# Can I see an analog signal with changeable frequency on oscilloscope?

I have a sinusoidal signal with changeable frequency (which implies that it is not periodic). I wonder is it possible to see it on oscilloscope? Does it show non-periodic signals? As it is apparent my signal has different frequencies. The range of frequencies are less than 100 kHz.

-

The only type of oscilloscope that won't be able to display a non-periodic waveform is a "sampling" type (which are now referred to as "equivalent time sampling" to distinguish them from ordinary real-time sampling digital oscilloscopes), and that type is mostly used at frequencies where real-time sampling is difficult (GHz).

Even a cheap digital oscilloscope samples fast enough to see a snapshot of a signal with a bandwidth of 50MHz or 100MHz in a single shot (0.5G samples per second is common in inexpensive scopes).

It perhaps needs to be mentioned that the oscilloscope will only have a certain "depth" of memory so you will only get a kind of keyhole view of the signal, starting at the trigger and extended some thousands of samples after the trigger. After that snapshot, the oscilloscope may trigger again or it may not, depending on how you've set it up, but you cannot gather true continuous data from the input with a typical oscilloscope, just snapshots of it.

For continuous acquisition (say to capture the signal over many seconds, minutes or hours) you need a data acquisition system capable of taking samples at at least double the highest frequency component of interest, and of sufficient bit width for the accuracy you care about, and storing it to suitably capacious memory in real time. It's not hard to fill gigabytes of memory if you're gathering high precision data over multiple channels.

-

What, exactly, do you mean by "possible to see it on oscilloscope"? The trivial answer is yes. Just set the scope mode to single shot and trigger on a zero crossing. If you don't see the signal you're looking for, just trigger again (and, possibly, again and again. Patience is advised.)

If by "see" you mean, can you reliably trigger on the areas of interest which you have marked on your illustration? The simple answer is probably not, or at least not with your scope's built in trigger facility. There are various circuits which you could build which would do the job of providing a fairly reliable trigger, provided you have a good idea of what you're looking for. It would help if you're using a digital scope, of course, since single-shot viewing is hard without it.

In the case you've shown, where the special areas are marked by a lower frequency than the unwanted signals, you use a comparator to convert the signal to logic. You then feed this into the clock input of two retriggerable one-shots, such as an MC5438. One half has a pulse width very slightly less than the period of the desired signal, while the second has a much shorter period. The output of the first drives the reset of second. The output of the second is used as a scope trigger. This will discriminate against frequencies higher than desired.

-
Any decent Tektronix can "reliably trigger on the areas of interest" as illustrated (using one of its fancier trigger modes). We are not in the 70's anymore. :) –  oakad May 19 '14 at 2:50

Any half decent modern oscilloscopes will be able to show the waveform as you have shown it. Scopes can show non-periodic signals and these can be easily triggered and "captured" and held on the display.

-

CRO will show the waveform. But measuring time period may be difficult since waves will be running most of the time. Adjusting trigger for such signals are difficult and this frequent change in trigger will feel the viewer that the waveform is running. But a DSO can show you the wave form correctly.

-
Can you explain more? –  AMIR May 18 '14 at 19:42
The waveform will be shown as multiple waves running across the screen. –  nidhin May 18 '14 at 19:55

I'd probably try to read out the data with a (digital) oscilloscope, and then use an FM demodulator in software to see the frequency shift over time.

There is a package for Matlab providing an fmdemod function, or you can use GNUradio.

If you need this measurement repeatedly, look for an oscilloscope with FFT analysis options, or use a spectrum analyzer with FM vs time display.

-