PART-2
In the end you basically have two classes of products. In my opinion the more ideal of the two would be that which was designed by a very gifted engineer who has managed to extend the response of the external loop out to the highest possible frequency, while also maintaining a very stable and solid phase-margin under all load conditions. I believe this type of amplifier represents the highest potential for accuracy and resolution. Unfortunately such products are very rare indeed.
Conversely we have products being offered by the (in my opinion) "not so gifted" set. They see the struggle to achieve the above goal to be so great and fraught with hazards that they simply choose to avoid the issue altogether. Instead, they intentionally use either very little or no external feedback and thereby purposefully "wreck" the potential for low output impedance/high DF in their designs. "If you can't do it right, then don't do it at all" seems to be their motto and "less is more" their mantra. The upshot is that while it's true such designs do not incur the "secondary" form of distortion mentioned above, then again they do nothing to reduce the "primary" form either.
OK, so now we find that there is a specific class of amplifiers that the BBs have the potential to significantly improve system performance when inserted with. This is the class that uses some form of negative feedback, or at the very least, exhibits an output impedance that is lower than that of the speaker in a range from about 1KHz on up.
Good luck though asking your amplifier manufacturer just what its amplifier output impedance profile is from 20 to 20KHz and above. They will often quote a value as expressed in DF, but there's usually an unspoken "qualifier" involved. The truth is, DF is usually measured at some low frequency, often around 100Hz or a bit below. If you ask them what it is at 5kHz or thereabout, you'll likely get some comment that it doesn't matter and/or that the resistance of your speaker wires would limit it anyway. A likely comment would be that such a specification is "meaningless." Don't you believe it, though. The fact is that output impedance rises (and DF drops) significantly, starting in at around 500 to 1,000Hz in even very robust SS amps that use relatively large amounts of external loop feedback.
But the "size" (physical or power) of the amplifier is not a direct indicator. Neither is the basic design be it SET, Push-Pull Tube, Switching, nor Linear SS. Any of these can be designed to have an output impedance at frequencies above 1kHz to be lower than a typical speaker (4-Ohms or higher). There is a "general" rule though. SETs by their very nature will typically have the highest output impedance at ANY frequency, as compared to most any other design. Also, any amp that obviously claims to use no negative feedback will have a higher output Z as well.
In any case, the question is "how high can the output Z be and still get good results from the BBs"? In my estimation, if the output Z of a given amp is 2-Ohms or below from around 1kHz to say, at least 5kHz, the BBs offer a significant potential for an improvement in system performance. If the output Z is around 2-Ohms or even a bit higher, but the speaker's impedance in that region is 8-Ohms or above, then there's still a good chance for improvement using the BBs. Let's put it this way, if the amp's output Z at high frequencies is equal to (or God forbid) higher than the impedance of the speaker, save your $$$ and forget the BBs. In fact, I *could* recommend what to do with the money saved, but I'd likely ruffle some feathers if I did.
Oh, and if you have an amp that offers extremely low output Z and high DF all the way out to 20KHz and beyond, the BBs might not do much for your system either. After all, why would you need them? You already have the "perfect" amplifier. aa
There is one other aspect to the BBs design that I haven't covered in any of the above. That part has to do with the most "patentable" aspect of the BBs design. I'm calling that the "Smith Cell" and all I can say is that it is intended to lower 1/f noise as another "well-known" product is claimed to do, although via a completely different mechanism. This aspect also has the potential to be expanded upon and applications developed for other uses, even in other industries. Therefore I can't say any more about it at this time other than to tell you that its action is not affected by the amplifier's output impedance whatsoever. Therefore, even in the least likely of cases outlined above where the BBs could "theoretically" not be a predictable "good fit," there is still a chance they might help a little.
Well, that's my story and I'm sticking to it. See what you started? Now leave me alone so I can get back to work!
Take care all,
-Bob
