1
\$\begingroup\$

I am reading a book about measurements in communications. There is a lot of information about how to measure frequency and period of signals.

I understand that we should measure the frequency to know the frequency of the signal, type of the signal (periodic or not), need to use filter or not, etc.

But why do we measure the period? What type of information do I get after measurement?

enter image description here

\$\endgroup\$

2 Answers 2

1
\$\begingroup\$
i have understood that we should measure the frequency ... why we measure the period?

Because period is 1/frequency, and frequency is 1/period. If you know one, you know the other.

Given that, you measure the frequency or period, whichever is more convenient for your setup. Sometimes it is easier or more accurate with the available hardware to measure the time of one or more cycles than to count the number of cycles in a specific time. This is particularly true for low frequency (long period) signals.

Many microcontrollers include hardware that makes measuring period fairly easy. It is often possible to have a timer free-run from a fast clock, but have that timer value be captured by the hardware on specific events. By computing the latest capture value minus the previous, you get the period between two events to the resolution of the internal clock.

\$\endgroup\$
4
  • \$\begingroup\$ Why we must measure a signal in one period? \$\endgroup\$
    – user193562
    Commented Jul 17, 2018 at 11:37
  • \$\begingroup\$ to avoid aliasing? \$\endgroup\$
    – user193562
    Commented Jul 17, 2018 at 11:38
  • 1
    \$\begingroup\$ "Why we must measure a signal in one period?" We don't. Any known number of periods works, as I said. \$\endgroup\$ Commented Jul 17, 2018 at 11:41
  • \$\begingroup\$ i have found a picture in a book, and the picture illustrates the measurment in one period. ( i have added in quastion) \$\endgroup\$
    – user193562
    Commented Jul 17, 2018 at 12:07
0
\$\begingroup\$

When you mix two high frequency clocks it is very accurate to measure the phase modulation if they are in sync and frequency shift is also very accurate. Jitter reduction is ~-10dB per decade of BW reduction.

There is a good thick book I read on this subject “Time and Frequency Measurements” in the mid 70’s so I won’t give a long answer but a precision reference frequency at 10MHz in an ovenized OCXO is the most common with options for GPS time/freq sync. Phase or time interval is inverse of frequency and I once measured the phase modulation of a 4xx MHz Tx using a precision HP Time Interval Counter with 1 degree accuracy and 0.01 ppm frequency error on power up for GOES 1.

Measuring frequency of heartbeats often takes averaging a bunch of cycles but it is faster to measure time interval and compute the inverse frequency.

In the old days with mechanical watches and no digital equipment watch repair men used a stripchart recorder to print the horizontal phase of a 2Hz sawtooth and the sound printed the dot and would show the beat frequency of time in Delta f in less than - minute to null the offset sawtooth wave result to get an accurate watch. This is called Beat Frequency offset method.

Jitter of random intervals is always reduced by the square root of the samples averaged. .. unless there is no jitter then averaging makes no improvement, in resolution, so one must add jitter.

This is just a sample of examples.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.