Dear Friends,
Grab a cup of coffee, a donut or two and your smokes if you’re given to such and settle in for a marathon. I hope the webmaster doesn’t dump this for hogging up all of the drive space. I apologize for the delay in my reponse but I've had a very difficult decision to make. In light of Jerry’s decision to return the Timepieces, I am reluctantly forced to announce the dissolution of SP Technology Loudspeakers, Inc.
PSYCH!!! We’re a little stirred…but not shaken, and we intend to be around for some time to come. It is upon this premise though, that we will be making some modifications to the “No Risk” deal. If we don’t, we might not be around long after all. Financially, this has virtually no effect, as we are building to stock anyway. The issue has to do with having too many “demo” units that must be reduced in price when they are re-sold. That kind of loss cannot be recouped. We’re not about to pull any underhanded tricks either by repackaging used speakers and pawning them off as new.
Seeing that the following essay is rather long, I will be posting our policy change separately so those of you that are not interested in wading through it all can get to the facts quickly. And please do not take offense; I make no pretense of being the world’s authority on amplifier design or am I attempting to “grandstand.” This is only offered to those of you that are interested in my views.
To start, please bear with this lengthy dissertation but I would like to take some time to explain our motives. It can be summed up in Jerry’s own words in his review,
“Technically, the SP is superior in almost every category” and
“I guess I like my speakers to lie to me a little.” As anyone can see, we have what appears to be a bit of a dichotomy here. Don’t get me wrong, I find absolutely no fault with Jerry in this whatsoever. He’s been a very gracious gentleman throughout the whole process and I thank him tremendously. I also believe that he has a very keen and refined sense of hearing.
Never the less, the situation does point out a failing in foresight on our part and even a bit of naivety. We come from an engineering perspective wherein accurate reproduction of the original source is paramount above all else. Having known within ourselves that what we had achieved with the Timepiece was truly revolutionary, it was our belief that such performance would be the deciding factor in virtually everyone’s decision to purchase. In defense of this oversight, I must say that up until now, that was the case with all of the other sales we’ve had. Virtually every other customer has commented that the Timepieces were the best loudspeakers (accurate and realistic reproduction being the key) that they had ever heard, particularly at its size and price point. We’ve had many in the past that have heard them and not purchased but their professed reason was due simply to a lack of funds and proper associated equipment. This hobby ain’t cheap!
Jerry’s case is special for us though, only because this is our first encounter with his type of preference in sound reproduction. I anticipate that he is far from alone in his views. Now - I am not about to engage in the practice of “tube bashing,” so please don’t take it that way. I do feel though, the need to point out what is to me obvious in his case from a purely technical standpoint. Having extensive background myself in amplifier technology, I can say without reservation that, for tube technology to match the performance levels of the “better engineered while still affordable” solid-state designs, the manufacturing costs must be several times those of the equivalently powered solid-state.
The Tube vs. Solid-State Wars:The fact is, that when properly implemented (still very difficult and somewhat rare), solid-state amplification will reproduce the original source more accurately in a majority of performance areas than tube at a given power vs. price point. Tube technology IS NOT at all inherently inferior – just different. It can achieve extremely high performance levels but because of the very high cost of doing so, it is seldom attempted. Both technologies have certain strengths and weaknesses at frequency extremes (solid-state is good down to DC – tubes excel at U.H.F.) For the audio frequencies of our concern, both can be engineered nigh unto perfection if cost is not a factor.
The major limiting factors in tube design are: the requirements of the output transformer that a solid-state amp does not require, the complexity of series/paralleling multiple costly output tubes to achieve equivalently high output power, the need for quality components that are rated to operate at high voltages and temperatures (very expensive) and the practical inability to apply direct coupling methods between amplifier stages (a significant source of distortion).
In my past work I have been directly responsible for transformer testing and design. You would not believe how much distortion these devices produce – even before they begin to saturate, let alone when they are forced into mild saturation conditions. It is a common belief that the more musically pleasing, even-ordered harmonic distortions that tube amplifiers are claimed to produce, are due to the tubes own characteristics. This is only partially true. If they are operated in Class B mode (necessary for high power) the crossover notch distortion that occurs at the zero crossing point is of a symmetrical nature. This will produce odd-ordered harmonics just as in a solid-state design.
Rather, even-ordered harmonic distortion is the by-product of a non-linearity in one polarity of the source signal and is far more often associated with output transformer saturation due to plate current flow and the resultant hysteresis driven “flux walking” of the core material. "Flux Walking" is a term used in engineering that refers to an offset of the core material that reults from a type of "memory" characteristic it exhibits. With the repeated cycling of the core through its B-H loop or "hysteresis cycle," a residual magnetization offset occurs that will gradually "creep" towards one end or the other of its linear range. This will continue to occur until the core reaches some point of equilibrium and come to rest. That point then will not be centered in the linear range of its operation and hence, an offset that leads to asymetrical transfer characteristics.
It is interesting to note that a loudspeaker driver will produce the same effect (frequency “doubling” distortion) due to “flux walking” induced by high drive currents that forces the voice-coil to operate at a point offset from its un-driven rest position. This too is a result of magnetic field or “flux” asymmetry and is a consequence of inferior (cheap) magnetic structure geometry and/or inferior magnetic materials.
Manufacturers of electric guitar amplifiers intended for use by musicians often employ a neat little trick that takes advantage of this. They will often design an output transformer that is intentionally underrated; their cores being operated near the non-linear knee of their saturation curve, in order to generate that “warm” tube sound (sorry guys, I guess I ratted you out – tough beans). The only drawback is that the transformer will eventually get too hot and burn up due to core losses - after extended use. I’ve fixed a bunch of these on the side myself for an extra buck from time to time. I’m sure they sell more replacement amps that way too.
My only intention in pointing all of this out is to help you see the importance of that transformer. In order to avoid such distortions AND extend bandwidth to frequencies below 10 Hz (very important for amplifier damping factor and speaker cone control) an enormously large and expensive transformer is required for high power tube amplification. We’re talking about huge amounts of silicon steel (to maintain core linearity and avoid saturation on peaks) and large gauge copper windings (to minimize IR losses and maintain as low of a source impedance as possible).
The fact is that most tube amplifier manufacturers choose NOT to completely avoid core saturation. It makes the amp sound “sweeter” or “warmer” if used judiciously. Heck, Aphex Systems was built on this very concept. Way back in the mid 70’s their founder (forgot his name) mis-wired a tube amp he was building at home and derived this wonderful “sound” that imparted all those qualities to the music he played through them. A brain fart later and he was converting that “mistake” into a solid-state equivalent, rack mount gizmo and selling it to recording engineers for use in “sweetening” their tracks – and it worked great! They loved what the Aphex Aural Exciter did to make their recordings sound warmer, fuller, airier, etc. And the thing works it’s magic by using all solid-state electronics – go figure tube lovers!!! Hence – Aphex Systems was born. Linda Rondstadt was one of the first artists to use it on her vocals. She really does sing real good anyway though. Check ‘em out if you don’t believe me
www.aphex.com.
Sooo…if you want accuracy and realism, go solid-state – top notch, that is. If you want to re-engineer your recordings - go tube. And I think that all of you that choose the tube route have chosen a path that is just as legitimate as the rest of us. Music is supposed to be enjoyable and the audio hobbyist should have fun – period. There is no right or wrong, whatever makes you happy is right for you and I applaud your decision. The market is big enough for all of us to buy and sell what we enjoy. Just don’t try to convince me that the available tube amps are more accurate than most good solid-state.
STOP! I know all about T.I.M., complex H.F. inter-modulation products, slew induced distortion, excessive negative feedback, etc, etc, etc. -- probably more than many of you (not bragging – just fact). The fact is, if tubes produced less distortion you wouldn’t buy them. And please don’t ask me to engineer a speaker that necessarily sounds good with them. Our speakers might… but that still depends totally on the amp. To be honest, I don’t know how the other guys do it. The variables are too complex and trying to shift speaker distortion modes around to find a pleasing combination too much like magic. That’s far more art than science and I’m no DaVincci.
To end this chapter, the upshot is that with the need for a huge transformer and many rare and expensive tubes along with the other issues mentioned, you don’t see any high-powered, ultra low distortion tube amps – they are way too expensive in comparison to equivalently powered solid-state designs. Just try finding one that will drive a 1-ohm load for any length of time. Besides, they sell far more of them with the distortion than they would without. Anybody remember the C----- S----- 7’s? It was a nice looking, high-powered (350W/Ch.?) tube amp…and it sucked (IMHO)! C----- permitted way too much distortion in that dog. Sorry C----- owners, but I’ve heard the thing and it was a sad joke. $11,000.00+ for a sexy, screaming bitch. My ears still ache from it. I guess that’s cheaper (in the long run) than marrying one though.
Remember friends, all this doesn’t mean that tubes are inherently inferior. It just means that we’re stuck with having to live with lower powered ones and their inherent, but pleasant distortion. Even Jerry’s 200 Watt/channel ARC isn’t a large amp by today’s standards. And he’ll be the first to admit that he doesn’t care much about specs and distortion measurements. And why should he? He knows what he likes and that is all that’s necessary for anyone to make a decision on. We’re not supposed to be listening to theory and design concepts with our systems. It’s about the music. Some of us like ours plain so we can savor all of the natural flavor, and others like it hot and spicy. Who’s right?
The only thing is…
THE TIMEPIECE NEEDS A VERY LOW DISTORTION, HIGH POWERED AMP!!!That’s just the way it is and I make no excuses. So for all you accuracy lovin’, realism seekers out there that crave a holographic soundstage, killer dynamics, tight “in your chest” mid-bass and deep bass that you can feel moving the air around you…we got the ticket for you. Y-e-a-h, bay-bay! And the price of admission is a goin’ up. We ain’t catering to those fickle tube jockeys no more. Only the gods and their shamen have a clue about what will make them fella’s happy. I tried but I just can’t lay my grubby little fingers on any good pixie dust to sprinkle my speakers with.
So for you meat and ‘taters guys, here’s the deal: If you got the amp – we got the speaker and we’re still willing to stick our necks out to prove it. No sissy amps allowed though. 300Watts per channel minimum into 8-ohms, solid-state or forget it. You’ll have to convince me you’re in the club as well. Don’t think you can fool me either. I can smell a bottle-head a mile away now ‘cause I’ve picked up the scent real good. Once bitten, twice shy (just kidding Jerry, you’re cool).
All joking aside, we can’t be everything to everybody and there’s no sense trying. I realize that Jerry’s observation about the lack of “air’ from the Timepiece was in comparison to the Dyne’s being auditioned on the same amp. This means that the tube issue is not the only cause of his observations. What it does do is reveal his preferences in sound reproduction in general. The fact that he prefers the tube “sound” is nothing more than an indicator and only ONE easily identifiable manifestation of his underlying preferences.
Maybe I’m wrong, but I deduce from this that people that prefer tubes are, in general, highly mercurial yet paradoxically determined in their preferences. Why else would someone be willing to trade all of the advantages and superior performance offered by the Timepiece (as Jerry himself, has confessed to) in order to satisfy that one rather minor and “ethereal” criteria? That’s OK though and I applaud his right to do so. It’s just that we need more assurance that an individual has a greater likelihood of being satisfied with our products before we ship dozens of speakers around the country, only to have a significant number end up returning for such highly subjective reasons.
We still take the position that performance can be analyzed and verified scientifically, at least to a first approximation. IMHO, every effort must be made to optimize all measurable performance variables before subjective optimization of a product is undertaken. If a product excels in all measurable areas FIRST, then by default, its performance will be observed to be superior to the majority of unbiased reviewers as compared to other products that fall short in a number of those same areas. Notice I said the MAJORITY of reviewers – not all. That majority constitutes the folks that our products are being marketed to.
So Where’s the “Air”?In Jerry’s review, his main concern was the lack of “air” being reproduced by the Timepiece’s. What is this quality and why didn’t they deliver? While I’m sure that there are many that will take issue with the following, I am persuaded that our explanation is more correct than not.
We can all safely assume that this quality of sound reproduction resides in the realm of the tweeter’s operating range. I will admit that there may be a slight difference in the two different tweeters that has an effect on this quality, but I do not believe that is the major cause. If it were a simple matter of accuracy, then the Timepiece can hardly be faulted. One only needs to here the delicate harmonics of a gently brushed cymbal being faithfully reproduced by them to realize the natural and life-like reproduction they provide. Yet, no “air”?
Once again, I default to the amplifier. In order to explain, a brief review of amplifier behavior is once again in order.
If one examines the published distortion specs of any amplifier, they will find that harmonic and/or intermodulation distortion is specified in percent of “total rated output.” This means that the distortion levels are measured as a percentage of the total signal being reproduced by the amplifier when it is being driven to its maximum output power just before the onset of clipping distortion (i.e. hitting its full power limit).
What may not be obvious from this is that ANY decent amplifier will produce the best (lowest) measurements under these conditions. The reason for this is that distortion and noise are usually of a “static” nature. “Static” meaning that the levels of these by-products does not change much, regardless of the output power being delivered, until the amplifier reaches its power limit and begins to “clip” the signal. Then, of course, the distortion levels skyrocket beyond that point. As a side note, the static levels will usually increase slightly in a tube design as you increase output power. This is in contrast to a solid-state design, which typically does not (or very little). The cause of this effect in the tube design is primarily due to the output transformer and will be discussed in length in the following text.
Now at this juncture, I refer back to the previous discussion regarding tube amplifier distortion. Upon inspection, anyone can see that of those manufacturers that do publish distortion specs, the tube units generally display slightly higher distortion levels than equivalently powered solid-state. We have already discussed that, to a certain extent, such distortion is intentional. At the very least, it is an unavoidable consequence of tube design. In either case, it is the source of the beloved tube sound.
The issue here is that as output power is lowered, the percent distortion specification increases. This means as you decrease the amount of power being drawn from the amplifier, a higher total percentage of distortion (or ratio of distortion to signal) is being delivered to the speaker. This is just as true in solid-state amplifiers as well. When one couples a higher efficiency speaker to an amplifier of any kind, less total output power is required to achieve a satisfactory volume at the listening position as compared to a speaker of lesser efficiency. Therefore, a more efficient speaker will reproduce more distortion from the amplifier as a total percentage of sound being heard by the listener. If this speaker is also a relatively accurate design (as in Jerry’s Dynaudio’s), then the amplifier’s own distortion products will not be masked by the speaker’s otherwise self generated (and less pleasing) distortion.
To sum up, the “warmth” and “AIR” that is the result of tube amplification will be reproduced more audibly from a more efficient speaker than from a speaker that is equivalent in every other way, but is less efficient – being driven by the same amp. This is the very premise of the S.E.T. concept and why it works so well for its advocates.
As a side note, you may argue that the tube sound is still apparent to a certain degree, even when driving the amp harder, such as into a less efficient speaker as the Timepiece. This is true, but the character of that sound (i.e. distortion) is slightly different. Remember our discussion on output transformers? In tube designs, as you leave the realm of static distortion by pushing the amp harder, transformer saturation distortion begins to “take over” or dominate the spectrum of distortion products being delivered to the speaker. This type of distortion has a rather gradual onset and begins well before the amplifier begins to hard clip.
The difference in sound quality is due to the inherent physical characteristics of the transformer. Since it is nothing more than a bunch of turns of wire (both primary and secondary turns) wound on a ferrous core, it exhibits the properties of “self-inductance” as well as hysteresis induced distortion. The self-inductance is in series with the current flow out of it. Any student of electronics knows that an inductor resists or “impedes” any change in current attempting to pass through it. This is a property of filter theory and basically means that as the frequency through an inductor increases, the current decreases.
The upshot is that the transformer’s own self-inductance, along with the more complex core hysteresis issues (a very large factor) reduces the amount of higher frequency distortion products that it will generate or even pass on to the output, even if it is produced by the tubes upstream. It will still generate and/or pass harmonic distortion, but those distortion products will be mostly dominated by lower (bass and midrange) frequency artifacts. The output transformer simply cannot generate and/or pass any significant quantity of the higher frequency distortion (above 5 kHz.) harmonics when being driven hard.
For those of sound recording/engineering background; the total effect is not unlike that of a downward expander - specifically acting on the higher frequency harmonics. This effect can even be seen to a certain degree in the lab by doing a simple frequency response plot of a tube amp while being driven to full output, under rated load conditions. It is common to see a slight roll-off of the higher frequencies starting around 5kHz. 4-ohm loading usually produces more roll-off than 8-ohm due to greater core saturation effects.
So… that’s why a tube amp may sound a little “bloated” in the bass and still retain its midrange “warmth” when being pushed. The one thing it won’t do well is deliver the “AIR” at the higher end of the frequency spectrum when under larger output demand. AND THAT’S WHY THE TIMEPIECE DOESN’T REPRODUCE THE “AIR” FOR JERRY!!! He has to push his ARC harder on the Timepiece’s because they’re less efficient and “POOF” – no “air.” I hope this all makes sense and I apologize for the lengthiness – it is somewhat complicated.
Just remember, we’re living in an age where men have already been to the moon and space flight is becoming commonplace. NATURAL LAW orders the universe. The same physical laws that govern everything else also govern audio performance. It can be analyzed scientifically and it is not magic. It is also not ROCKET SCIENCE!!!
God bless!
-Bob
