7W IC amplifier

That's fine for experimentation but will not give real world "legal" number s. Running it unloaded will not hurt anything. That can hurt a tube amp or an amp that has output transformers but not a regular solid state amp.

Go ahead and run it into clippind iht the spectrum analyzer connected. I ha ve neve done this actually but I can tell you what will happen. As you get further and further into clipping you will se odd order harmonics on the sp ectrum analyzer. You are actually seeing mathematics at that point. All the numbers are not there but the concept is. Figure out the numbers later.

amp or

get

All

later.

Do you think it is better to add 8 ohm load and connect the spectrum analyzer to 8 ohm load. Also should I use sweep in a function generator to model audio signal ?

--------------------------------------- Posted through

On Tue, 04 Aug 2015 20:18:00 -0500, "chess" Gave us:

A speaker is a dynamic reactive load, so simply reading a signal dumped into an 8 Ohm resistor will not give definitive results.

Place a speaker on it and remove the cone if you want to NOT ...actually transduce sound from it. It will be slightly less damped but far closer to real world than a resistor. You can also at that point force toothpicks in around the interior of the voice coil to jam it from moving at all., which may also be closer to real world, and still far closer than a pure resistor.

Better use a dummy load at the output - say use 10 ohms rated at 100 watts (so it does not overheat).

It depends on what you want to do. If you want to mimic worst-case load, sure use an 8-ohm load (but it'll have to be a high-power resistor). If you just want to see something wiggle, it's probably not necessary. A 100-ohm load wouldn't hurt and would take a lot less power.

Again, it depends on what you're trying to see. A sweep will show you something of its frequency response but it doesn't mean much without the real world around it (the speaker or whatever you're using). Note that a sine wave (or a swept sine wave) doesn't resemble audio much at all. It has a much higher "power" than normal audio. Audio has a much higher "crest factor" than a sine wave.

Please don't help those leeches.

On Wednesday, August 5, 2015 at 12:04:45 AM UTC-4, DecadentLinuxUserNumeroUno

With a good design there will be no discernible difference between loading it with a speaker or a resistor. But it sounds as if this is a school assignment. So why do do both and see if what difference you can detect. The idea is to learn, right?

Dan

...actually transduce sound from it. "

Nope, two speakers face to face in phase. Of course this requires two amps but when you measure the output of a stereo amp you should be driving both channels anyway according to the FTC.

Actualy if you had two 4 ohm speakers they could be put in series on one channel.

Or hell, just put one speaker face down on a table or something.

Speakers have enclosure resonance, variable impedance, and coil inertia that's much messier than R, L, and C dummy loads.

The best way to simulate speakers without speakers would be to send test signals into both ends of the amp. A perfect amp would always have an output impedance of zero ohms; a signal current injected into the output would have no effect. In reality, you might find that the impedance goes up near clipping or when rapidly changing from push to pull. It might even react badly with overshoot. Such defects could become noisy glitches on a real speaker.