I’ve read all of your responses & each one makes a good point. I liked especially the circuit proposed by Alex P for its simplicity.
Phil A points out that the average audiophool does not know what to do with that kind of information nor does it matter. But none of us in this discussion is an average audiophool far as I can tell. And the audio gate still provides a lot of information without the advantage of switching at zero crossing. But that would be good to have as well. The trigger output to the scope provides a stable display. For the tests so far I’ve been using a Pico Tech ADC-216 Scope/Spec A. 16-Bit gets 96 db dynamic range on the SA.
Patrick T mentions the use of pink & white noise as a good test & I fully agree. I just never got around to building or buying such a piece of TE. Would still like to do that at some point.
So I will not make any excuse at all for the path I took. There are many detours in life (And TOOB Audio) so I took this one! Even though the switch I built is arcane it works well & enabled me to get the information from this amplifier (And others) that I wanted. It also satisfied my curiosity with regard to a simple balanced modulator, which it is.
I first became curious of this technique as I read an article by a General Radio Applications Engineer, James K Skilling on ‘Testing with Tone Burst Signals’. That was in Dec 1966. I still have a copy. If you look on the web you can find some of the GR 1396 Tone Burst Generators for sale used. The price looks good but at this age I guess I will pass.
This technique can be used to test loudspeakers & room acoustics as well as amplifiers. So worth having for those involved in that kind of work.
For others reading this it may come as a surprise that their amplifier will not be capable of full power output to the load if it is driven by 2-tone signal. A quick look at some numbers will illustrate this fact. If your amp can deliver 8 watts into an 8 ohm load at clipping you would measure 8 volts rms. That means that clipping occurs at 2.8 times 8 volts or 22.6 volts peak to peak.
The peak to peak spec is what will limit the amplifier’s output. If you have an output of a 2-tone signal each of 4 volts they will add & subtract in such a way that the peak to peak limit will be reached regularly at a rate determined by the test frequencies selected. The formula for power is P = (E^2 / R) so each signal contributes 2 watts to the load. That is only 4 watts total, one-half of the single tone condition. As the test signal becomes more complex, maximum power delivered by the amplifier may be even less. The actual voltage & power delivered to the load can be confirmed by using one of the new precision DMM/Wattmeter instruments. In this example the average power is 3db below the maximum.
All this reminds me of the limits on amplifiers found in a modern communication system. The complex modulation schemes are a problem. For example, QAM (Quadrature Amplitude Modulation) uses simultaneous angle & amplitude modulation. For 16-QAM there are three possible AM levels & eight possible angles. Without going thru the maths, the peak to average symbol power is 2.55 db while the dynamic range is 9.54 db.
These levels look a lot like what happens in the music we listen too. Add another digital channel & you need another 3db of headroom & so on. When a base station becomes overloaded with data there is a good possibility of errors. The Boonton 4500B & 4540 Test sets are useful in analyzing these problems.
See a paper on these problems at this link-
http://www.scribd.com/doc/22678090/A...pr-%E2%80%9893
A paper by Anthony New published in Electronics World makes a very good argument that THD testing is a waste of time. The article was reprinted in the Jan & Feb 2001 issues of AudioXpress magazine. New makes an excellent case for IMD testing with as many as four SGs phase locked to each other. He also covers 3rd order IMD to determine IM3, an important spec for any amplifier, mixer or converter. But these techniques for whatever reason seem to be used only in RF systems.
An RF Generator I’ve sold is the Aeroflex (Marconi) 2026A/B. It can run up to three frequencies simultaneously, all phase locked to the same time base. That allows an important IMD test of RF Comm Systems. See it at this link-
http://www.aeroflex.com/ats/products...erator~12.html
Two signal IMD tests are relatively simple to perform for audio & provide useful information. The 3rd & 5th order show up quickly if there are problems.
If interested I can copy most of this stuff to pdfs from my files. If there are problems pulling up the information I’ve referenced let me know at-
johnnhelen4 at gmail dot com
A while back I bought a Pico 5203. It has quite a few goodies on board. Now that I’m officially retired I will be trying something new again. Will let you know. Long as it don’t get in the way of the yard work.
But at 80 years the bicycle & skis seem to dominate my life. So electronics in the shop may have to wait!
Cheers to All, John