speakers with a "smooth" impedance curve

So now your mantra becomes "a smooth impedance curve above 8 ohms with a first order crossover ."

Good luck.

Most Tannoy's are very easy to drive.

Having used a particularly load-sensitive tube amplifier in designing fairly high-efficiency loudspeakers, in my experience smoothing the impedance curve via additional crossover componentry does not have a downside. Electrically, the amplifier sees the net impedance curve (including its phase), rather than seeing numerous individual components. Smoothing the curve (and reducing its phase angle) is beneficial.

Let me give an example. A friend of mine has speakers whose impedance curve wouldn't work well with an OTL tube amp, so I designed an outboard filter to smoothe the impedance curve. Not only did this work with the OTL amp, but he also reported a significant improvement even with his solid state amp - which is what his speakers were originally designed for.

You can read his account here:

http://www.audiocircle.com/index.php?topic=56415.0

If the additional circuitry was somehow detrimental, but just happened to synergize well with the tube amp, then surely that detriment would have been revealed by my friend's high-resolution class-A solid state amp. What he said:

"...also mentioned even my current SS amp's performance might improve w/ this filter. Sound quality was indeed improved: smoother, more ease, less tension, less grain, tighter focus + increased musicality, image/stage improved a lot...etc. The difference can be heard even in the next room."

I think the effect he describes comes from the reduced phase angle of the impedance curve; at any rate, there was no detriment reported.

Duke
dealer/manufacturer

Well, I guess we cannot agree all the time, Duke...

Anyway, I probably understand where you are coming from, even if I disagree as far as the benefits go. I just would like to clarify a bit more so that I can at least say that we are more or less on the same page of what is at work here, even if our opinions diverge.

Along those lines, a few questions, if I may:
1) From your statement, I infer most of the direct experience you have with the network you mentioned implementing is in partnership with OTL amplification.

Is this correct?

2) The solid state amplification experience is indirect, through the customer.

Is this correct?

3) Normally, large impedance swings are "smoothed" by networks that flatten their peaks (though I know of one that works the other way), which, if we are interested in partnering with OTL amplification would kind of run counter to that. Though, without making an already long post even longer, I can understand the argument against the statement I just made, and in some instances, I even support it.

Can you provide a high-level description of this network you are referring to, please? Of course, I respect the fact that some folks are more sensitive to discussing things than others such as myself normally are. But, if it's a well-known, previously illuminated topology, as opposed to something truly unique, novel, and patentable that you would feel uncomfortable in discussing, please describe it.

Finally, again, I understand what most designers are aiming for with these impedance compensation networks. In fact, I accept that I'm likely representing the minority opinion here, as without naming names, what I consider to be the flagship loudspeaker of the current high-end audio marketplace implements them, as well as a highly successful company that offers tweaks as an add on to one's existing pair of loudspeakers. The benefits are more or less exactly as outlined in the words of the customer you quoted. Nevertheless, I disagree with their implementation in most instances, as I feel the detriments they add in terms of robbing the music of immediacy, openness, and dynamism seem incredibly difficult to overcome elsewhere in the audio chain.

as usual I agree with Duke. I'd just add that when a crossover is designed for a speaker to have a smooth impedance curve this can be done without using any filters or extra components, it just takes a lot more time etc., some companies do this some don't. Mine are built and tested using solid state amps, but smooth is smooth and the real difference between the SS and tube amps is what impedance range they like to see, both work best with a smooth load.