How to fix my detailed, accurate but BRIGHT system

These speakers are best driven by an amp that doubles power output as impedance is halved... so that the volume in the bass is equal to the volume in the mids and highs. Your Butler 2250 does not double power as impedance is halved.... The result is that the amp produces somewhat louder volume in the highs than it does in the bass, which causes a tonal imbalance that emphasizes the highs....

As long as an amp produces more decibels into higher impedances (treble region) than into lower impedances (bass region), the speaker will sound louder in the treble.

Tvad, this is one of the rare occasions on which I must disagree with you. Unless the amp is approaching the point at which it would be running out of either current capability or voltage swing capability, frequency response flatness will have nothing to do with the ability to double power into lower impedances. Basically, the amp will provide a flat frequency response into the speaker's frequency-dependent impedance as long as the amp's output impedance is low.

The 2250's output impedance (or equivalently, damping factor, which as I'm sure you know is output impedance divided into 8 ohms) does not appear to be specified. However, even though it is a no-feedback design, I would assume it is a small fraction of an ohm, which would result in just a small fraction of a db difference in voltage delivery into 4 ohms compared with 6 ohms or 8 ohms.

Any difference in power delivery which that may result in is a function of the speaker design, not the amp design, as long as the amp remains within its output current limitations, which it should as even the last line of your last post seems to indicate.

I think the posts by Johnny, Rich, and others about addressing the room issues are very well put and on the mark.

Regards,
-- Al

Al,

I cross-posted with you, so I hadn't read your post, before I posted.

Are you saying that an erratic electrical speaker impedance curve doesn't actually present a more difficult load to the amplifier in this case, because it's still within the acceptable limits of the amplifiers current delivering ability, and there won't be a problem with a frequency db imbalance, unless the curve dips far enough below the amplifiers minimal impedance, lets say 2 ohms?

Rich

Hi Rich,

That is sort of what I am saying, but I would put it a little differently. As long as the output impedance of the amplifier is small compared to the impedance of the speaker at the lowest point of its impedance vs. frequency curve, and as long as the amplifier is not called upon to deliver more current than it is capable of delivering, then a tonal imbalance will not result.

The fact that an amplifier can double its output power into 4 ohms, compared to 8 ohms, is an indication that it has good output current capability. In this case, the fact that the amplifier can put 250W into 8 ohms but only 400W into 4 ohms, is an indication that its output is being limited into 4 ohms by its output current capability, and therefore the peak volume levels it can produce accurately into a speaker which dips to 4 ohms will be less than for a more benign speaker load (everything else being equal). But as long as the output current capability of the amplifier is not exceeded (and keep in mind that we are presumably talking here about tonal balance at moderate volumes), and as long as the output impedance of the amplifier is low, no frequency response imbalance will result.

A notable exception to all of this would be amplifiers which have tube output stages and are output-transformerless, such as the Atmasphere's, which may have an output impedance in the vicinity of 4 ohms or more. That would definitely result in a tonal imbalance working into a speaker that is basically 4 ohms in the bass and 8 ohms in the treble (see Stereophile's impedance measurements on the Aerial 7B). I recall Tvad once mentioning that he had an Atmasphere at one point, and perhaps that is what prompted his comment.

Regards,
-- Al

It’s funny the notes of system fixes follow the trends of their posters. Gear houhnds point to the gear as the bug a boos, room treatment aficionados’ allude respectively to adding items there, cable mavens exclaim buy better wires, etc. well, to each their own aim.

LOL! Interesting observation, Jim. I suppose that a reason for that is that it's hard "at a distance" to have a feel for the DEGREE of excess brightness, and if everyone had first-hand exposure to the sound, opinions would probably be more convergent.

BTW, re Wireless200's comments, I want to make sure it is clear that I did not mean in my earlier posts to totally rule out the possibility of the amp as being a contributor to the brightness, and I don't disagree with Tvad's assertion that "these speakers are best driven by an amp that doubles power output as impedance is halved."

What I was saying essentially was that assuming the amp has low output impedance, the following is incorrect (aside from the first sentence):

Your Butler 2250 does not double power as impedance is halved.... The result is that the amp produces somewhat louder volume in the highs than it does in the bass, which causes a tonal imbalance that emphasizes the highs....

As long as an amp produces more decibels into higher impedances (treble region) than into lower impedances (bass region), the speaker will sound louder in the treble.

Regards,
-- Al

I'm not disagreeing with your discussion of low output impedance. However, in rereading Harley's discussion of power output into varying impedance loads, and the resulting loudspeakers' varying dbW (decibel watts) measurements, he makes no mention of low output impedance negating the effect. Perhaps this is an issue of voltage paradigm versus current paradigm?

I don't have Harley's book, and so I don't know exactly what he is saying, but yes the question can be considered in the context of the two paradigms of amplifier and speaker design. Those being the voltage paradigm, and the power paradigm, which is more accurate terminology than "current" paradigm, as explained in Ralph's (Atmasphere's) excellent paper on the two paradigms.

Consider the output stage of an amplifier to be a theoretically ideal voltage source (zero output impedance), the voltage being proportional to the amplifier's input voltage, in series with a resistor (equal to the amplifier's output impedance).

In a voltage paradigm amplifier, by definition, the value of that resistor approaches zero (i.e., it will be a small fraction of an ohm). The result is that the speaker will see a voltage proportional to the amplifier's input voltage, regardless of what the speaker's impedance may be at the frequency that is involved (as long as the amplifier is capable of supplying the required current, the required current being higher as the speaker's impedance decreases -- recall Ohm's Law). Nearly all amplifiers with solid state output stages work this way, and the majority of conventional box speakers are designed based on the assumption that they will be driven this way. Many tube amplifiers approach this model, although only approximately because their output impedance is typically higher. Other tube amplifier's, with even higher output impedances, fall into the power paradigm category.

As Ralph's paper mentions, a significant downside of voltage paradigm amplifiers is that they typically (but certainly not always) require more feedback than power paradigm amplifiers, increasing the well-known side-effects of feedback.

In a power paradigm amplifier, the output impedance is much higher, for instance 4 ohms or more in the case of many of Ralph's designs. That will cause both the voltage that is seen by the speaker and the current that is drawn by the speaker to depend on the impedance of the speaker at the particular frequency that is present. The higher the speaker's impedance at the particular frequency (or frequencies), the more voltage it will see (because it represents a greater fraction of the total impedance that is in the path, meaning its own impedance plus the amplifier's output impedance), but the less the current that will flow (because the total impedance in the path is greater). Since, if we neglect the effects of inductance and capacitance, power is equal to voltage times current, the power that is delivered to the speaker (as opposed to the voltage) will remain fairly constant as a function of variations in the speaker impedance.

As I said, most speakers, especially box-type speakers, are designed with the expectation that they will be driven with voltage-paradigm amplifiers. But Ralph's paper includes this statement:

Loudspeakers that operate under Power Paradigm rules are speakers that expect constant power, regardless of their impedance. Examples include nearly all horns, ESLs, magnetic planers, a good number of bass reflex and acoustic suspension designs. Horns, ESLs and magnetic planers do not get their impedance curve from system resonance and so benefit from a constant power characteristic and indeed, many of these speaker technologies are well-known to sound right with Power Paradigm amplifier designs.

So that is some background. Returning to the original question, I think all of this should make clear that a tonal imbalance can result from a paradigm mismatch between amplifier and speaker, such as the excessive brightness that would undoubtedly result from using a power paradigm amplifier (high output impedance) to drive this particular speaker (4 ohm impedance in the bass, 8 ohm impedance in the treble). But a voltage paradigm amplifier (near zero output impedance) would deliver essentially the same voltage into both the 4 ohm and 8 ohm impedances, which is presumably the expectation the speaker was designed based upon (or it would not sound right with just about any solid state amplifier). And the ability of the amplifier to deliver twice as much current into 4 ohms than into 8 ohms has no direct relevance to tonal balance; its main relevance is to maximum volume capability. Although, of course, for any of many other possible reasons one voltage paradigm amplifier may sound different with the particular speaker than another, and it stands to reason that an amplifier that can double current into 4 ohms will, everything else being equal (which of course they rarely are), be more comfortable dealing with a speaker like this.

I'll add in closing that although I haven't read Harley's book, I have read a lot of his writings over the years in TAS and Stereophile, and I suggest that you do not exclude the possibility that anything he says of a technical nature may be flat-out wrong.

Best,
-- Al

TVAD:On the other hand, we can also not exclude the possibility that he is right. In the acknowledgments section of his book, Harley thanks several experts in specific fields of audio who provided technical review of his manuscript. I am going to take the viewpoint that they know their fields, and that therefore the information provided in the book has been determined to be correct.

TVAD,

Let's try an example, because I suspect that there is some misunderstanding somewhere in the path from what Harley meant to say, to the words he used, to your interpretation of his words, and to my interpretation of your words.

We have a speaker that is 4 ohms in the bass, 8 ohms in the treble, and an amplifier that we are assuming has an output impedance which is negligible in relation to 4 ohms, and which can put out 250 watts into 8 ohms. And we are assuming that the speaker, like most speakers (especially enclosed box type speakers) is designed based on voltage paradigm principles.

Let's say the amplifier is putting out a treble tone of amplitude 8.94 volts, which is 10 watts into the 8 ohm speaker impedance in the treble region. The corresponding current that will flow at that frequency is 10/8.94 = 1.12 amps.

Let's then say that the system is also called upon to put out an equal volume bass tone (either simultaneously or at a different time; it doesn't matter for purposes of this example).

My first contention is that the amplifier will then put out a bass tone which is also 8.94 volts in amplitude, which will correspond to a power delivery, into the 4 ohm impedance which the speaker has in the bass region, of 20 watts, and a current flow of 2.24 amps.

My second contention is that, everything else being equal, an amplifier that can provide 250 watts into 8 ohms but cannot double its maximum power into 4 ohms will handle that situation no differently than an amplifier that can provide 250 watts into 8 ohms but can double its maximum power into 4 ohms.

And my interpretation of your original post, based on your interpretation of Harley's book, is that in this situation you and/or he claim that the amplifier which can double its maximum power into 4 ohms will deliver the 20 watts, while the amplifier which cannot double its maximum power into 4 ohms will deliver something less than the 20 watts, thereby resulting in excessive brightness. Which I contend is wrong.

So am I correct in thinking that there is a misinterpretation somewhere, or do you in fact disagree with any of my contentions above?

Best,
-- Al