Here's a quick & dirty intro to the CRO (Cathode Ray Oscilloscope), how to set it up properly & how to identify clipping.

For this test I've used:
BWD-509B CRO
Digitor 707 Digital Multimeter (DMM)
Alpine CDE-7831 Headunit (with BBE, BASS control, 2x 4V Preouts)
dB Drag's dB Jam 6 CD for test tones

Before we start FAQ:
Q. Can I use this same method for testing amplifier clipping ?
A. Yes, you use the same method. Just disconnect the speakers from their terminals and connect the CRO probe to them instead. You can also connect the DMM to find out the voltage output. Just make sure the headunit is running unclipped before you start

Q. Do I need to run test tones ? Cant I use music instead ?
A. No. Music doesnt give a clean, clear signal and makes it hard to identify the trace. You must use a tone source (CD or generator)

Q. Can I do this for my full range amps too ?
A. Yes, just use a higher range test tone around the center of the frequency range your amp will be driving

Introducing the CRO



Here's my CRO. Its an Army Surplus BWD509B single trace CRO thats not too flash and over 20 years old. But it does the job, handles more power than I can ever feed it, is reliable and cost me a carton of VB tinnies.

CRO's looks pretty daunting, so what does everything do ?

Intensity Knob (Top Right) - This turns on the CRO and adjusts the brightness of the trace on the screen. Adjust it however you like.

Red Dot above Intensity - Indicates if the CRO's turned on. This is handy as the CRO can take a few seconds to warm up and start showing a trace.

Three Knobs Under Intensity - These are used to adjust the traces horizontal position on the screen, adjust focus, on screen start & end points & on screen trace width. Adjust these so the trace is clear, and the trace starts on the left of the screen and goes across to the right all the way.

Volts/CM Knob (Bottom Left) - This adjusts the sensitivity of the CRO in Volts, and adjusts the size of the waves on the screen.

AC/DC Switch (Under Volts/CM Knob) - This switches between AC & DC volts. You want to measure sine waves, so put this to AC.

Vertical Position Knob (Center Bottom) - This adjusts the vertical position of the trace on the screen. Adjust it so the trace is roughly centered.

Time/CM Knob (Bottom Right) - Adjusts the bandwitdth of the display. A bandwidth out of range will results in a CRO trace that doesn't pulse and/or display a signal.

Trigger Level (Under Vertical Position) - This is used to adjust the trigger level if you are testing a conductor carrying multiple signals. For our purposes, Automatic Trigger is the best setting.

Three Sockets Bottom Right - These are used for calibration & advanced CRO techniques we dont need, so you can safely ignore them.

Not pictures is the CRO probe. A CRO probe looks a bit like a pen, when you pull the top down a little hook pops out, thats the CRO probe tip. Connect this to the outer ring of the RCA, the little aligator clip is for the reference, so connect it to the center pin of the RCA. As we are measuring a sine wave, the CRO is in AC mode and polarity isn't an issue. This is the same when CRO'ing an amp.


Here's three examples of incorrect CROing techniques:


This is what happens when you try to CRO with music instead of a test tone. Notice how the wave is not uniform and very jagged ? This makes it difficult to know what your measuring. The line through the middle indicates silent sections (image take with a slow shutter)


This is a nice sine wave, however the Volts/CM sensitivity is set too high and can make it difficult to define the tops of the waves. The line indicates silence (slow shutter speed, silence is due to track change)


Another nice sine wave, however this time sensitivity is set too low and the peaks/crests of the waves are being chopped off by the screen.

CROing The Headunit

To do this I've connected the CRO probe to one of the headunits RCA's (tip on ring, clip on pin) and adjusted the CRO to the correct ranges.
The DMM has been connected to the other one.
I am running a 60hz test tone out of the headunit. 60hz is the accepted standard for testing in sub-bass regions. This headunit is for a podium, so sub-bass range clipping/pre-out voltage is what I'm interested in.
The headunits setting have all been set to 0 (EG. volume), flat (EG. EQ settings) and off (EG.BBE)

All images taken with a slow shutter speed, so sorry about the blurriness.

Slowly winding up the volume, at 34/35 it shows a trace like this:

This is a nice clean trace showing a properly formed wave. Note the DMM voltage is showing 3.6V.

Upping the volume to 35/35 (full) shows a trace like this:

It is difficult to see clearly, but the crests of the waves are slightly mis-formed with an irregular ripple on top. This indicates that the headunit has started to clip. Note the DMM is now showing a full 4V. 4V @ Onset Of Clipping is its rating, Alpine knows headunits.

Now with 34/35 volume, upping the the BASS Control from 0 to +1 shows a trace like this:

INSTANT SQUAREWAVE ! Ever heard me say never use BASS Boosters or BASS Controls ? This is why. This is a badly clipped signal which is square waving, running speakers with a signal like this will result in damage to them. Note the DMM reading now, 5.0V ! This is due to the nature of squarewaves increasing the Duty Cycle of an AC signal.

For curiosities sake, I've reduced the volume to 20/35 with BASS Control +7 (maximum). This is the trace:

Even worse square wave ! Again, this is why I say dont use bass boosters or up bass controls. Again, this signal will cause damage to speakers. Note the DMM again, 5.8V this time. The larger square top means an even longer AC duty cycle, hence the voltage has again risen.

Finally, what happens if we have 34/35 volume, everything flat and turn BBE on:

Another square wave which will cause damage. Voltage is up to 5.9V now, hence the duty cycle has increased again due to even worse clipping.

Some notes:

It is not the signal itself that causes damage to a speaker, but the amount of power it is receiving. Now an amplifier that produces, say 100w (unclipped) will produce significantly more than 100w when heaviliy clipped. This is because a clipped wave has an increased duty cycle. (and therefore more RMS power)

For example, say you had a speaker that could handle 100w RMS.
(To avoid confusion, lets say this speaker can handle a 100w RMS tone for an infinite amount of time without damage)
We then connect this to an amplifier that outputs 100w RMS, unclipped.

Since the speaker can handle 100w RMS, no damage will occur.

But if we were to clip this amplifier, then it would output significantly more than 100w RMS and therefore it could damage the speaker when clipped.

On the other hand, if we hooked up that 100w RMS speaker to a 20w RMS (just before clipping) amplifier and regardless of how hard you made it clip, it still would not damage the speaker because it cannot output 100w RMS.

Conclusion:
A clipped amplifier produces more power (up to a point of course) than an unclipped amplifier.
It is this extra power that will make it more likely to damage the speaker, not the shape of the wave.

Of course, we all try to avoid significant clipping, even when our speakers could handle the extra power, simply because audible clipping sounds bad.

Secondly, keep in mind that we don't listen to test tones, we listen to music. Thus some degree of gain overlap is often preferred. Ie, if you wanted to use 10dB of gain overlap, then you would set your gains with a -10dB test tone.

If all of this use of the CRO and test tones confuses you, then you can still set your gains by ear. You would select a common piece of music that you normally listen to. Then you would set the gain on the amp(s) to minimum and turn up the volume on the headunit up until you hear audible distortion. Then you would back off a little. Next, you would do the same with the gain on the amplifier - raise it, until you hear distortion and then back off a little.

If you find that some other music that you listen to is recorded a little louder and therefore has audible clipping, then you will of course compensate by turning the gain down a little, or turning down the headunit volume a little, depending on how high it is.

If you use the tone controls (and loudness etc) on your headunit, then set them to your usual settings when setting the gains with music..

Next, you have to consider the subwoofer amplifiers.
Now it depends on if you have a separate subwoofer out on your headunit. (or if you have wired it up to get a similar effect with the fader)

Also consider that it is quite common to have more power available for the subwoofer(s) and thus the subwoofers can easily overpower the front speakers if the gains are set just before audible clipping. When the sub bass is significantly louder than your front stage, you will never have good SQ because the subwoofers will not blend in well at all.

If you do have a subwoofer level control, then you have some choices:

Since you can adjust the subwoofer level, you can actually set the gains, so that you can turn the sub bass as loud as possible (before audible clipping and amplifier overheating) for those times when you don't care about SQ, you just want a lot of bass and then still be able to lower the subwoofer level on the headunit back to a SQ level when you want.

To do this, you would set the subwoofer level on the headunit fairly high (11/12 for example), then raise the volume to the volume you set the gains for the front stage amplifier at. Then you would raise the gain on the subwoofer amplifier(s) until you hear audible distortion and then back of a bit.

If you cannot adjust the subwoofer level on your headunit, then you have less flexibility. I would suggest you set the gains on the subwoofer amplifiers so that the subwoofer level bends in with your front speakers (ie, tune for SQ), rather than setting them as loud as they can go.

Again, if you play a different recording and hear distortion, then lower your gains a little. If you are pushing your amplifiers too hard and they go into thermal protection, then again, lower the gains a bit.