When to Add External Amplification to an A/V Receiver

[Tomorrow]

There are measurements showing compression with moderate clipping, which contributes to the thermal load.
So I would pick the 2000W amp and configure the Preamp/AVR to a volume limit of reference or -10.
There is no published maximum recommended power for Salons(2)'s, so they must be able to handle 2kw.

- Rich

Rich...just a clarifying note that was a surprise to me when I discovered it. And this has been discussed in these parts a bit before. The amplitude levels that you have referenced a couple of times are completely arbitrary to the AVR/Pre manufacturer and even among their models. Likewise, there is no absolute relationship between these (arbitrary) scale numbers and 'reference' volume. Some amplitude range scales are even user selectable (e.g. Yamaha Aventage AVR models).

 

[RichB]

Rich...just a clarifying note that was a surprise to me when I discovered it. And this has been discussed in these parts a bit before. The amplitude levels that you have referenced a couple of times are completely arbitrary to the AVR/Pre manufacturer and even among their models. Likelwise, there is no absolute relationship between these (arbitrary) scale numbers and 'reference' volume. Some amplitude range scales are even user selectable (e.g. Yamaha Aventage AVR models).

I assumed there was some standardization when using a negative volume scale where zero is Unity Gain.

Most AVR's and Processors support volume limits which provide a useful tool to protect your system.

- Rich

 

[Tomorrow]

I assumed there was some standardization when using a negative volume scale where zero is Unity Gain.

Most AVR's and Processors support volume limits which provide a useful tool to protect your system.

- Rich

Standardized gain staging would certainly be helpful. But for some reason, it ain't in the cards.

Yes, limiters are important...especially if they're wisely integrated in the AVR/Pre.

 

[alienmuppet]

Oh, so you worry about the "too little power will harm your speaker" myth? Don't be. It's nonsense.

I agree it is likely none-sense in the hi-fi world. I think it does apply in the pro-audio world though (like live sound) where huge amounts of power are involved. Clipping is not funny at those levels; I always ran amps that were more than powerful enough to drive my speakers. There's a lot more transients to worry about in live sound too, whereas recorded material is very sanitized, usually.... often too sanitized ;-)

 

[alienmuppet]

Now I think you're on technically shaky ground. While what you're saying is true for thermal stress, you are incorrect when it comes to signal complexity. Music consists of many frequencies being reproduced simultaneously, and loudspeakers present different electrical characteristics to amplifiers based on frequency, so a music load is electrically far more complex as seen by an amplifier driving a real-world speaker (especially a complex multi-driver speaker with a passive crossover) than just driving a frequency sweep into a test resistor. I believe this is Gene's point, and I think it is valid, and I think that Gene's FFT measurements occasionally show some interesting broad-spectrum effects.

Regarding complexity: at a single moment in time there is but 1 instantaneous voltage being sent to the speaker; there are not multiple voltages. As far as I know no part of the system cares how complex the music seems to us human beings: it is just a particular voltage at a particular point in time.

I am no expert in this field though, so if someone can explain to me what the complexity is (from a system point of view, not from a human point of view) that would be appreciated.

 

[PENG]

Regarding complexity: at a single moment in time there is but 1 instantaneous voltage being sent to the speaker; there are not multiple voltages. As far as I know no part of the system cares how complex the music seems to us human beings: it is just a particular voltage at a particular point in time.

I am no expert in this field though, so if someone can explain to me what the complexity is (from a system point of view, not from a human point of view) that would be appreciated.

There were talking about amps. You are right regardless, and I would think that no music waveform should present too much of a challenge to a well designed/built amplifier anyway. Another reminder side note, a complex waveforms really comprises infinite series of sine waves of fundamentals + harmonics.

 

[Goliath]

Now I think you're on technically shaky ground. While what you're saying is true for thermal stress, you are incorrect when it comes to signal complexity. Music consists of many frequencies being reproduced simultaneously, and loudspeakers present different electrical characteristics to amplifiers based on frequency, so a music load is electrically far more complex as seen by an amplifier driving a real-world speaker (especially a complex multi-driver speaker with a passive crossover) than just driving a frequency sweep into a test resistor. I believe this is Gene's point, and I think it is valid, and I think that Gene's FFT measurements occasionally show some interesting broad-spectrum effects.

Music is a more complex signal compared to a pure tone, but pure tones have a crest factor of 3 dB, while music can have a crest factor of between 9-12 dB on the average, up to 20 dB. This vastly reduces the average drain on power supplies and the heating of the power transistors and heat sinks serving the output stages.

As I said before, music is like driving across a rolling plain, which is rarely flat, but also never steadily uphill. A resistive load, how amplifiers are commonly bench tested, keep the impedance constant with frequency. Well, as we all know, real loudspeakers don't operate with a constant impedance. The impedance varies with frequency, and can be much higher than the nominal impedance most of the time, over a significant portion of the frequency range.

Yes, the impedance can dip below the nominal, but it is more stressful on the power supplies to keep impedance constant than to have it varying above and below at various times based on the duration and frequency content of the music.

The issue is that for some systems in some rooms with some listeners some of us, and I'm in this camp, think there are differences in amplifiers that affect our ultimate satisfaction on a system level. I think Gene has the right attitude in looking for reasons why that may be true.

Okay, then we are in agreement. I just don't agree with the nebulous claim that power amps can somehow improve the performance of speakers even if the voltage/current demands have been met with a given AVR. We both know there are people out there who make those claims, regardless of the required output levels.

 

[AcuDefTechGuy]

There were talking about amps. You are right regardless, and I would think that no music waveform should present too much of a challenge to a well designed/built amplifier anyway. Another reminder side note, a complex waveforms really comprises infinite series of sine waves of fundamentals + harmonics.

What courses in college cover this?

Was it engineering physics II?

Yeah, how does an amp know if the signal is complex waveforms?

When they test amps, they use continuous testing, which they label as "RMS"? In real life, amps don't really see RMS continuous signals like they do in testing?

So I would think that the amp testing is a lot more severe than most real life situations.

 

[AcuDefTechGuy]

it is more stressful on the power supplies to keep impedance constant than to have it varying above and below at various times based on the duration and frequency content of the music.

That makes sense. Keeping the torture test at 2 ohms constant is a lot more strenuous than having the impedance vary from 2 ohms to 20 ohms.

 

[Goliath]

I have BIG power amps in my system b/c my speakers demand them and sound BETTER with them. Not even an $11k Pass Labs amp could properly drive my speakers like the Emotiva Monoblocs especially when listening to Blu-ray audio with no compression. My situation isn't the norm. I am running my speakers full range with ALL LFE routed to them.

If not even an $11k Pass Labs can drive your speakers properly then isn't that a testament to poor loudspeaker engineering? Plenty of speakers do not need $11k of amplification to be driven well. Well, unless you feel price and performance is a linear function, but I don't agree with that.

Modern speakers don't need boatloads of power? Really, which speakers? Have you tested very large speakers that have low impedance demands at bass frequencies? What size room would you make this qualification? at what SPL? Again over generalizing...

Look, I don't think it is unreasonable to assume that typical people a) don't sit very far away from their speakers (more than 5 meters), and don't crave ridiculous SPL (105 dB+) in their homes. I'm not talking about the small subset of people who enjoy trying to recreate a live concert in their rooms, but the everyday, ordinary guy who enjoys listening to music.

You don't need bucket loads of power to handle the kinds of SPL that the average guy enjoys (75-95 dB), and the typical guy probably won't own speakers that have impedance dips down to 1 ohm with weird phase angles. Contrary to popular belief, speakers today don't all require a big stonking power amp in order to sound decent.

Now sure, I'm generalising here, but the above seems reasonable to me. Yes, you may need a bucket load of power to drive your speakers with rerouted LFE, which is not a typical scenario. It is atypical. How many people will not use a subwoofer to handle the LFE? I imagine not many. That's another generalisation, but a reasonable one to expect given the popularity of subwoofers.

If you ask me, most people today use subwoofers for handling LFE and rerouted bass information. Again, another generalisation, but am I wrong?

 

[fuzz092888]

If not even an $11k Pass Labs can drive your speakers properly then isn't that a testament to poor loudspeaker engineering? Plenty of speakers do not need $11k of amplification to be driven well. Well, unless you feel price and performance is a linear function, but I don't agree with that.

Normally, I'd be in agreement, to a point (depending on other factors), but in Gene's case the 8T's are being fed the full range signal, along with the LFE content. This means that if he's listening at reference levels, his speakers need to be able to reproduce peaks of around 115+ db's. Factor in listening distance and room size and even with a rated sensitivity of 93db/2.81v/1m + 700WPC from the Pass amp, and you're asking a lot from that amplifier. Take away the LFE channel and I think the Pass Labs amp is probably just cruising in most cases. However, with the LFE channel, I can see where having 1750 watts can have some real benefit.

 

[Irvrobinson]

Regarding complexity: at a single moment in time there is but 1 instantaneous voltage being sent to the speaker; there are not multiple voltages. As far as I know no part of the system cares how complex the music seems to us human beings: it is just a particular voltage at a particular point in time.

I am no expert in this field though, so if someone can explain to me what the complexity is (from a system point of view, not from a human point of view) that would be appreciated.

Your statement is only correct for DC signals. When music is playing an amplifier is simultaneously acting on many thousands of signals at different frequencies and phasing. Also, amplifier distortion can be dependent on frequency, power level, and how many different frequencies the amplifier is working on at once.

 

[Irvrobinson]

That makes sense. Keeping the torture test at 2 ohms constant is a lot more strenuous than having the impedance vary from 2 ohms to 20 ohms.

Only as a test of thermal issues and the power supply. As Gene was trying to explain, there are other factors.

 

[Irvrobinson]

Yeah, how does an amp know if the signal is complex waveforms?

When they test amps, they use continuous testing, which they label as "RMS"? In real life, amps don't really see RMS continuous signals like they do in testing?

So I would think that the amp testing is a lot more severe than most real life situations.

The amplifier doesn't know about simple or complex signals. The circuits in the amplifier have specific transfer functions, and depending on the signals you feed them and at what power levels you get different results. There is a valid question about whether or not the variability is audible, but amplifier circuits always have measurable distortion and noise.

 

[Irvrobinson]

Yes, the impedance can dip below the nominal, but it is more stressful on the power supplies to keep impedance constant than to have it varying above and below at various times based on the duration and frequency content of the music.

I think you're missing the entire point. There is more to amplifier performance than simply power supply stress.

Okay, then we are in agreement. I just don't agree with the nebulous claim that power amps can somehow improve the performance of speakers even if the voltage/current demands have been met with a given AVR. We both know there are people out there who make those claims, regardless of the required output levels.

You are, of course, free to disagree. I'm one of those people who suspect that amplifier's have some sort of distortion *within* their power envelop that makes them more or less pleasant to listen to. I have no definitive evidence, and I admit that, but it is nice to see the search continue.

 

[Goliath]

I think you're missing the entire point. There is more to amplifier performance than simply power supply stress.

Sure, distortion is important and with less stress on the power supplies at a given power output, you can attain lower distortion. What I do not, IME, believe factors into amplifier performance is the look and colour of the chassis, the brand, or the price. Price often has no direct correlation to performance whatsoever. Of course, in sighted listening tests, the look of the chassis, the brand and the price can have very strong influences over the perceived sound quality.

People seem to think that the amplifier needs to have zero load impedance in order for them not to impact audibly on the frequency response. As we know, just because there are measurable differences does not mean there must be audible differences.

 

[Irvrobinson]

Normally, I'd be in agreement, to a point (depending on other factors), but in Gene's case the 8T's are being fed the full range signal, along with the LFE content. This means that if he's listening at reference levels, his speakers need to be able to reproduce peaks of around 115+ db's. Factor in listening distance and room size and even with a rated sensitivity of 93db/2.81v/1m + 700WPC from the Pass amp, and you're asking a lot from that amplifier. Take away the LFE channel and I think the Pass Labs amp is probably just cruising in most cases. However, with the LFE channel, I can see where having 1750 watts can have some real benefit.

Actually, some of the loudest peaks I've measured on my own system are from horns, especially trumpets. Saxophones are pretty loud too. (The sax player in my wife's band often wears ear plugs when he plays the sax, because unlike other horns the bell is pointing up at the musician.) A 110db peak at 1KHz takes just as much power as 110db at 30Hz, assuming the speaker system has a flat frequency response, lest we forget.

 

[AcuDefTechGuy]

The amplifier doesn't know about simple or complex signals. The circuits in the amplifier have specific transfer functions, and depending on the signals you feed them and at what power levels you get different results. There is a valid question about whether or not the variability is audible, but amplifier circuits always have measurable distortion and noise.

A lot of hypotheses. At the end of the day, whatever we believe, whatever we hear, whatever makes us happy is the correction answer for us.

I am happy with my ATI amps.

 

[RichB]

Music is a more complex signal compared to a pure tone, but pure tones have a crest factor of 3 dB, while music can have a crest factor of between 9-12 dB on the average, up to 20 dB. This vastly reduces the average drain on power supplies and the heating of the power transistors and heat sinks serving the output stages.

As I said before, music is like driving across a rolling plain, which is rarely flat, but also never steadily uphill. A resistive load, how amplifiers are commonly bench tested, keep the impedance constant with frequency. Well, as we all know, real loudspeakers don't operate with a constant impedance. The impedance varies with frequency, and can be much higher than the nominal impedance most of the time, over a significant portion of the frequency range.

Yes, the impedance can dip below the nominal, but it is more stressful on the power supplies to keep impedance constant than to have it varying above and below at various times based on the duration and frequency content of the music.

Testing in a lab can be more strenuous than testing in a lab, except when it is not. For example, for speakers that dip below 4 ohms. If heat is an issue, the duration of a 4 ohm test matters as well. A few minutes may not be indicative of a 4 ohm speaker listening for an extended period of time.

Here is a real world example:
There Revel Salons are rated with a nominal impedance of 6 ohms.
The A51 amp has been measured by many sites including AH at 400 watts RMS 20-20kHz 2 channels driven.

I was able to put the A51 (on an open shelf) into thermal overload in 15 minutes driving my Revels with 2 channel musical at near reference levels (-5 on the Marantz Pure Direct, but I am uncertain of the DB level). There was no discernable clipping.
To be fair, I was doing this at the bequest of a friend that was having some trouble with midrange distortion at high volume level on his speakers. This is not a listening level that is normal for me. The album was Alison Krauss "Paper Airplane" which is has good bass content but not bass heavy).

In reality, the Revel Salons dip below 4 ohms and they have some difficult phase angles.
Damn reality.

- Rich

 

[RichB]

A lot of hypotheses. At the end of the day, whatever we believe, whatever we hear, whatever makes us happy is the correction answer for us.

I am happy with my ATI amps.

... and well you should be

- Rich