Electro-Mechanical Convergence: How can it affect your audio gear?
|Electro-Mechanical Convergence: How can it affect your audio gear?|
|1 May 2002|
First of all, what is Electro-mechanical convergence (EMC), and what is its relevance to things audio?
To offer some background on this controversial topic, a few years back, one of my editors told me (in no uncertain terms) that mechanical resonance's and vibrations could not be shown to have any measurable or audible effects on electrical performance. Period! He then dared me to prove otherwise. The fact is, he would not even let me mention the use of tweak products such as isolation bases, air-bladder suspensions, or after-market audiophile feet in connection with the products I reviewed. This was both upsetting and unsettling to me at the time because, although I felt I was hearing real quantifiable differences when using such products, I hadn't a clue as to how to prove it scientifically.
Over the years there have been many instances where it appeared that products like Sorbothane pucks, Andy Bartha Audio lead-shot Whachamacallits, Shakti Stones, air-bladder isolators, and Black Diamond Racing Cones, all made their respective contributions in tuning my system to sound its best. One evening not long ago, I unwittingly performed an experiment -- the results of which cracked me in the chops like a Mike-Tyson uppercut.
Phase I: Discovery
I recently reinstalled my AHT tube line stage preamp in my reference system in an effort to tailor the sound more to my liking, finding that although it sounded passable, the system didn't sound as striking or convincing as I had hoped (or as it had in previous iterations).
The AHT preamp was used to feed my InnerSound solid-state active crossover/bass amplifier, which drove the dynamic woofers of the InnerSound Eros Mk-II electrostatic hybrid speaker system. The high-pass section of the crossover routed the signal to Monarchy Audio SE-160 tube hybrid monoblock amps to drive the speaker's ESL panels.
Previously, I had attained very good results using "high-bouncing balls"
(a.k.a. Super Balls®) as home-brew audiophile feet under the InnerSound crossover/amp and under the AHT tube preamp. But now, as I played through "Mix One" of Rough Mix [MCA 2295] by Pete Townshend and Ronnie Lane, on vinyl, something about the twang and timbre of the guitars just wasn't cutting it.
I got the idea to try and improve the sound by experimenting with different types of feet under the tube preamp and under the InnerSound crossover/amp. In my storage closet, waiting for an opportune moment to arise, I had some small rectangular-shaped blocks of exotic, hard woods, which could be used as the mechanical interface between the chassis of the components and the MDF shelves of my Plateau equipment rack. First I removed the high-bouncing balls and placed a tripod arrangement of Cocobolo blocks under my tube preamp. Immediately I heard a change in timbre that caused the aforementioned guitars to sound much more natural and true to life. "Hmmm," I mused, "This is getting interesting!"
Next I propped the front of the InnerSound crossover/amp upon a single block of Zebra wood (ran out of Cocobolo). When I listened again, I thought I heard a further refinement in transparency and naturalness. Ronnie Lane's vocal on "Nowhere To Run" had become extremely lifelike, palpable, and focused-almost like he was standing right in front of me. Fantastic!
However, when I proceeded to prop up the hind quarters of the crossover/amp with the Zebra wood blocks and likewise, when I attempted to replace the #3 Black Diamond Racing cones under the tube preamp's outboard power supply with wood blocks, the sound became unduly bright and less natural. Too much of a good thing, I suppose. But at that point I was quite thrilled with my previous success and returned to the single Zebra block under the front of the crossover amp.
While listening and enjoying my finely tuned system, as fate would have it, one of my Monarchy SE-160 tube hybrid amps developed a problem and had to be shut down-just when all was bliss. So I removed the Monarchy SE-160s and installed a pair of Monarchy SE-100 monoblock amplifiers that I had on hand for review. The SE-100s provide slightly less power than their larger siblings and are entirely solid-state. Nonetheless, they use the exact same type of Hitachi MOSFET output devices, and I was hoping for comparable performance.
Leaving everything else exactly as it was, and restarting side one of Rough Mix, I was hoping I wouldn't lose too much of the magic I had just attained. To my disappointment, those notorious guitar licks did not ring as true as they had through the SE-160s. Rats!
Finding it difficult to believe that there could be that much difference between the respective amplifier pairs, I decided to do another experiment and change back to my original quasi-Super Ball footer arrangement under the AHT preamp and the InnerSound crossover/amp.
When I re-cued the record and sat down to listen, I almost couldn't believe my ears! It seemed that much of the former glory and trueness of timbre had been restored. The sweet authenticity of the fretwork and vocals was back. No, the SE-100s didn't sound exactly like the SE-160s, but it was a much closer contest. The senior amps still seemed a hairs-breadth smoother, but the timbral perspective and all other areas were shockingly similar-just when I was thinking that the SE-100s might only merit a so-so review. Hmmmm!
Since the changes I had made were not to the amplifiers themselves, but instead, to the upstream components, I found this whole process amazing. If this EMC phenomenon can be documented and supported by scientific measurements, then the implications of these findings may be significant. Obviously this new information would call into question the validity of A/B component comparisons where the end results were judged solely on A/B swaps without any attempt to optimize the sound for the particular component under test. And, sports fans, I can assure you that many reviews are written without such consideration.
At the present time the many audiophile tweak products that work on this concept are randomly applied with no guarantee of beneficial results. To cite one example, Black Diamond Racing Cones come in two models of different densities. One is said to be generally better for use with solid-state gear and the other is supposedly better for use with tube components. However, the user is encouraged to try both types of cones since there is presently no way of predicting which model will perform best with a given component. Further research may lead to a better understanding of EMC and to products that perform with greater consistency and more predictable results. Other practical applications that have not yet been conceived may be realized in the future, quite possibly in unrelated fields. A stretch? Well, maybe, but stranger things have happened.
Stick with me friends, as I endeavor to quantify and measure this phenomenon, with the help of my friend and colleague, Mr. Jens Waale, a Danish research scientist. Jens earned his MS Engineering degree from the Technical University of Denmark (DTU), and currently resides in the lovely Garden State, where he is a member of the New Jersey Audio Society.
Phase II: The Test Setup
Our testing protocol involved placing a speaker very close to the test component, and applying acoustic energy to the chassis in the form of a sine-wave sweep tone that spanned various frequency bands at roughly 90 dB. Measuring the component's output (with the inputs shorted) during the sweep showed significant spikes in the preamp's electrical output. The largest peaks we measured were roughly 25-30 dB below an average level of about 90 dB, which is likely to be audible under real-world conditions. The amplitude of the spikes varied with the frequency and intensity of the test tone and the type of surface and mechanical interface (feet) that the particular component was placed upon (Fig. 1). In rooms where one or more pieces of gear in the amplification chain are exposed to sound pressure that approximates the playback level at the listening position, EMC could be influencing the tonal character, or frequency balance, that is ultimately perceived.
Figure 1 illustrates the measurements of my AHT tube preamp, showing two significant but narrow peaks above the noise floor, one at just under 2.4 kHz, the other at about 2.5 kHz, which would likely be audible when playing music containing a strong signal at those frequency points. In the first run, the preamp was placed atop three Sorbothane pucks (two at the rear, one in front) and in the second run, the pucks were removed and the preamp sat on its four integral rubber feet upon the plywood test shelf.
On its own feet, the preamp shows a major spike at 2.4 kHz, which is reduced significantly when the Sorbothane pucks are inserted. Yet, the secondary spike at 2.5 kHz is much higher with the Sorbothane and the center frequency seems to shift slightly between the respective runs.
Initially, Jens and I had attempted to measure EMC in both solid-state and tube amplification components. While it was quite easy to demonstrate the effects of EMC with tube gear, its effects on solid-state designs was more difficult to pin down-using the particular amplifiers that Jens had on hand to test. His laboratory-grade solid-state amplifiers are built to very exacting specifications and tolerances for extreme stability under adverse conditions. Therefore they are likely less susceptible to the phenomenon than typical consumer audio gear may be. We were able to measure the same type of frequency-dependent peaks that we discovered with tube components; albeit, the peaks were farther down relative to the noise floor and may or may not be audible in this very specific case. That said, we also tested a high-gain, solid-state amplifier that exhibited very dramatic peaks in response to the acoustic pressure to which it was subjected. The unit appeared to be functioning properly, and sounded fine, but it was literally singing along in response to the acoustical stimulus-to the extent that we consider the results an outlier.
Further tests should be performed on other solid-state amplifiers in order to gain a clearer picture of the typical response in this subset of electronic gear.
Jens warned me of the pitfalls of drawing conclusions from these data; but I can't help believing that my preamp does indeed sound different under conditions that excite the resonant nodes that we observed. Although these peaks are still below the actual music signal, they are intertwined with it (similar to dirty AC) and it is likely that they produce audible effects. In theory this accounts for perceived differences in the sonic profile of the preamp under the conditions we've described. My subjective listening impressions of foot-dependent, audible differences in the upper midrange and lower treble region tend to support this. I have observed differences in the bass character as well, but in this article we won't go there.
Our experiment builds a strong case for the use of after market feet and vibration damping products, especially in applications involving tube gear. It supports what many audiophiles have been hearing in their own audio systems all along.
One thing is clear: we need to keep our ears and our minds open if we are to learn and advance.
Don't forget to bookmark us! (CTRL-SHFT-D)