rough method for clipping detection

So I understand that the only way to accurately detect clipping on the outputs of my amplifier is to use an oscilliscope to measure the waves.  However, I need a method to roughly determine clipping with a voltmeter.

My understanding of clipping is that the amplifier is asked to deliver power beyond its voltage output capabilities, so the amplifier then attempts to meet the output demand using amps instead.  Going off that idea, I decided to try setting my amplifier gains by turning them up until the voltage stopped rising.  My guess was that that would be the point of clipping...at least roughly speaking.  However, the amplifier has a clipping light that comes on at roughly 33.7 volts according to my voltmeter, while the highest voltage that my voltmeter read was 65 volts.

I am wondering if my test could be considered good enough.

The amplifier I tested was a MB Quart RAA1000.  It is rated to output 250 watts at 4 ohms, 500 watts at 2 ohms, and 1000 watts at 1 ohm.  I determined that it was capable of outputting roughly 284 watts at 4 ohms, 568 watts at 2 ohms, and 1136 watts at 1 ohm without the clipping light on the amplifier turning on.  At the highest voltage that the amplifier put out, the power should be roughly 1056 watts at 4 ohms, 2113 watts at 2 ohms, and 4225 watts at 1 ohm.  The wattage appears believable when the amplifier's clipping light activates.  It does not seem very likely, however, that the amplifier is capable of the output at the point that I determined to be the highest voltage.

Any thoughts, advice, and insight, would be greatly appreciated.

When I worked at Adire, I produced a video of one of our Brahma's playing full range, and ran the amplifier into clipping so you can hear what it sounds like.  Here's a link to it: http://realmofexcursion.com/videos/Adire/brahma12.13.wmv  Unfortunately Realm of Excursions mixed several videos I did together not understanding the point, so you will find the clipping video at the end.  The other videos were to show the very low inductance of the XBL^2 motor and it's impressive upper extention. 

I'm not sure what you're asking but there's no way that amp is going to put out 4000 watts.  If you have a 250 watt sub why would it matter? That sub is rated at 250 watts and maybe it can handle more but you're playing with fire when you start exceeding the rated power handling. 

That amp has a max of 120 amp draw i believe.  120 amps x14 volts x.85 efficiency=1428 watts max. and that's being very generous in every regard.  Anyways that's running at 1 ohm and you'll need lots of power to deliver 120 amps to it.  Personally I would say if that amp has a light that tells you its clipping at 284 watts @ 4 ohms, set it to right below that and get what you can out of it.  Probably not going to get much gain if any after that with a few more watts to that sub anyhow.

What you need to consider for your test is how the voltages are being measured.  The a/c wave should be a perfect sine wave right up to immediately before clipping.  That would be your max. gain setting and therefore your maximum voltage.  Now you continue to raise the gain and the voltage increases.  What's happening is you're "clipping" the top of the sine wave and it's approaching a square wave.  When you've really got that clipping hard it basically looks like a square wave.  A square wave goes from max to min voltages almost immediately instead of in the nice gradual manner of the sine wave before clipping.  Now your voltmeter is measuring the average voltage of this square a/c wave and it of course is much higher.  The voltage tells you nothing of when your amp actually started clipping and therefore is pretty much useless for setting the gain at it's perfect level.  So no your test isn't really good enough.

So, according to your statement, my test measures the peak voltage since the voltmeter is measuring the RMS voltage and because the only thing it is averaging is the peak voltage then the number it is outputing is the peak voltage.  So, with that figure, I can take the Root Mean Square of that number and have a rough idea of the highest unclipped voltage.  Does that sound okay to you?  I appreciate the input zhalverson!  I've been hoping for that advice all day.

Please do keep in mind that I am competing in SPL competitions and that I am not interested in the RMS Wattage that was established with the intention of avoiding the meltdown of the adhesives holding the voice coils in place.  I am more interested in the wattage that will drive the driver to its Xmax limitations.