@macbrett,
'These spikes only happen during clipping. The sharper the corner, the higher the spike.'
This has been a most instructive post.
Thankfully I've never had issues with clipping, at least none that I was aware of or that left permanent damage.
It might yet be interesting to hear from those that have - if the memory is not too painful - just to see how common a problem this is.
I used to be so uptight about my system and if something ever went wrong it would upset me for a while.
I don't remember if it was just immaturity (lack of perspective) or the lack of money but it clearly wasn't worth getting into a hole over.
I can imagine owners of seriously expensive loudspeakers, and amps too, might want to make sure their chances of encountering damage caused by amplifier clipping it are kept to a minimum.
|
@cd318 The reason tweeters are susceptible to damage from an amp clipping is that the distorted (flat-topped) waveform contains huge amounts of high frequency energy that are not there if the undistorted wave had been able to reach a normal peak. In particular, it’s the squared-off leading and trailing corners that are the culprits. After passing through the speaker’s crossover which blocks the low frequency (DC) flat part of the wave, these corners appear as large positive and negative spikes of energy to the tweeter
These spikes only happen during clipping. The sharper the corner, the higher the spike. Normal musical content doesn’t contain nearly as much high frequency energy as these spikes. Tweeters are just not able to dissipate this, and the fine wire of the voice coils will melt like a fuse very quickly. |
The graphs in Atkinson's measurements of amps in Stereophile are quite helpful. He pushes amps to their limit, and you can see the line heading skyward in the graph where it goes into clipping. He also specifies % of distortion allowed to measure maximum watt output, a % figure that he tends to relax with tube amps. |
@rodman99999 ,
Good article. I was still a bit puzzled why clipping particularly tends to blow out tweeters.
'Since a clipped signal contains a high number of high frequency harmonics, tweeters are especially at risk for damage. These high frequency tones not present in the original signal are directly fed to the tweeter by the speaker’s crossover. The result can be a burned voice coil, i.e. permanent damage to your system.'
|
|
@cakyol, 'It is ALWAYS safer to use a higher power amplifier on a speaker but ALWAYS MUCH MORE dangerous to use an amplifier which is under powered for a speaker, especially if the amplifier has no anti clip circuitry.' Yes, that's been a key principle of mine during the past 30+ years. My first serious amp, the now near legendary NAD 3020a had a feature they called 'soft clipping' as described in this link provided earlier by @rodman99999. http://www.thefactoryaudio.com/blog/2017/1/28/soft-sabotageI think I did eventually turn it off as suggested by the author William Crampton, but during the times it was left on you could actually hear the sound getting soft and rubbery if you turned the volume up too far past 12 O' Clock. With an adequately powerful amp and reasonably efficient speakers the only thing to watch out for is sustained high volume levels which can burn out the voice through overheating. Thankfully that only happened once when my brother got carried away by blasting out some heavy metal that burned out one of the mid/bass drivers. It was cheap enough to get a replacement and never happened again. Although the end result (driver damage) might be similar the causes are different. Clipping, I believe, is caused by continuously overdriving the amp beyond its power reserves. Overheating damage on the other hand is caused by the prolonged overdriving of the speakers. Too much power for too long. A situation that's not likely to happen with modern speakers unless you wish to invite ear damage. Both are dangerous but clipping is the more common risk of the two as even low efficiency speaker designs tend to have high power (heat) handling capabilities. @rego, Good link. It's easy to forget just how magical the whole process of converting electrical signal into room filling sound is. |
Altho what I said above is true, I want to mention also that you may be lucky and the capacitors in the speaker crossover MAY protect against DC reaching the tweeter voice coils. But I would not ever rely on that premise just in case. |
Clipping happens when you drive a transistor to a point where it wants to swing to a voltage higher than what your power supply can provide. It happens when you drive the amp with too much input signal.
It has NOTHING to do with how many watts your speaker can handle. You can literally instantly DESTROY a 500 watt speaker with a 10 watt badly designed amplifier with no clip protection, if it clips. That is because you are sending DC to the speaker and they absolutely hate that.
It is ALWAYS safer to use a higher power amplifier on a speaker but ALWAYS MUCH MORE dangerous to use an amplifier which is under powered for a speaker, especially if the amplifier has no anti clip circuitry.
|
Even if amplifier can deliver to speaker peak voltage equivalent to rated peak power only (no headroom) overdriving amplifier still delivers more power to speaker. It is because signal that resembled sinewave (Pavg = 0.5Ppeak) is changing to square wave (Pavg=Ppeak). Heat dissipated in speaker coil is equal to this average power (Pavg). In addition tweeters are not designed to handle huge power (small size), because there is no need for that, since high frequencies carry very little energy. Square wave of overdriven output has big high frequency energy (harmonic content) and is likely to overheat tweeter. Tweeter is the one that usually fails first. Many amps provide "Soft clipping". NAD was the one that I remember. Also some, if not all Icepower class D amps do that.
Yes, it is possible to created malicious recording, but you should be watching your gain settings (volume knob position) - otherwise you can play unknown music that appears to be very quiet, but then you get full symphony orchestra blast over the limit. It is also technically possible to create recording containing very high levels of 20kHz that might overheat any tweeter silently, but I wouldn’t expect the worst everywhere. |
First off Guys I am hesitant to get into this discussion because I won't be p/o of a back and forth argument. However, in the interest of clarity, I refer you to this link:
https://en.wikipedia.org/wiki/Clipping_(audio) .
It provides a pretty good definition of audio clipping. I don't agree with it entirely, as it over simplifies for non engineers, but it is essentially correct. I would only add that my professors would say something like this: now a good engineering design will prevent this from happening, or at least prevent damage to the equipment. Hope this helps ALCON.
|
The above referenced commentary by DeVore is not about a " lie " but about how manufacturers juggle terms (intentional misdirection ???) that could be misleading for important Sensitivity / Efficiency figures, otherwise useful information for easy evaluation of a product. DeVore relates to a period before " Watts is Cheap " became a driving theme. This is little more than a natural evolution / development of play back gear. @cd318, I think the video that you want is: How Speakers Work- https://www.youtube.com/watch?v=7fFFys-vv8g |
+3 what millercarbon said |
@millercarbon, Thanks. That's the best explanation of this enormously complicated business I have yet read. Only the other day my browser fed me this 'rant' from none other than John DeVore regarding misrepresented speaker sensitivity. I had to watch (listen to) it a few times before it made sense. OK, so I was at work doing something routine at the time... ------- John DeVore goes on about the Hi- End Audio lie that inspired him to start his own company. https://youtu.be/PEcFkSQMc8g |
There was an amp that provided soft clipping circuitry, NAD comes to mind. Clipping can destroy a tweeter in short order. If the amp/speaker is matched properly then it should not be an issue. Typically, its better to have an abundance of power than not enough. However, even a lower powered amp matched to a high sensitivity speaker will seldom if at all experience clipping. |
Ok people make this more complicated than it needs to be. To make it more simple clipping is your amplifier trying to put out too much power. When you do this the distortion levels reach a point that can damage your speakers. You can damage 200 watt speakers with a 50 watt amp if you push the amp too far. Typically most amps will clip when they get to about 1% distortion. Amps with more headroom can exceed their rated power a little easier without clipping.
|
|
Just grab your amplifier’s keyboard, hit the tilde key, and then type in ’noclip’, hit enter...and bob’s yer uncle...
|
My understanding is that SS and tube amps, generally speaking, tend to behave rather differently. Distortion will increase just a little until the SS amp reaches its limit, at which point distortion tends to skyrocket. Whereas tube amps will have rather more distortion gradually increasing over their standard operating range, but there is less of an "elbow" when clipping ensues. Is that about right?
It's been suggested to me that this is one reason why people say that tube watt ratings are not the same as SS watts.
|
All amplifiers have a maximum and minimum output voltage, if this is exceeded then the amplifier should output dc (a constant voltage) until the signal comes back into range. So if you imagine a sine wave with the top and bottom cut off then that is what an ideal amplifier will produce when clipping (I'm talking solid state, valve amp distortion is more benign). A good designer will ensure an amplifier will enter and exit clipping in a controlled way (without oscillation etc.) although this is not always the case.
When designing an amplifier you can specify the maximum voltage and gain but you don't know what the level of the input signal will be. So if you take a consumer amplifier and provide a high level signal (most DACs put out a signal far higher than the nominal consumer level of +-0.447V) then the amplifier is in danger of clipping.
There are ways of limiting the gain at the extremes to try to avoid clipping but these distort the signal close to maximum output. Probably the best approach is some sort of warning light a red light or VU meter however they are rarely used in modern equipment.
|
@rodman99999
Thank you for those very informative articles, they explain a lot in deed.
They ought to add more depth of information on the OP enquiry, to what already has been said.
Particularly the second link I found most informative and hope so will the OP.
M. 🇿🇦 |
|
Now how about the difference of 'soft clipping' as an amp design feature, compared to 'hard clipping'?
M. 🇿🇦
|
What millercarbon said, +1. Especially: that clipping is actually a high frequency distortion, that will be passed (by the crossover) to a system’s tweeters, first. ALSO; one can lose a tweeter to clipping, in a matter of very few seconds. While in the business of manufacturing speakers; I built in Zener diode circuits, parallel to the tweeters, that tapered off any voltages beyond the driver’s handling capabilities. About clipping: https://www.electronics-tutorials.ws/amplifier/amp_4.html |
@bwang29, what is the issue with active powered speakers that you are referring to ... a specific example ... ?
"
3. Another related confusion is, how is it possible that sometimes I see powered active speakers blown because it’s trying to play too loud? Would it be true that the amp in those active speaker should always be designed to operate within the limit of its power handling?"
Small playback systems are not intended to produce high SPL levels.
This reference is more about an issue of misuse.
|
Right you are, you completely misunderstand some concepts. A big one is thinking of an amplifier as some kind of sentient being thinking about what it should do.
What I don’t understand is why would an amp allow itself to consume more power than it could handle in the first place. Its a circuit. It doesn't "allow itself" anything. Because there is no "self". Its a circuit. It never "consumes more power than it can handle in the first place". It can't, the fuse would blow long before that could happen. What happens instead when you turn the volume up the incoming signal is able to be handled by all the circuits just fine until it gets to the output stage. Tubes or transistors, gloss over some details, basically two things happen. To understand what they are it helps to know a little about amplifiers. Basically, AC power comes in from the wall, gets converted to DC in the power supply stage, and stored in power supply caps. Right away you can see there's three things that can limit power: the incoming AC, the power supply stage (basically a rectifier that converts AC to DC) and the power supply caps. Eventually there's either output transistors, or tubes and transformers. Either way these impose their own power limitations on the signal as well. Nowadays we have these standardized power ratings that are great for leveling the field but they do have the unintentional side effect of blurring what's really going on. What you measure can actually contribute to confusing people more than helping them. Such sadly is the case here. Because music isn't a uniform sine wave, its all different frequencies and amplitudes, constantly changing. The power supply has way more than enough power stored to put out huge power for an instant. The tubes or output transformers can also put out huge power for an instant. But do it long enough and either the power supply caps will drain, or the output devices heat up and become inefficient (heat is energy, the power is going to heat instead of sound) or both. Power drops off. A lot. This is what we call clipping. The point where the power supply or the output devices, either one or both, simply cannot do any more. They are either too hot, or too drained. (Usually too hot.) Same thing happens inside the speaker. Power goes into the voice coils, a very thin wire inside a magnetic field. The more power the hotter the voice coil gets. Eventually if it gets too hot resistance increases, heating accelerates, and if this goes on long enough the insulation starts to melt and typically the coil deforms and scrapes or even melts and sort of welds or glues itself in the gap. Power that is clean spikes up very high but only briefly. Power that is clipping goes very high and stays there. Worse, if you look real close at the flat clipped parts its not really flat. Its jumping up and down at a very high frequency. Its this high power high frequency part of clipping that blows tweeters, especially if there's a lot of high frequency in the music too. This all happens the same with active or separate components. The only difference is if the speakers are active they can be designed to not fail no matter what signal goes in. Separate components offer a lot more performance but call for a little basic user knowledge. Which you now have.
|
You can prevent amplifier clipping by placing a limiter (a type of compressor) before the amplifier. Some powered loudspeakers have this in their circuitry. Regarding your question, an integrated amp is a preamp and a power amp built in a single cabinet. They face the same "problems" regarding clipping. Many classical pieces start out softly and build to very loud climaxes. It's part of the music and most listeners want to hear the wide variation in volume levels. In a well put together system with a knowledgeable listen it won't cause clipping. Clipping won't damage most loudspeakers unless it's sustained for a relatively long period of time.
|
