Amp's nominal power rating - any use?


I just paired a couple of Coincident Frankentein monos with my SF Guarneri Homage. The sound is great (fat, rich, dynamic, transparent) and sounds well with any type of music (opera, rock, electronic...). These are 8W monoblocks and sound like with more power or at least the same as my previous fabulous pass aleph 3 (30W class A SS). Of course it depends if tubes not tubes, class A or not, speaker sensitivity, impedance load, room dimensions etc, but what i see is that it's not a relevant criteria at all on its own. Maybe there should be some transformation formula to take into account some of these factors to get some Apparent or Perceived Power, but maybe it would be hard to take into account all factors. Any ideas, opinions, on this?
dongiovanni
It actually plays louder than my previous amp, with a more open soundstage and a more vynil sound. I think it's perfect, it goes beyond my expectations...
Dongiovanni,
Various opinions here as usual and of course the typical stereotype comment by Irvrobinson regarding SET amplifiers. Always trust your ears, you know what sounds best to you. Clearly you`re pleased with this amp/speaker combo. If you have the chance to try some SS amps I think you would be hard pressed to equal let alone better your present system. You have in your posession a remakably sounding amplifier.
The issue with power ratings is that there are so many variables, and many ways to test power output. It also takes ten times the power to increase the volume by ten db. If you are happy with what you have, and it's only a matter of worrying about the wattage specifications not aligning up, don't worry about it. As long as the amp can get loud enough without clipping, which sounds bad, you are good to go.
Dongiovanni, I confess to being biased against SET amps. I think most people that like them are just hearing a modified frequency response caused by the electrical anomalies of the output stage, like a high output impedance. Frankly, considering the high quality of your speakers, I think it might be very interesting if you try a solid-state amp and tell us what you think. If the receiver is good enough (they do okay above approximately the $500 price point) you will at least get a first impression, though the speakers are good enough a better amp might have an audible difference. If you still think the SET amp sounds a lot better, I respect that. Choosing a system is personal decision.
Couple of points....and some agreement.
Yes, phase and impedance ARE frequency dependent.
I think the main take-away would be that a large phase angle coupled with an impedance dip is an amp stresser....(is that a word?)

Also.....and I don't remember which, Tube amps simply do not like a very reactive load. Capacitive? Inductive? don't 'memeber which.
Their is a good reason for this to do with transformers, right?

I don't believe 'next to nothing' is a fair characterization. Given that the PAIRING is a most important aspect to choosing amp/speaker combos, you can't simply ignore phase/impedance characteristics of the load and the amps response TO such loads.

Overcome a 'weird load' with simply more amplifier power? A solution, of sorts. I've heard talk of 'cheap watts' which is one answer. Another approach is the rational design of speakers.... We've all heard of some of the wacky stuff out there which apparently need something from Lincoln or Miller to drive.
I'll leave that decision up to YOU!

Now, about a proposed amp test? I think I'd like to see some data before I simply tossed that out. A resistor may be a nice, standard way to test an amp, but has little or nothing to do with real world conditions. Even my panels have reactive elements, though the impedance curve is pretty flat with only a rise at the crossover frequency. I'd love to see amps tested for power with real world reactive loading. A simulated speaker load could be agreed upon. The worst loading should be below....say 1khz, where most of the amps power goes, anyway, in real music.

And as for the OP's speakers. Fine stuff. Pricey and apparently worth it to some. I can't possibly criticize that choice. But, I would caution the owner to look at the pairing amp carefully. A large phase angle load, regardless of impedance, will still eat into amp power. Large impedance swings also, IMO, kind of defeat the purpose of tube amps. The more recent stereophile test of the similar SF speaker says you may want to use the 4 ohm taps, even though a straight look at the impedance curve may indicate the8 ohm taps.
If I wanted tubes with that speaker, I'd for sure try both.
The OLD test by stereophile did not measure such phase data. The new test indicates moderate phase and impedance curves....probably not a bad load for tube OR SS..
Magfan, the phase angle discussion you're trying to have is frequency-dependent, which you've ignored, so "next to nothing" still seems a stretch to me, which was my point. Since most listening, even on the OP's speakers, will be done with the first few watts, the effective frequency response will be the most important determinator of the sound.
Irv,
I just reread my response and will pretty much stand by it.

Just for an unreal example, a phase angle of 90degrees will result in ZERO power making it to the load. Using the 4foot fluorescent tube next to my workstation as an example, it draws about 1/3 amp. This lamp is consuming about 40va or 32 watts....the power factor is 0.8 and the phase angle is about 37 degrees. The math? PFxVA=Watts.

Tube amps do not like some reactive loads. An 8 watt SET working into a 60degree phase will deliver HALF its rated power to the load. SS is subject to the same limits. But, the amp may not even do that well....half is the upper limit.

I have not seen enough data, but would go out on a limb and say that many amps, even some highly touted ones, won't even do as well as a straight theory calculation (cosine of phase angle x power) says it should.

http://stereophile.com/standloudspeakers/807sonus/index4.html

Stereophile measures of recent version of the SF Guerneri.

Original speaker....from the 90s.

http://stereophile.com/standloudspeakers/487/index5.html
Kijanki, it is difficult to disagree that music has varying loudness, but 45.9%? I don't think your calculation is nearly accurate enough to claim a number with such precision. For one thing, the dynamic range of music varies so much by music type and recording, though I suppose someone with a SET and minimonitors is unlikely to be listening to Kayne West.

Regardless, I'm still not bought into the efficacy of your calculation.

As for Magfan's comments about adding capacitance or inductance into the power calculation, "next to nothing" seems like a big exaggeration. But considering we're talking about a tube SET here, which is going to be the most sensitive design to load variance, amp and speaker matching is of more importance than usual, I'll agree. Once you're in SET territory, I think the only practical way to tell if an amp is compatible with a speaker is to listen. No one publishes enough specs to make a worthwhile prediction, and even if they did, very few of us could probably do a reasoanble interpretation.
Irvrobinson, no, i never tried the SF with higher power amps. Maybe i should now for curiosity, since due to planning the roll-over to a tube amp, not long ago i got a aprox 100w onkyo receiver for home cinema (with some mordaunt-short speakers) so i have separate systems now for music /movies. I watch cinema a lot and wasn't imagining tiring those delicate 300B with movies...
Magfan,
Your suggestion for measuring speaker output is very much spot on, I wish this would become an industry standard procedure. The current method is of so little real world value in trying to determine an amp`s potential match with a given speaker.
Amps are rated into resistive loads which says next to nothing about how they do into a reactive real world speaker.

I would propose that amps be rated into 8 and 4 ohm resistors....
Than at the same 2 impedances into +-45 degree reactive loads.

The resultant chart would be a very visual indicator of how the amp does into real loads.

That's the way .....at least on paper to see what an amp can be expected to do.
Irvrobinson - let me explain. Average music power delivered to speakers is very low - just a few percent of peak power. It is because music on average doesn't go more than half of the loudness and that means 1/10 of the power. In addition music has gaps. Amplifier's max continuous rms power is completely irrelevant. What is important is peak power that amp can deliver.

Imagine two amps:

1. 10W rms, 100W peak
2. 100W rms, 100W peak

In my opinion they will have exactly the same max loudness with music signal.
High-end amps almost always achieve their claimed wattage ratings, but there are amps and then there are amps, and in this regard, what Charles1Dad wrote is true: an amp with massive power supplies (and if it's a tube amp, one that also has the highest quality output transformers) will sound considerably more powerful than most amps of the same wattage rating. Very few amps fit this profile, however (and the ones that do are very expensive), so you're basically stuck with wattage ratings.
I have no idea what Kijanki is talking about, with his percentages.

Your speakers are good for about 87db for 1 watt into 8 ohms (at one meter). This means an amp that clips at 8 watts will drive the speaker to 96db at one meter, which most people will consider quite loud. I would suspect a lot of people don't really listen to speakers this small any louder than that.

The 30 watt amp will be able to drive the speakers 6db louder, all things being equal, and they probably aren't. 6db is pretty significant, but since the 6db is 6db over 96db, you may not notice it, depending on the music you're playing, the size of your room, and your average listening level.

Since these are single-ended amps you're talking about, they will have a sonic signature, and the transfer function differences between the amps just may overwhelm the power differences.

Have you ever connected your speakers (which are pretty good for speakers of that size) to a decent solid state amp of, say, 100w/ch, just for grins?
Its probably the best reasonable rating anybody could come up with without getting to much into specs that are hard for a lot of people to understand.If they give it at full bandwidth,it sure helps.The speaker being driven creates a hard power rating spec for a one fits all amps.Their impedance,(speakers resistance) is always varying making power ratings tough for amps rating.Amp matching with speakers can be more critical than watts.Speaker efficiency is very important also.Doing some research on whether your amp will drive xyz speakers can be more useful,especially if they have low impedance dips.
Isolated output power ratings hardly tell the whole story. You could buy a receiver rated at say 150 watts/channel and know it could`nt keep up with a Pass XA.5 30 Watt/channel at all. That Frankenstein is an 8 watt beast due primarily to excellent custom transformers and an enormous power supply, it uses the 6EM7 tube as a driver for the 300b(supplies a lot of current). A case of quaility over quanity.
Just calculating perceived loudness:

(30W/8W)^(1/3.5)=1.459 30W is 45.9% louder than 8W but only in identical conditions. These power rating apply to max rms power and are pretty much useless with music. Peak power is more interesting since average music power is only a few percent of it (unless one listens to sine-waves). You don't really know how much of peak power each amp can deliver and it is not even possible to measure distortions at the peaks. Without measuring you perhaps push each amp to limits of audible distortions that are much less audible in even harmonics generating tube amp (even many percent) than odd harmonics generating SS amp (fraction of percent). Interaction with the speaker, being complex load, might also be different for tube and SS amps.
As you point out perceived power is the result of a set of complex variables that would be impossible to reduce to a single formula. The amp output is one of these variables and as such has a value.