn, the period is consistent. The time from one positive-sloped edge to the next positive slope edge always is 4 days, no matter which edge you start with. Same for the negative edges.

ak

 

[Nigel Goodwin]

On Thursday, 22nd May, my daughter sent me a WhatsApp asking "In radio land would you ever use wavelength?" - rather strange, I thought?.

I replied, giving some examples, of how wavelength was very useful for building aerials.

Apparently there was a big 'unit discussion' going on, at the University she works at (as the token Chemist among mostly Physics people). Apparently, one of the two groups she works with, COPS (Complex Photonic Systems), are a wave number group, AQO (who ever they are? - Adaptive Quantum Optics) say you should use wavelength, and Henry (no idea who he is either) says you should use frequency.

Apparently, there are no winners, just a discussion

 

[JimB]

View attachment 150117

The biggest confuser here is the badly drawn X-axis, which has a somewhat non-linear spacing between the marked points on the axis.

JimB

 

[schmitt trigger]

Apparently, there are no winners, just a discussion

Hopefully they won’t start a discussion between the merits of conventional current flow vs. electron current flow.
Literally, sparks would fly!

 

[circuit975]

From your diagram
View attachment 150121

by looking at the one between the rising edge, we will reach the most accurate result. Yes, you are amazing, your narration is extraordinary.. God help you, thank you