Real World Use Of A Function Generator

[rstofer]

Well that's the digital way.  Any ideas on doing it analog?

Somewhere along the way, I came across a theory that ANY periodic signal can be made up of various sine waves of specific amplitude, phase and frequency.

Earlier in this thread there was a link to https://www.eevblog.com/forum/chat/a-valentine_s-day-activity-for-your-scope-and-function-generator/msg1136847/#msg1136847

This is a very cool application of an AWG.  It actually has to be quite sophisticated to be capable of adding in harmonics at varying amplitude and phase.

So, yes, it can be done in an analog manner but there's a reason we are using DDS instead.

[bson]

Somewhere along the way, I came across a theory that ANY periodic signal can be made up of various sine waves of specific amplitude, phase and frequency.

If anyone reading this finds it difficult to visualize, consider the opposite which is also true: that any repeating sinusoidal can be decomposed into a superposition square waves.  As the square wave harmonics added get higher in frequency the "jaggies" in the resulting sinusoidal get smaller.  For the sinusoidal to be continuous an infinite number of square waves (up to infinite frequency) are needed.

Exactly the same applies when trying to add up sinusoidals to a square wave.  Except of course in the real world, where nothing is ever perfectly square, the Dirac delta function can't exist, and nothing is perfectly continuous (because of QM).  So these are mathematical models of the real world.

[IDEngineer]

A recent "real world" use of a function generator was to test the sensitivity of the human eye to flicker when designing a multiplexing scheme for multiple RGB LED's. Read all the references you wish, you'll still be left with lots of variables. 24FPS is fast enough for old film-based movies, so that must be OK, right? Except that LED's have a faster rise and fall time. 30Hz is fast enough for LCD/LED based displays, so there you go, eh? Except that 2D arrays average differently in your field of vision than (nearly) discrete points of (nearly) monochomatic light. And so forth.

So I turned on the function generator, configured it for pulse output, hooked up some LED's, got several volunteers, and varied the brightness (amplitude), duration (duty cycle), and refresh rate (frequency) as my "subjects" viewed the LED's in various ambient light levels. We learned a LOT in a very short time, and it was ridiculously easy to vary every parameter.

Could I have done this without a function generator? Sure. But it would have taken a lot longer, and I wouldn't necessarily have had such accurate (and repeatable) results. I could vary frequency by single integer Hertz. I could vary duty cycle by single digit percentages. I could vary brightness (amplitude) by tenths or hundredths of a volt. All from a convenient front panel with a nice user interface and keypad. I spent my time learning and gathering data, rather than creating and fumbling about with some Rubed-up test jig.

And this is just one, perhaps less obvious application for a function generator. They're also great for their usual purposes too. A scope and meter are more important, but once you have a function generator you'll likely never go without one again.

[bob91343]

So I can use the VCO input to make a sweep generator, from an external ramp wave from another generator.  Not sure what limits I can get.  The staircase generator is a bit more sophisticated.  But I have a third generator, for whatever good that might afford.

Let's see, I can run some pulses and charge a capacitor.  The capacitor will go to the VCO input so I get an approximation of a staircase.  End of wave could give me a reset pulse to discharge the capacitor.  I would need some extra circuitry for that.