Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Antone Atmarama Bajor]

Harmonic Compression is good.  How well does a horn loaded subwoofer cope with compressed High frequency energy that it isn't designed to reproduce?

   I don't know I've only seem a couple of Amps that claim to be able to provide short term power bursts above their RMS rating (proton comes to mind).  What do they do, set the amps clip light comparators to trigger 20V bellow the power rails?  Do the amplifiers have a magic voltage step-up up feature?  I guess the amps can have an under engineered powersupply that'll supply a nice juicy peak before Sagging and getting some good ripple going on.

    I thought the RMS rating usually was set with the peaks of your signal just a scoonch bellow the rails (-transistor overhead).  Which would mean on most amps the Peak Maximum Power output might be a tad bit more than 3dB hotter or (Vrms*sqr(2))^2/R=Ppeak or to make its easy P*2.

    So the maximum power out is a little less than your voltage rails.  How many amps give a butt load of (RMS) headroom before hitting the rail.  I am not familiar with amp designers doing this but unfortunately the tech industry is dying its last gasps where I live so I don't get to learn more about these things as I can't find work.

    I may be somewhat ignorant to amp design theory as I am an unemployed E.Tech. with only a few months of real world experience.  I have heard that some tube amps have a little headroom as they reach Hard saturation, but in solid state soft or light saturation isn't a reality.  I know with newer faster devices being used that the amps can recover from clipping faster without it being as obnoxious sounding.

I'm here to poke and prod, because many of you here actually have the knowledge and experience I lack.  And if I don't understand It I have to question it until I do.

Do periodic DC pulses (square waves) look like a dc pulses through and inductive voice coil (Xl).  Probably not?

Sum      

[John Roberts {JR}]

If you want to learn perhaps you need to read and listen a little more.

Few modern SR power amps have significant "dynamic headroom" (there used to be a specification for it primarily used with audiophile amps). These days and in this business, the difference between rated (sine wave) power and momentary peaks (sine wave tone-bursts) will usually be a dB or less depending on how you set up to test.

FWIW, I did design a power amp in the late 80's (PMA 70+) that used a variation on a cap doubler in the power supply rails so it could momentarily put out over 100W clean tone bursts from a nominal 35W amp. I used this topology in a small studio amp. While I had visions of taking this to higher power levels the technology didn't scale up economically. For SR dynamics modern power amps are fairly well matched to loudspeakers for peak and average power demands when properly sized.

JR

[Antone Atmarama Bajor]

     That sounds like a cool amp setup.  My dad just told me that  sound craft made an amp with a secondary supply for handling 24 dB peaks above its rated RMS.  Doesn't using a voltage doubler limit the actual available current?  Well I guess one could get a really quick peak before it drained its cap banks and thats the point of doing it that way.

    Too bad its not cost effective.

Ok I still want to know though does the increasing impedance of a voice coil to higher frequencies make the presence of "compressed" harmonic content make it a non issue (talking about single driver no passive crossover full range systems)?  The fundamental is not ever going to get much louder than the amps RMS output.  The Harmonics generated by the clipped waveform can increase in amplitude until they rail out.

    I'm not sure how much the inductance of the coil and mechanical resistance to reproduce the high frequency components of the clipped signal to say a Bassmaxx sub effect the actual spectral energy that the coil can convert into mechanical energy and how much of it becomes heat.

    Isn't the Fundamental effectively hard limited to just a scoonch above the rms output level of the amp?

    The driver probably can't track a clipped/squared waveform very effectively?  Which would be roughly 3dB more power dense than the un-clipped waveform?

    Or is the coil inductance and mechanical resistance low enough to allow periodic DC pulses to pass, inductors resist the change in current flow, and aren't to good at allowing instantaneous voltage changes, though it maybe in the RF ranges in this case?

    The spectral density of the clipped waveform is more power dense right?  The Fundamental of the clipped signal is hard limited to about the RMS rated output so its not the fundamental that kills the driver as it is being reproduced at the level that the driver is intended to handle. This is my logic on why I think the Harmonic content kills drivers I don't know if you can really think about it as periodic DC pulses in an AC universe as instantaneous voltage change is more theory than reality.

     But I could be bass ackwards.

I haven't heard a counter argument to this yet.

Just telling me that harmonics from clipped amplifiers don't "blow up" subwoofers isn't that compelling without some sort of explanation and elaboration.  If you have a good link I'd love to read it.

Chucks article talks about running amps whos RMS output is 2X the drivers rated RMS.  And how clipping is bad for drivers.

It doesn't the discredit harmonic content damaging the drivers theory.

I understand that mechanically abusing the driver can damage it too (like over excurding the cone trying to reproduce a low frequency with enough power to dislodge or melt the coil, and tear the spider etc..  Trying to accelerate the driver instantaneously to extremes can be thought of as trying to reproduce higher frequency high amplitude energy, which is beyond the drivers electrical and or mechanical limits can't it?

Am I just way off the mark?  I don't know you guys are the designers and engineers.  I'm just a green tech.  I know just enough to be confused.

Sum


 


   

[John Roberts {JR}]

sumsound wrote on Sun, 24 April 2005 15:42

That sounds like a cool amp setup.  My dad just told me that  sound craft made an amp with a secondary supply for handling 24 dB peaks above its rated RMS.  Doesn't using a voltage doubler limit the actual available current?  Well I guess one could get a really quick peak before it drained its cap banks and thats the point of doing it that way.     Too bad its not cost effective. Ok I still want to know though does the increasing impedance of a voice coil to higher frequencies make the presence of "compressed" harmonic content make it a non issue (talking about single driver no passive crossover full range systems)?  The fundamental is not ever going to get much louder than the amps RMS output.  The Harmonics generated by the clipped waveform can increase in amplitude until they rail out.     I'm not sure how much the inductance of the coil and mechanical resistance to reproduce the high frequency components of the clipped signal to say a Bassmaxx sub effect the actual spectral energy that the coil can convert into mechanical energy and how much of it becomes heat.     Isn't the Fundamental effectively hard limited to just a scoonch above the rms output level of the amp?     The driver probably can't track a clipped/squared waveform very effectively?  Which would be roughly 3dB more power dense than the un-clipped waveform?     Or is the coil inductance and mechanical resistance low enough to allow periodic DC pulses to pass, inductors resist the change in current flow, and aren't to good at allowing instantaneous voltage changes, though it maybe in the RF ranges in this case?     The spectral density of the clipped waveform is more power dense right?  The Fundamental of the clipped signal is hard limited to about the RMS rated output so its not the fundamental that kills the driver as it is being reproduced at the level that the driver is intended to handle. This is my logic on why I think the Harmonic content kills drivers I don't know if you can really think about it as periodic DC pulses in an AC universe as instantaneous voltage change is more theory than reality.      But I could be bass ackwards. I haven't heard a counter argument to this yet. Just telling me that harmonics from clipped amplifiers don't "blow up" subwoofers isn't that compelling without some sort of explanation and elaboration.  If you have a good link I'd love to read it. Chucks article talks about running amps whos RMS output is 2X the drivers rated RMS.  And how clipping is bad for drivers. It doesn't the discredit harmonic content damaging the drivers theory. I understand that mechanically abusing the driver can damage it too (like over excurding the cone trying to reproduce a low frequency with enough power to dislodge or melt the coil, and tear the spider etc..  Trying to accelerate the driver instantaneously to extremes can be thought of as trying to reproduce higher frequency high amplitude energy, which is beyond the drivers electrical and or mechanical limits can't it? Am I just way off the mark?  I don't know you guys are the designers and engineers.  I'm just a green tech.  I know just enough to be confused. Sum

I can't satisfy your request for immediate hand-holding explanations. This topic has been discussed and abused at much length here. It has become a mild irritation to me. I don't know what the state of the search capability is these days but if you try and it is functional you should find numerous past threads on that exact subject with thorough answers.

Regarding, my specific design, while I guess all engineering decisions can be characterised by cost effectiveness, what I said was that it didn't scale effectively. For any who care, the reason why is because the output stage must be able to handle the full max current needed to deliver even momentary peaks. The technology of the time (bipolar power transistors) required paralleling several devices to get up to serious power. Like wise the active devices in the boost circuit also had to handle the same peak current requiring a doubling of the output stage. You end up with a component count/cost/complexity similar to a class G/H design with lower continuous power. For up to a couple hundred watts my approach kicked ass, above that is was a science fair project (interesting but not very sell-able). Perhaps ironically it might better match the peak/continuous power capabilities of real speakers but I won't confuse this discussion with that just now.

Regarding your father's citation of an amp with 24dB headroom "above" rated RMS, I think I see where you get your charms  .

You mentioned in one of your other posts that you were a tech or something like that... so let me give you a little homework task. Figure out how much more power "24 dB peaks above its rated RMS" would be for say a 100W amp. Take your time, and lets know what you have an answer.

Sorry, I won't address your entire post, it gives me a headache. Maybe somebody else here more generous than I will help you.

JR

[Antone Atmarama Bajor]

10^(2.4) * 100 = 25,118.9 Watts.  I don't have any Idea of the reality of those figures he quoted me I've never heard of such a thing until a couple of hours ago.  He probably misread something or its an awfully low RMS amp like 1-10 watts.  That would be a pretty spiffy amp though.  I would call it the speaker Annihilator!  Or maybe its just a power supply for a rail gun?

Oh well I'll search for more.

Sum

    Hmm well my dads kinda a stoner and he kinda blurbed something about 24dB of headroom after talking about the Sound Craft amp.  Apparently it was in context to standard headroom above 0dBm for old 600ohm line level consoles and had nothing to do with the amplifiers headroom.  He kinda was mumbling and hard to follow.  Kinda a silly out of context interpretation of what he was trying to say.  I wasn't thinking that 24dB is over 200 times as much power.    

That would be one heck of an instantaneous peak for a system to deliver.  Lets see That only take 448Volts Why not lets build the thing I'll order a three phase 480 distro for my house tomorrow.    
We can use our drivers as projectiles for target practice.  

[Too Tall (Curtis H. List)]

[quote title=Craig Leerman wrote on Wed, 13 April 2005 22:34]

Quote:

snip-  I suggested to him that he should back off his level, and that running into the red was causing distortion.  He CORRECTED me and informed me that you don't get "All the signal"  if you don't run the meters into the red!

Well ya know he's right. He paid for all those red lights and he wants to see them working.

Then there are some mixers where the red clip lights are also known as "signal presence lights".


It isn't just DJs either.
I don't know how many times I have looked at the mixer for a concert where channel 1 (the kick drum in most cases) red clip light is bright enough to read the newspaper on every hit. You ask the guy and he says he needs to run it that hot to get "The Sound" he needs.
   This is not on a dj mixer at a local bar. This is on a $60k to $80k plus 50-channel console. Now many boards will light channel clip many dB below actual clipping, but when they are bright enough to use as a camera flash we might be into clipping, yes?

[Al Limberg]

I've always felt manufacturers could save a couple pennies on DJ products by simply equipping them with combination clip/pilot lights.

?;o)
Al

[Antone Atmarama Bajor]

     Ok so the Harmonic content is not the culprit by it self, but rather the Fundamental frequencies summed with the harmonics which increase the Average power/voltage by around 3dB, depending on how hard the amp is running into clipping.  So a 250Watt RMS speaker powered By a 250Watt RMS hard clipped amp is actually trying to Dissipate about 500Watts average or continuous worth of Heat (50% duty cycle?)?

    Not to mention any other Mechanical abuse the driver may be experiencing.  If you could magically remove the fundamental there wouldn't be enough average power to cause a problem with overheating?

    Is that a little more accurate description of whats going on (at an over simplified level).

Sum

[John Roberts {JR}]

sumsound wrote on Tue, 26 April 2005 12:37

Ok so the Harmonic content is not the culprit by it self, but rather the Fundamental frequencies summed with the harmonics which increase the Average power/voltage by around 3dB, depending on how hard the amp is running into clipping.  So a 250Watt RMS speaker powered By a 250Watt RMS hard clipped amp is actually trying to Dissipate about 500Watts average or continuous worth of Heat (50% duty cycle?)?     Not to mention any other Mechanical abuse the driver may be experiencing.  If you could magically remove the fundamental there wouldn't be enough average power to cause a problem with overheating?     Is that a little more accurate description of whats going on (at an over simplified level). Sum

You're still making this more complicated than it really is.

Forget about duty cycle and harmonics for a moment... imagine that you have a bigger amplifier. As you continue to turn up the gain that resulted in clipping with the smaller amp, you were turning up the level of the entire signal... more level = more power = more heat.

For typical music waveforms clipping scrapes off the tops of some transients but doesn't dramatically increase or reduce the total energy content (the hifi tweeter exception has been noted but doesn't really apply to pro applications). The speaker and many sound men don't know the difference between clean and moderately clipped waveforms. The speaker does notice the higher average power. The more you turn it up, the hotter it gets. The clipping is incidental.

The same speaker blown up by a clipped waveform from a smaller amplifier would still blow up had that same level signal been cleanly passed by a bigger amplifier. The concept that smaller amplifiers blow up speakers faster or easier is a myth.

Another myth is that with a properly sized amplifier you don't have to worry about blowing up speakers. The nature of music's crest factor (peak to average ratio) and loudspeaker's short term physical (excursion) vs. longer term (thermal) limits, suggest that mating a speaker with an amplifier that allows you to exploit it's peak capability exposes it to thermal damage. The system operator must manage this as sizing the amp so small you can't overheat gives up too much potential performance.

Clip limiters can be helpful to inexperienced operators not just because they prevent clipping (which they do) but in the process they keep the average levels lower.

JR  

[Tim Padrick]

Why do you need headroom everywhere in the system?  Tom Danley explains:

"If one had a bank of oscillators set to 10 different frequencies set at random and the outputs were all set at 1 Volt each and added together, there would be a time when the phases of each oscillator were the same and the voltage would momentarily reach 10 Volts and this means peaks much larger than any individual signal."


If for the sake of argument we call that 1 Volt O on our PFL, 10 Volts would be +20.

The program material in an individual channel may have peaks of 10dB above the average value.  It may be possible that at occasional moments these peaks may also sum.  This means that if you PFL your channels to 0, the peaks may be +10, and 10 channels' peaks could sum to hit +40.  In such a case, the summing busses had better be capable of handling a LOT of signal.



Danley on Multi-amping:

"For a power amplifier, the broader the signal bandwidth, the greater the signal capacity has to be to accommodate all the additions that happen.  Switching to a Bi amp situation with the same relative power will result in an increase in loudness (assuming everything else is up to the task) partly because now the power amplifier can carry in effect a bigger signal."



Danley on Speakers:

"...in a loudspeaker system, the narrower the BW a given driver handles, the greater the usable loudness from each device."  (Owing to fewer additions.)