The never ending amp journey

[fuzz092888]

I think that's a great choice. If you're going to add a center channel, I would think the best match would be another Phil 3. It would have the same build quality and be perfectly timbre matched to the L/R. But, that can come later.

Now, about that amp....

Good luck Fuzz!

True another Phil 3 would be timbre matched (and would probably sound awesome as a center) but in the Philharmonic thread somewhere around here or AVS, I have trouble keeping straight where I end up hear my information sometimes, Dennis has been designing book shelf and center speakers to go along with his floor standing speakers. I believe his centers use the fountek ribbons and they go for 650, but he'll do the RAAL ribbons for those who so wish for 850ish. Less expensive than another Phil 3, but with the same quality, but hey now there I go again getting ahead of myself

 

[GranteedEV]

Go for the third Philharmonic if you can... way worth it over an mtm center.

 

[fuzz092888]

Go for the third Philharmonic if you can... way worth it over an mtm center.

I would love nothing more than to do such a thing, I wonder if Dennis would let me order a single later on?

Also my significant other may kill me if I showed up at home one day and had three of them.

 

[DS-21]

So....having a 500wpc amp is a good thing just in case you need it.

Well, I'm not really in that camp. I don't object to higher power, IF it doesn't come with degraded low-level linearity and a higher noise floor. But for most it's unnecessary. Not only that, in multichannel systems the limit is really from the wall. Consider that a 15A circuit is limited to about 1800W at 100% efficiency, though yes one can burst a lot more than that short-term before tripping the breakers. Still, that's 3 mains, perhaps 2-8 other channels (surrounds, surround backs, wides/heights), and multiple subwoofers.

For most people, a 30-60W/8Ω chip amp per mains channel (Class AB or Class D) is probably the optimal solution. That size amp will give the better theoretical low-level performance in terms of zero-crossing distortion and the like than bigger amps (though audibility is debatable), low noise, and so on, and still provide plenty of power to drive most loudspeakers to reference levels in most rooms. People who want things louder should IMO get more efficient loudspeakers, and save the wall current for their sub amps.

But.....having a 7V pre-out isn't necessarily a good thing?

Having excess capability, assuming no compromises to actually-used capabilities, isn't a priori a bad thing. If it results in a compromise in something that actually matters, it is.

Going back to the 1V pre-out AVR. A person could easily increase the Speaker Channel Level to boost the pre-out voltage to 2V o 3V if needed?

I suppose, if the system was incompetently designed, or unreasonably used. With competent design and reasonable use, no.

The only place preout voltage really matters is in systems designed to play reference or louder (115dB in the LFE channel) well below 20Hz. Let's take Audioholics' esteemed subwoofer tester Josh Ricci's system. His references system uses extremely long-throw 18" woofers in closed boxes. To get a combined peak output of ~115dB at 10Hz, each sub needs to reach ~103dB at 10Hz, which shockingly enough is not beyond their capabilities. To do so requires a staggering 38dBW of gain, or about 6kW. Per subwoofer! Josh knows what he's doing, so he powers them with a Powersoft K10, which is capable of 6kW bursts. But to get there, even at its highest gain setting, the Powersoft K10 needs 2.27V input. So in that case, a higher preamp voltage is required.

 

[AcuDefTechGuy]

I would love nothing more than to do such a thing, I wonder if Dennis would let me order a single later on?

I see why not.

Would be so awesome with 3 Phil3.

 

[DS-21]

Dude, read carefully the text on your own link. OSHA, does not have any authority over independent testing labs

I have written absolutely nothing that is inconsistent with the link I posted.

What part of "[t]he Program recognizes private sector organizations as NRTLs, and recognition signifies that an organization has met the necessary qualifications specified in the regulations for the Program." do you fail to understand as "authority" over the labs? If they can certify/decertify labs as designated NRTL's, then to any reasonable person they have "authority" over the labs.

As for the standards, I never claimed that OSHA developed them. OSHA's recognition of the standards ("consensus-based standards of safety") is sufficient to make everything I wrote entirely correct.

I think you'll find many manufacturers of amplifiers do not have any safety certification and that their quality is superior to that of many that do... Just sayin.

Doubtful. The better amp makers (Bryston, McIntosh, Crown, QSC, etc.) all carry NRTL labels. Here's, I think, an interesting quote from someone who knows more than anyone in this thread (or most other places) about audio electronics:

*** the [Sherwood A-965] will pass FTC with seven channels driven into a 4-ohm load. The power into 4 ohms is 160 watts. That is just about the maximum power that can be expected from a seven-channel amplifier designed to be meet the UL- developed safety standards.***the [McIntosh] MC207 can produce 200 wpc continuous into 8 ohm continuous [sic] with all 7 channels driven but you need to throttle the DC voltage rails back (it has a switch in the rear) for 4 ohms continuous performance, which is also 200 wpc. Without that switch the MC207 would likely not be NRTL-certified for a 120V line.

David Rich, "Sherwood Newcastle A 965 7-channel power amp," The $ensible Sound, issue 109. Available at http://www.sherwoodusa.com/pdf/A965_from_TSS_review.pdf

The unwillingness to confirm basic electrical safety through independent testing is a big red flag to me. And, if the thing catches fire, can result in denied insurance claims. Such a stupid thing will never be found in my own home, and in my rental properties I have a clause specifically forbidding tenants from using non-NRTL certified electronics, that I will fully enforce should the situation come up. One should take reasonable steps to protect his own property, after all.

 

[AcuDefTechGuy]

...6kW bursts. But to get there, even at its highest gain setting, the Powersoft K10 needs 2.27V input. So in that case, a higher preamp voltage is required.

So even in this extreme case, the preamp only needs to have 2.27V?

Which means more than this is most likely unnecessary, which is what Denon electrical engineers probably think?

 

[DS-21]

So even in this extreme case, the preamp only needs to have 2.27V?

Well...keep in mind he's using a couple of 'em. So the preamp needs to be able to drive four 2.27V amp inputs.

Also, the extra voltage of something like the Marantz pre-pro would not be wasted in that case. For long runs, or a very noisy room, or maybe just because it works better for whatever reason, one may prefer to run a hotter preamp signal and lower amp gain. From Powersoft's website:

-----------------
AUDIO
Gain (selectable via user menu)
26 dB - 29 dB - 32 dB - 35 dB

Input sensitivity
6.39V/18.3dBu - 4.52V/15.3dBu - 3.20V/12.3 dBu - 2.27V/9.3dBu
-----------------

Also, input attenuation is available at each gain setting. Again, consumer amps seem rarely to have input attenuation/gain controls. That is IMO another sign of the idiot dogma that has infested so-called high end audio.

 

[Irvrobinson]

Maybe closer to 20-75, depending on how large your room is, where the peaks lie in the spectrum, how many speakers were playing, and whether or not the signal going to them was in phase. Remember that (Klipsch and a few others excepted) sensitivity ratings are taken from anechoic chambers.

Unless you heard "don't hurt me" type noises emanating from the amps themselves (it happens - try, for instance, a loud 5-200Hz swept sine on an 82dB/W/m subwoofer powered by a Dayton 240W plate amp) the amps did not in fact sound strained. The loudspeakers did. That is a very non-trivial distinction.

My room has a rather large volume and there were two speakers playing. The sub was working, but this was music so the frequencies handled by the sub were not the loudest. Also, since my speakers are about 12 feet apart the additive loudness from the second speaker will not be a big factor with the mic at one meter from the speaker being measured.

As for your assertion that my 125w/ch amps did not in fact sound strained, I'm not convinced. I would argue that if I use a more power amp and it sounds better at a higher average and peak listening level, the only likely explanation is that the greater headroom of the higher powered amp results in lower distortion. Your argument is that a low-powered amp is probably good enough, and the rest is all in my head. I tried that, out of curiosity, using my much-lower-powered Sony receiver in place of the high-power power amps, and it was e-a-s-y to tell the difference. I think you've just made up a theory that you think power requirements are really lower than others think, and it's a (bad) guess.

I've also come to the conclusion, DS-21, that you're just an inverse audiophile, so to speak. You make all sorts of assertions about what doesn't matter in audio systems, while audiophiles just make all sorts of assertions about what does matter. [Big Grin] Sometimes, you're right and they're wrong (like with speaker and power cables, I'd presume), but power levels have a good chance of actually making a difference in loud listening levels in large rooms.

 

[DS-21]

As for your assertion that my 125w/ch amps did not in fact sound strained, I'm not convinced.

Please re-read the section you quoted. For reference, here it is, with a different emphasis:

"Unless you heard "don't hurt me" type noises emanating from the amps themselves (it happens - try, for instance, a loud 5-200Hz swept sine on an 82dB/W/m subwoofer powered by a Dayton 240W plate amp) the amps did not in fact sound strained. The loudspeakers did."


Note I never wrote you didn't hear strain. Any reasonable interpretation of my above statement is in fact the opposite: an acknowledgement that you heard something sound strained. I just wrote that you misapplied where it was coming from. (Unless you were actually talking about noises physically coming from the amp, in which case you would have been correct in your attribution.)

Now, could a higher-powered amp have less to less audible strain from the loudspeakers? Yes. However, the sound would still be coming from the loudspeakers and not the amp. Words matter.

I would argue that if I use a more power amp and it sounds better at a higher average and peak listening level, the only likely explanation is that the greater headroom of the higher powered amp results in lower distortion.

There are plenty of other explanations, but yes, that is one reasonable one.

I've also come to the conclusion, DS-21, that you're just an inverse audiophile, so to speak. You make all sorts of assertions about what doesn't matter in audio systems, while audiophiles just make all sorts of assertions about what does matter. [Big Grin]

If your "inverse audiophile" is "music lover," then that is correct.

However, I've made plenty of comments (albeit not assertions because I actually have and post support for the comments I make - again, words matter) about what does matter.

For example, constant midrange directivity and multiple subwoofers used expressly to smooth out the frequency response in the room's modal region: any system without those two traits is by definition low fidelity, because those two characteristics are necessary - not sufficient, but necessary, because, again, words matter - for high-fidelity music reproduction in a small room. ("Small room" is for these purposes a term of art defined more-or-less as "something one would find in a domestic dwelling," to contrast with a "large" room such as an auditorium or concert hall designed to support an orchestra.)

***power levels have a good chance of actually making a difference in loud listening levels in large rooms.

I don't disagree, though as a general matter if one's listening levels require high power, and one desires album-length listening sessions or longer, one should move to higher efficiency loudspeakers. Then one will realize less thermal compression at loud levels, and lower energy costs at all listening levels.

As to actual levels required, I'm curious, have you done the test on the diyaudio forum to determine how much power you actually use? (FWIW, my comments on the limits of the test are here)

The results are quite illuminating. It did change the way I thought about amplifier power a bit. Prior to doing the test, I was very considering dropping 4 grand on a Powersoft Q4004, a rather special 4-channel amp: only 1RU high, yet capable of providing 600W to each of my mains at their nominal 8Ω impedance and 1kW to my high-mounted sub at its 4Ω nominal impedance. And yes, it's certified as safe by an OHSA-approved lab!

But after actually doing the math I realized the only reason I'd want that amp would just be to have a cool, high-tech multi-kilowatt Italian superamp powering my mains. Sonically, it wouldn't give me anything that the built-in amps in the Anthem receiver can't give me. So instead I kept two grand for other things and gave two grand to charity. Obviously, under few circumstances would it be sonically different from two Crown XLS2500's (450W/8Ω, 750W/4Ω I think) which would also be a lot cheaper. But the Powersoft is marginally if insignificantly more powerful, and takes up 1/4 the vertical space.

 

[Irvrobinson]

As to actual levels required, I'm curious, have you done the test on the diyaudio forum to determine how much power you actually use? (FWIW, my comments on the limits of the test are here)

The results are quite illuminating. It did change the way I thought about amplifier power a bit. Prior to doing the test, I was very considering dropping 4 grand on a Powersoft Q4004, a rather special 4-channel amp: only 1RU high, yet capable of providing 600W to each of my mains at their nominal 8Ω impedance and 1kW to my high-mounted sub at its 4Ω nominal impedance. And yes, it's certified as safe by an OHSA-approved lab!

But after actually doing the math I realized the only reason I'd want that amp would just be to have a cool, high-tech multi-kilowatt Italian superamp powering my mains. Sonically, it wouldn't give me anything that the built-in amps in the Anthem receiver can't give me. So instead I kept two grand for other things and gave two grand to charity. Obviously, under few circumstances would it be sonically different from two Crown XLS2500's (450W/8Ω, 750W/4Ω I think) which would also be a lot cheaper. But the Powersoft is marginally if insignificantly more powerful, and takes up 1/4 the vertical space.

The link with your concerns didn't work for me, but I have significant concerns with this test. Measuring a loud continuous tone tells you something, perhaps, about the average level you might listen to, but nothing of the peaks. For example, when my wife does a rim shot on her floor tom or snare it is loud, hurt your ears loud, but it only lasts a fraction of a second. A continuous tone at that level would be intolerable. The difference between these peaks and the average level is often 20db or more. I'm guessing a rim shot is at least 115db at 2 meters away.

For that reason the only way to properly ascertain the voltage level going to your speakers is using an oscilloscope with a data logger while playing music. One of these days...

 

[PENG]

Hmmm. I was just estimating this myself. My speakers are about 86db/2.83v/1m, and at my ear position I was just measuring 104db peaks with an 92db average level. At 1m from the lower midrange driver I measured 111db peaks at the same listening level. So what do you think, somewhere between 250 and 500 watts on peaks?

You get room gain and that means you should be getting more than 86db/2.83v/1m. Was the other speaker making some noise too? Also, a good 125W amp should be able to take care of some peaks now and then.

Aside from that, I hope you don't listen to 92db average level often and long and I am sure you know why I say that.

I always say more power is better, but I wish people (not you for sure) won't exaggerate so much about the sound quality improvements they are getting, when most of the time they are getting such improvements on occasions only, i.e. when playing certain music contents, spl, etc.etc.

 

[PENG]

For that reason the only way to properly ascertain the voltage level going to your speakers is using an oscilloscope with a data logger while playing music. One of these days...

I have a Hioki meter that has recording capability and that can capture peaks pretty good. It has a decent bandwith too. In my room even my AVR can hit over 100 dB without any trouble.

 

[GranteedEV]

You get room gain and that means you should be getting more than 86db/2.83v/1m. Was the other speaker making some noise too? Also, a good 125W amp should be able to take care of some peaks now and then.

Is room gain a frequency independant function? Does the reflected sound sum up in phase? How is the delayed reflected sound interpreted by the SPL meter? Does reflected sound contribute to perceived dynamic impact? Does the recording assume the speakers are combining for dynamic sequences?

 

[PENG]

I just called Denon tech support & asked about the pre-out voltage.

They claim that the AVR-3310 pre-out is 2V.

But, of course, they could be lying since most of them are incompetent anyway.

I suspect you guys don't have all the facts/information. Denon AVR specs typically say 'rated' output of 1.2V, or 1.2V/2.4V in the case of the 5308 and their prepro because of the balanced output. Marantz, on the other hand say 1V/2V for the AV8003 and 1.2V/2.4V for the AV7005. The latest Onkyo however would say somethng like 'rated' output 1V, maximum output 5.5V. If Denon were to specify the 'maximum' in addition to just 'rated', they would like be 6 to 6.6V, just my educated guess.

Aside from the variety of expression (rated, maximum etc.), presumable they all specify 'RMS' values, so if you measure the peak you will get a higher value, and if you measure the 'peak' when the input signal is a music signal, the peak could be much higher and if you measure peak-peak the value will be even higher yet. For a pure sine wave, peak=square root 2 (1.414)X RMS, so peak-peak will be 2.828X higher. Again, depending on the input waveform, it could also be higher than 2.828X. For those who claimed they had taken measurements, I am interested to know the what, how, using what.. Lastly, if you can believe what they tell you in the manual, then perhaps Denon AVRs have equal, or in some cases slightly higher output than Marantz's, now that's just food for thought so don't bother arguing with me.

 

[PENG]

Is room gain a frequency independant function? Does the reflected sound sum up in phase? How is the delayed reflected sound interpreted by the SPL meter? Does reflected sound contribute to perceived dynamic impact? Does the recording assume the speakers are combining for dynamic sequences?

Good questions, that's why I did not quote any value but I bet it would be higher. FWIW, room gain could even be negative but more likely positive. I just wanted to jog his memory about those variables so he wouldn't assume he needed 250 to 500W to get that SPL 1m in front of his speaker.

 

[GranteedEV]

then perhaps Denon AVRs have equal, or in some cases slightly higher output than Marantz's, now that's just food for thought so don't bother arguing with me.

but you know I will

The Marantz SR6004 and Denon AVR 5308 that I used preamp measurements, were both measured by Gene using the same methodology.

The receiver that costs 1/3 to 1/4 the MSRP has around 40% more preout headroom.

Denon's AVP processor however wipes the floor with everything, also an apples to apples comparision.

 

[Irvrobinson]

You get room gain and that means you should be getting more than 86db/2.83v/1m. Was the other speaker making some noise too? Also, a good 125W amp should be able to take care of some peaks now and then.

Agree, and they did. But 300w/ch seems to sound better louder. I also suspect that unlike the Mark Levinsons the ATI channels are conservatively rated. I'll bet there's 4db of additional headroom.

Aside from that, I hope you don't listen to 92db average level often and long and I am sure you know why I say that.

As you know, live music is pretty loud stuff. I probably listen to listen to 92db average levels with 104db peaks once or twice per week for about an hour, if that. But with the ATI amp it sounds quite remarkable.

I always say more power is better, but I wish people (not you for sure) won't exaggerate so much about the sound quality improvements they are getting, when most of the time they are getting such improvements on occasions only, i.e. when playing certain music contents, spl, etc.etc.

I understand, and the additional headroom was a benefit I didn't think would matter to me. I just wanted the quieter background I heard with the ATI for classical music. When louder music sounded better I figured the answer was obvious - I needed the additional 4db of headroom. I wouldn't have guessed it. I didn't order the AT3000 over the AT2000 because I thought I needed 300w/ch, it's just become my mantra to go-big-or-stay-home. I apparently got lucky.

 

[Irvrobinson]

I have a Hioki meter that has recording capability and that can capture peaks pretty good. It has a decent bandwith too.

What's the time resolution of the meter?

 

[Irvrobinson]

I just wanted to jog his memory about those variables so he wouldn't assume he needed 250 to 500W to get that SPL 1m in front of his speaker.

That wouldn't be jogging my memory... I didn't think they'd be relevant at such a short distance. If they are, well, that's the trouble with ignorance, it always finds a way to seep through.