1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  
    TalkBass.com has been uniting the low end since 1998.  Join us! :)

Newb question - please explain "clipping"

Discussion in 'Amps and Cabs [BG]' started by Matthew Bryson, Aug 18, 2005.

  1. Matthew Bryson

    Matthew Bryson Guest

    Jul 30, 2001
    Okay, I've only been playing about five years and I played an 80 watt Peavey TKO combo the entire time. There was no "clip" or "overload" lights or anything like that. I played somebody else head & cab a couple of times and it had a "clip" light - one time it went fine, the other time I had the clip light coming on constantly - I really don't understand what that is. I always believed that I'd figure all that out when the time came - but I recently started playing a Trace Elliot head that has an "overload" light (I'm thinking that's the same as "clip"?) So…. What is that? Is the fact that I have not seen the light come on no matter how I set the controls on the amp a good thing, or does it maybe indicate that the light doesn't work? Any info is appreciated.
  2. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    Clipping is what happens when an amplifier tries to put out more voltage than its power supply will allow. This can happen in the power amp, or in the pre-amp, or in any of the various amplifying stages therein, and it transforms a clean undistorted signal into a rough, compressed and distorted one. A clip indicator could be revealing a clipped signal just about anywhere in the amp, it all depends on where the designer chose to put it. Overload is usually a different term for the same thing, although it might also indicate excess amperage draw. In most cases bass players will find clipping to be undesireable, but sometimes not, and guitar players can't live without it. Distortion pedals are devices that intentionally cause clipping.
  3. Distortion occurs when the output signal of an amplifier does not faithfully reflect the input signal. Clipping is a name we give to a certain type of distortion that occurs when the demands on the amp exceed it's capabilities.

    Clipping causes a flattened voltage waveform. Flat waveforms can be considered as DC (direct current) and DC is really bad for speakers. Extreme clipping can cause high speed switching DC to be fed to the speakers, which destroys voice coils in short order.

    The clip light on an amp comes on for various reasons. Firstly you may have the input section running too hard. Turn down the input gain, switch in the gain pad, or in extreme cases you may have to turn down your bass. Other causes can be too much EQ boost on the amp, improperly matched speaker impedance or a fault in the amp.

    When set correctly, the clip light is allowed to flash briefly occasionally.

  4. Which is considered AC. There is no such thing as alternating/switching 'DC.' DC is non-varying by nature. Pure square waves do have equivalent heating power to DC but in a speaker system this will not cause the same amount of heating as true DC because cone motion is a primary source of cooling, and cone motion doesn't stop under square wave input.

    Also, in real amplifiers, clipping never gets to the true square wave level. What you get is sine (or whatever the input signal is) waves with flattened tops. The parts of the waveforms below the rail voltage will remain normal...

  5. Good point. In a recent thread on PA system power, there was a statement to the effect that "square wave forms destroy speakers". I've put square wave forms into a speaker, intentionally with a square wave sig gen, and guess what? The speaker didn't react any appreciably different than it reacted with a sine wave--I believe this has been stated before by people far more knowledgeable than I.

    Speakers die from overexcursion or overpowering. If the amp has been driven into clipping so far that the amp's output power exceeds the speaker's capacity, then the speaker burns up. Just because the waveform is "clipped off" isn't necessarily bad for the speaker.
  6. Absolutely. If square waves destroyed speakers, guitar amps couldn't function. Square waves merely have more power than an equivalent sine wave with the same voltage, which also explains why amps put out more than their rated power when they're clipping. They're not putting out more voltage, its the same voltage, but just with a waveform that spends more time at that same voltage compared to the sine wave.

  7. Geoff St. Germaine

    Geoff St. Germaine Commercial User

    Here is what happens.

  8. I third this ... or, er, fourth it ... or fifth it, whatever.;)
  9. Yup, that's why they call it clipping. Its like you "clip" off the tops of the waveform with scissors. Look at how the clipped waveform spends more time at max voltage, instead of the gentle arc of the clean waveform. So obviously the avg voltage is higher, therefore the avg power is higher for the clipped waveform, even though they both have the same max voltage.

  10. ehhhh. I was simply trying to describe the phenomenon briefly. It's a matter of interpretation - so some clarification is in order.

    A severely clipped waveform will effectively approximate switching between two DC sources at high speed. The transitions from peak to valley (and vice versa) of an extremely clipped signal will exceed the speed that which the voice coil is capable of responding to.

    Say you're in London, and a friend calls you and asks you to visit him in Hong Kong for a cup of coffee. Just as you take a step to hail a cab to go to the airport, your other friend appears behind you and asks you to stay. As he finishes his sentance, your Hong Kong buddy calls again... repeat ad infinitum and you're standing in the door of the cab with a perplexed look on your face!

    Even with a technically alternating signal, the voice coil will simply tremble in place, rather than moving from one extreme to the other. When it gets to this point, you're effectively asking the speaker coil to be in two places virtually at the same time - which it simply cannot do.

    No movement, no cooling, but an effective DC offset being applied to the speaker. Perhaps this only happens intermittently for short periods during transients, but if it happens often enough the voice coil will overheat and fail.

    An overheated cone creates friction, which becomes a primary source of heating - entirely independent of any equivalent heating power of an electrical signal.

    Just to clarify, I never said anything about square waves. I agree that square waves alone won't destroy speakers.

    "real" amplifiers? If the input signal is of sufficient amplitude, ANY amplifier will produce heavily clipped waveforms. If the first 1 degree of the sine wave is the only part of the waveform below the rails, then it's gonna be for all intents and purposes a high duty cycle square wave.

    If what you say is true, then what DOES damage speakers?

    Now, "switching DC" seemed a contradiction to me at first too. However, think less about the text book definition of DC and AC and more about it in a descriptive sence.

    Cheap DC Inverters switch multiple levels of DC in a sequence so as to resemble an AC waveform. They sure don't create sine waves, but their "AC" outputs are derived from DC voltages.

    Square waves generators are one thing, they usually have a 50% duty cycle. Clipping doesn't create square waves, it creates clipped sine waves which do not have a fixed duty cycle. Try hooking up a signal generator that can create a "square wave" with a 90% duty cycle - that's closer to what a severly clipped waveform looks like.

    If an amps ouput power exceeds the capability of the speaker, then even a completely pure sine wave will destroy the speaker from over-excursion - just like you said.

    However, it's possible that sustained clipping produced by a 100W amplifier can destroy a 500W rated speaker by overheating.

    A static voltage you say... hmm, isn't that DC?!?! If only for a very short time period!?

    See look, we all agree, it's just semantics! And Friday is here :hyper: so lets not sweat about it any more! Lets just all go out and JAM!!!
  11. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    True, but it's not woofers that clipped signals destroy, it's mids and tweeters, and it's got very little to do with the wave shape per se and everything to do with signal compression, which goes hand in hand with clipping. When you clip you not only chop off the top of the waveform, which is mostly fundamentals and low order harmonics, but you also increase the relative levels of the high-order harmonics. A 100 watt signal normally might contain 5 watts of signal content above 5 kHz. Heavy clipping and the associated compression might quadruple the HF content to 20 watts. The additional 15 watts of artificial highs isn't going to bother a 100 watt woofer, but it will quickly toast a tweeter than can only handle 10 watts. It's reason #2 why guitar amps don't have tweeters.
  12. Bob Lee (QSC)

    Bob Lee (QSC) In case you missed it, I work for QSC Audio! Gold Supporting Member Commercial User

    Jul 3, 2001
    Costa Mesa, Calif.
    Technical Communications Developer, QSC Audio
    No. The power put out by the amp acts as an accelerative force on the voice coil and speaker cone.

    Uh, no. It does not work like this.

    Whether overheated or not, speaker cones, like all moving parts, have friction.

    The worst case you'll get with a 100-watt amp, clipping so badly that it's putting out square waves, is 200 watts--and that's if the supply rails don't sag from the higher current draw. Will 200 watts burn out a 500-watt loudspeaker? Only if it's not really a 500-watt speaker.

    You don't need a flat waveform to measure a short portion of an AC voltage and find that it hasn't changed direction yet. That still doesn't make it DC.