[Anonamemouse]

Oh, yes, another conclusion is that using a 4 ohm nominal speaker, it seems that one of the clear benefits would be higher dynamics and therefore, greater dynamic realism potential.

The problem with this conclusion is that a "4 Ohm" (nominal) or an "8 Ohm" (nominal) loudspeaker does not exist. The specified value only happens at a few frequencies. The image shows a graph which is VERY smoothened, in real life there are lots of peaks.

[SoundGame]

The problem with this conclusion is that a "4 Ohm" (nominal) or an "8 Ohm" (nominal) loudspeaker does not exist. The specified value only happens at a few frequencies. The image shows a graph which is VERY smoothened, in real life there are lots of peaks.

I agree that there is no absolute but a nominal rating should be arrived at in any scenario other than a purely random scenario.  The use of a mean, mode, median calculation on such a graph will help to at least make some comparisons between speakers possible.  It is true if that a given speaker has an extremely variable impendence then the nominal value will not be dependable.  I would still gamble on a speaker with a nominal impdenance rating of 4 ohms being closer to a "perfectly flat" 4 ohm rating than a speaker rated at 8 ohms or higher.  That said, it would be valuable in any given case to have an understanding of the phase of a speaker and it's overall difficulty to drive.
 
I've also been seeing more an more speakers quoted with both a nominal as well as a minimum impendence rating, which helps to make comparisons a little more meaningful.   In any case, I would think that the objective should be to strive, where all things are held equal, for a speaker with a lower nominal impendence if you are seeking more in terms of dynamic performance.
 
 

[Mike Pickett]

So, I quickly calculated the Headroom figures on the lucky 7's based on your numbers and using their rated RMS, as follows:
 
2.2 dB @ 8 ohms; 3.2 dB @ 4 ohms
 
This is considerably higher than the 4B headroom, which were calculated as follows:
 
For the 4B-SST2 they stand as the following:
 
1.6 db @ 8 ohm; 2.3 db @ 4 ohm
 
The 4B was stong, but it's clear the the 7B is a monster of an amp in terms of it's dynamic capabilities.  Any chance of getting any others to the test bench this week?

Actually, given that the 7B is a bridged amp, and each channel will see 4 Ohms when driving an 8 Ohm load, the numbers work out to only a 0.1 db difference (2.2 db vs 2.3 db).  This needs to be taken into account when comparing amps, or you will wind up with apparent differences that are not actually reflective of reality. 

[Mike Pickett]

I should also reiterate the fact that a higher 'headroom' rating is the result of more sag on the power supply when under continuous load.  If you are going under the assumption that an amp that sags more sounds better, so be it, but that's not our design goal...

So, rather than considering headroom as 'extra power', it's more useful to recognize it as 'the amount of power you lose under continuous load'.

For this reason, I'm going to use a reference of 0 db of headroom as a "perfect" amp, and any number higher than 0 as the result of real world engineering challenges.

[Mike Pickett]

Hi Mike,
 
I just got some details on how McIntosh establishes their headroom specifications.  Take a look at this (would be interested in how you feel your method compares):
 
McIntosh Dynamic Headroom Method conforms to EIA RS-490 test method http://www.ce.org/standards/cea-490-a_r-2008_preview.pdf:

·         Test signal: sine wave, 1kHz.
·         On time: 20mS
·         Off time: signal is never turned off just reduced by 20dB for 480mS.
·         Load: Amplifier rated load (8/4/2 ohms).  For McIntosh amplifiers with Output Autoformer, the rating is nearly identical for any rated load connected to the correct output terminals.
·         Distortion:  standards specify visible distortion on an oscilloscope pattern; however, the Power Guard indicator is used as the distortion analyzer, which is far more accurate than a visual opinion 
·         

Thanks for that.  I'll try to run some tests that are more easily comparable.  Not sure I'm going to get any of this done today; other duties call...

[SoundGame]

No problem Mike.  Go home early and enjoy the weekend.  No rush, though your answers would be interesting.

I also checked in with Anthem/Paradigm but it seems that going forward they will no longer be reporting peak dynamic power, as there is too much inconsistency in methods / standards.  They didn't share their former test method.  Their current belief is that the best measure is the RMS rating between 0.1 and 1% THD from 20Hz to 20kHz, into the lowest continuous operating impendance. 

[SoundGame]

Thanks for that.  I'll try to run some tests that are more easily comparable.  Not sure I'm going to get any of this done today; other duties call...

Hey Mike - any chance of getting some more testing on this done today or is your plate rather full?  It will be interesting to see how the figures end up using some of these standards.  Cheers.

[Mike Pickett]

Hey Mike - any chance of getting some more testing on this done today or is your plate rather full?  It will be interesting to see how the figures end up using some of these standards.  Cheers.

I'll try today; I'm helping out in shipping today, so that's eating up most of my spare time this week.  If not today, next week for sure...

[SoundGame]

I'll try today; I'm helping out in shipping today, so that's eating up most of my spare time this week.  If not today, next week for sure...

Not a priority, I know but of interest.  Great!

[SoundGame]

I'll try today; I'm helping out in shipping today, so that's eating up most of my spare time this week.  If not today, next week for sure...

Perhaps by the end of February...just for fun...

[klao]

"While monoblock construction is not uncommon, most manufacturers simply create a bridged version of an existing stereo product, coupling two channels in series to form a single amplifier with increased voltage and, nominally, more output power. Unfortunately, the inevitable side effects of this approach include doubling distortion, doubling input impedance (impairing the ability to drive low-impedance loads) and cutting damping factor in half (reducing control of cone motion at low frequencies). The XXXX topology offers almost the same output in watts but with double the output current, half the distortion and double the damping factor. XXXX is built for massive peak current swings into very low impedances without losing its cool."

How true is the comment above by a Danish amp manufacturere of brand/model (let's call 'em XXXX)?

[SoundGame]

I'll try today; I'm helping out in shipping today, so that's eating up most of my spare time this week.  If not today, next week for sure...

Any chance to re-test using the Mac method in March sometime?

McIntosh Dynamic Headroom Method conforms to EIA RS-490 test method http://www.ce.org/standards/cea-490-a_r-2008_preview.pdf:

·         Test signal: sine wave, 1kHz.
·         On time: 20mS
·         Off time: signal is never turned off just reduced by 20dB for 480mS.
·         Load: Amplifier rated load (8/4/2 ohms).  For McIntosh amplifiers with Output Autoformer, the rating is nearly identical for any rated load connected to the correct output terminals.
·         Distortion:  standards specify visible distortion on an oscilloscope pattern; however, the Power Guard indicator is used as the distortion analyzer, which is far more accurate than a visual opinion 

[SoundGame]

Hi Mike,

Just came across a spec. on NuForce amps.  They too quote peak power output, and like McIntosh, mention a 20ms duration for the peak level.  Captured from a SoundStage review:

"NuForce also claims that the Reference 18 puts out 175W RMS into 8 ohms, or 335W into 4 or 2 ohms, with respective peak power ratings of 325W, 650W, and 1300W (with a hold time of 20 milliseconds)."