Crossovers and the Audience
Auricap XO: Excellence in Capacitors
The old Collie lay sleeping on the bare wood floor, resting quietly...
At one point I noticed his legs moving, at first twitching, and then almost swinging in syncopation. He was dreaming again. I wondered what it must be like for him to be growing old, barely able to walk in his waking hours, thanks to a stroke that weakened the entire rear of his lanky body. Sleeping and dreaming, he was in his youth again, running after a ball, or chasing about with all the other Collies, many of whom have preceded him in death. Free from the burden of waking reality, he was alive in his memories of youthful exuberance.
It was then my mind recalled the time years ago I happened to call Richard Smith, developer of the Auricap, to place an order. Beyond the business at hand, we always spent a few minutes extra on the phone, chatting about the current situations of the day, or interesting events in our respective past. He knew I lived with a houseful of Collies, and on one particular conversation he recounted a time from many years before. Long before Richard got in to the high-end audio parts business, he worked for a telephone company installing lines and phones. One afternoon he arrived at a modest home recently occupied by an old man, and needing a phone installed. Upon entering and setting about his task, he was greeted by 3 gentle Collies, tails a-wagging, and full of joy. After finishing up, he commented to the old fellow how much his Collies reminded him of Lassie.
"Well my boy, they should...these ARE Lassie!" It was the retirement home of Rud Wetherwax, the breeder of all the dogs we have known over the years as the TV and movie star, Lassie.
We lost Richard a few years ago after a long battle with cancer, but we did not lose the fruits of his labors. John McDonald had partnered up with Richard as his business manager long ago, and John kept the business going: Not only the capacitors that Richard was so well known for, being one of the very first caps specifically designed and marketed for musical qualities, but also speakers, cables, and even some AC conditioning equipment. I wasn't the first to use Richards's caps, but as I went on to develop and manufacture some products that needed caps, we used, and continue to use, lots of them. It was back in those days that I was also a regular contributor to Positive Feedback, the real printed-on-paper magazine… not the website based magazine you are now reading... I ended up writing a review on those caps. Even now, many years later, I occasionally come across a reference to it.
Back in those days, very few parts were actually being designed and built with sonic qualities in mind, so as they say, "The Pickins were Slim." You could choose from the small number of special audio-focused component part products, or choose from the vast array of standard commercial parts, or a very tiny selection of hyper- expensive aerospace grade parts. My, how times have changed over these passing years. Many new companies have come to the trough to feed upon the audio specialist market, and some of the old companies have stayed with us, developing new and better-sounding parts.
Such was the case at Auricap, now under the new banner of Audience. I was most taken by the really superb Teflon™ caps that they started to produce for the cost-no-object end of the business. I have used these with great results in the tweeter circuits of some speakers we started to build a few years ago. However, at the affordable end of the range, we had seen little change to the regular line of Auricaps. They remained exceptionally reliable, even in high stress AC power-line filtration applications, but from a sonic perspective, the rest of the industry had moved forward. Recognizing this, John set about to re-engineer and improve the caps that had served the company and its customers so well for so long. The result was the Auricap XO.
At first glance it looks a lot like the old product line, a yellow wrapper with green epoxy end caps, but upon closer inspection you will notice double wires at each end... and internally, well, let's just say that John and I have chatted a bit on the phone about new methods of doing just about everything there is to winding a cap. Was all the effort to develop a new way of doing things really worth it?
In a word, "YES!"
And how do they sound? In two words, "Much better!"
OK, that's not really fair to boil it down to such a simple reduction... after all I've been living with these things for half a year, trying them out in all sorts of applications.
First, let's set some criteria...we are going to be discussing parts that depending on capacitance value and voltage, mostly sell for under 50 bucks each, the sort of thing that most hobby DIY people can afford, and within the budgets of most of the higher end equipment builders. We are going to set aside the super caps such as the top Jensens, Duelands, V-Caps, and even the Auri-T, as these are all very expensive, very good sounding, and for the price, one would expect that. These are the dream caps, sort of analogous to cars like Bugatti, McLaren, Ferrari, and Lamborghini. As much as there is a place for these great products, we can't always afford to be using them... no... we can rarely if ever afford to be using them. Hence, the need for good parts at the lower pricing structures, where higher production volume and larger market can make things happen. The tough part is designing these more affordable parts. It's really a LOT harder to design a decent lower-cost part than it is to do a cost-no-object part. It really requires the designer to bring all the tricks of the trade to the table, and all their knowledge of what affects what. If you can't just sit down and select the best of the best, you have to have knowledge and experience to know HOW to make compromises that include selection of vastly cheaper materials and affordable production methods. It ain't easy....
First… break in the caps.
I soldered all the caps together in parallel and hooked them up to signal generator. I set the device to apply about 20 volts of audio, in a sweeping frequency. Every so often over the course of the next 10 days, I would re-set the frequency range and waveform, so the caps got hit with sine wave energy all the way from below 20Hz to above 150kHz. There, that should do the job after 10 days of that stress!
The first application I tried the XO caps in was as a simple high pass filter to drive a tweeter. In many respects, this is a tough duty for a cap, far more difficult than a quick glance might reveal. First of all, all caps seem to have better performance when they have a polarizing voltage across the two ends, such as when you are using a cap in a DC power supply filter, or as a DC blocking-interstage coupling device in a tube amplifier. In a speaker we have no DC polarizing voltage present [OK, you can do some trick stuff with batteries and more caps and high-value resistors to create a polarization, but we're dealing with the real world here.] In a speaker, you hear the real sound of the cap as it handles current and voltage, and does so in the most difficult way. Everyone that knows me well knows I'm sort of a nut for ribbon tweeters, and I have a small selection of them to play with.
Raven R3, updated
Arum Cantus G3si
I set up a system by which I could exchange any of these drive units for one another, each one being impedance matched and level matched to the other. This allowed me to try a number of caps with each driver; all that was needed was a simple change of caps, all of which needed of course to be of the same value. I choose 6uF for my tests, because it gave me a nice crossover frequency at 6dB per octave, safe for all of the drivers I would be listening to caps with. Lows were provided by a proprietary 6 inch mid-woofer in a slot-loaded enclosure, a driver and cabinet design I did for a client a few years ago.
I spent a day listening to assorted caps and drivers, going back and forth with an assortment of tunes, to work out the general ranking of how the caps sounded. Comparing the XO to the standard Auricap, the highs are clearer, more open in the illusionary 3D sound-stage that spreads before me. Less distortion in passive parts will allow the passing of the finer details of a complex signal. These fine details are what our brain uses to make sense of the dense sound around our heads, to create the spatial acoustical representation in our heads. The dynamics are there too, a bit more snappy, not as held back on cymbals and percussive edges. I also sense no increase in edginess or graininess like I have heard in some other caps, caps I don't use, even if they are bargains and well distributed. Please, let's just not mention those crummy ones! One of the sleeper—that is to say, affordable-but-good-sounding—caps in recent years is the Obbligato caps, out of the Orient. I tried some Coppers in this range of listening test too, as I have some around and sometimes use them in client projects. They were overall, and in a nutshell, just a bit more open and clear than the regular Auricaps, something I already knew from previous experience They are one of the new generation of more recently designed parts that sounds fairly good and doesn't cost an arm and leg. The new XO? One or two words, better. More of everything; not revolutionary, but enough to be completely audible in a hurry, and also satisfying. It did not matter which tweeter I was using, as it turned out they all had the same changes in how the music was presented.
I'm not going to go into a list of CDs, records, SACDs, or the music on my DSD recorder from live symphonic venues I recorded, and then tell you a lot about how each track had so and so a change. That is mostly not really necessary, in my humble opinion. It makes good reading to fill the screen or page, but you likely have different music, and you want to know the general changes that are probably going to be pretty universal no matter the tune being played, so I will mostly just confine my impressions to how things do in a general sense.
Not to stop with listening to what these caps do in a speaker, and focusing on only the highs and upper mids of the speaker, I next installed them into one of my tube pre-amps, to function as the only cap in the signal path, the output stage DC blocking cap. We have a nice polarizing voltage here of about 130 volts. This is a good thing for any cap of sufficient rating, and we get to hear the whole spectrum of sound, top to bottom. I fired up some SACDs and my main speaker system, consisting of 40 inch Fibbonacci ribbons and 16 inch Great Plains Audio-Altec 515s.
The main amps for the big ribbons are huge MA-2 OTLs, but they are undergoing some re-work again, so I used some of my of my other amps, one solid state, and one tube, just to hear how things would do with different flavors of amps.
For solid state, we have a Coda that incorporates a whole lot of really good sounding j-fets and mos-fets. Representing the vacuum state, in the other corner, is the BAT VK60. It has become one of my favorite glow-in-the-dark amplifiers. Yes, it is not the absolute top in every degree of transparency, but it is always completely musical and emotional in its glowing heart, and completely enjoyable. Bass below 185Hz was handled by my usual powerhouse, an old Classe' 250 watt per channel box. Sure, it's getting to be an almost vintage item now, but it still kicks out the current to keep the woofers moving just so, and it sounds just fine as long it's just 'down there' where it works OK... so down there, in the bass, it stays...
In the preamp, first I listened to a 1980s vintage array of Component Research Mil-Spec/aerospace Teflon TM caps, left over from the days before there were really good audio specialist caps. These were the cat's meow back then, and at a few hundred bucks a pop, not too many people ever got to hear them.
As our friends across the pond would say, "Right!"... the sound was familiar, because I had these in here on and off for somewhere in the range of 2 decades, and I KNOW what they sound like. Next was some old regular Auricaps, and then, the new XOs... Wow, what fun it was to hear the full range, because now we can hear the bottom end of things, too.
I recalled from memory, the results of not too long ago instilling the Teflon TM caps by the same folks at Auricap, and how good they were. Fantastic really... I had mine wrapped with Cascade Audio Engineering VB-2 vibration damping sheet. Clear and open, fast and wide, almost like no cap at all. I couldn't just leave the caps in the preamp, as they were borrowed from some of my other speakers! So, out they came from the preamp, and back into the crossovers I stole them from. Into the preamp went my old standbys, the Component Research parts. Immediately apparent was how much progress had been made in the 30 years that separated the design and manufacturing of these parts, each representing a pinnacle of performance in its day. I got a little bit of congestion back with the Component Research parts, a little less dynamics, a smaller sound stage, but it was at least a known and somewhat comfortable sound, free of obnoxious additions or obvious subtractions. I would still hear clearly the low-level accent strings buried in the far distant right sound field on one of K.D. Lang's tunes, and Lauri Anderson still had the textural and visceral excitement visiting the 'Body Shop'. My few dozen favorite reference SACDs still all sounded transparent and liquid, but just a little less so.
I had to know where we were acoustically speaking, in regard the previous standard Auricap and the new XO, so I dug some up regular Auricaps from my bins and soldered them in place, after an ordeal on the same break-in rig. We still use these in our AC power conditioners with great success, and we keep a lot of them in stock. In this application, they are not passing music, just dealing with noise on the AC line, and I must say that with over a thousand AC conditioners built by JENA LABS, we have NEVER had a capacitor failure. This is an incredible testimonial as to the reliability and robustness of the design… but I stray… we are talking about how they sound, not how tough they are. Instantly I could recall how good they were back in the day, well over a decade ago. I also could hear how they have been surpassed over time by other designers' products, and the Auri-T from the same company. Some of the new parts by other companies even approach the affordability of the original Auricap and sound pretty darn good. I remembered why we spent the huge sums to get aerospace parts like the Component Research and Tektronix capacitors, as there were just not any standard commercial parts as good as these high-priced and rare caps. These exotic bits of spacecraft grade, and instrumentation-grade caps could not be used everywhere because of the exorbitant cost; we simply could not afford to use them whenever a cap was needed, and back then the Auricap was the stand out leader among affordable parts.
OK, memory lane being walked upon, it was time to slap in the new XO... HELLO, new century!! Could it be? Yes it was...these humble and affordable parts started to sing with only the break-in from the generator, I was flabbergasted at how close the sound became to the much more expensive Auri-T Teflon™.
I know that this is an audio write-up, not a test set results blurb; for those of you that have an interest in such things, I took one of the XOs and installed it in my Hewlett Packard LCR analyzer. The humble XO, it performs pretty good! These are the kind of performance numbers I usually see with parts costing much more.
XO Measurements: Digression and Clarification
Our choice of test instruments for measuring capacitors and inductors are the Hewlet Packard 4274A and 4275A. These were designed as absolute laboratory grade instruments that sold new back in the day for about $10,000 each. They differ from each other in the frequency range that they cover, and by having both of them, we can test components from 100Hz all the way up to 100MHz. They are a microprocessor based impedance measuring instrument that evaluates and quantifies LCR components, complex components, and even semiconductor materials.
The meters measure equivalent series resistance (ESR), impedance (|Z|), phase angle (Theta), reactance (X), susceptance (B), and conductance (G), in addition to the conventional L, C, R, D and Q parameters. They measures only the value of the component and/or device under test, with 5 1/2-digit resolution and 0.1% basic accuracy by reducing the possibility of errors due to self or mutual inductance, stray capacitance and/or residual inductance in the test leads or test fixture used. In practice, we have found the actual accuracy, determined by measuring actual National Bureau of Standards traced, test and calibration "standard" parts of known exceptional accuracy, they are often much better than 0.1%. In some ranges we have noted accuracy as high as 0.013%—about 8 times better than specified. Measurements are obtained by a four terminal Kelvin-style topology, and a built-in automatic zero offset will compensate for any test fixture or lead-in errors for both open and shorted conditions. A self-test provides an automatic operational verification check indicating a pass or fail condition for the instrument as a whole. It will tell you if something is wrong internally, and even report the internal section that is out of spec, needing service or recalibration.
Subsequent models by HP and other makers are faster, and some models area actually much easier to use in automated test stations, but nothing made since these fine machines has the accuracy and precision of these models. Even today, test and calibration labs worldwide keep their 4274As and 4275As working and in daily use, as nothing made since, seemingly, will do the job as well as they do. I trust mine and use them for every passive component evaluation we do.
I tested an XO with single lead wires, rated at 1.0uF and 200 volts for its basic characteristics, from 100Hz to 100kHz. The rated value was 1.0uF and the measured value was 1.02844 uF.
I then tested a pair of XOs with dual lead wires, rated at 6.0uF and 200 volts for their basic characteristics, from 100Hz to 100KHz. The measured values were 6.0220uF and 5.9750uF. I then measured the 6.0220uF example for typical characteristics.
If you happen to have had some electronics engineering training, these numbers will be of value to you. If not, what you need to know is the D rating, dissipation, is the most important for audio, and the smaller the number, the better. It represents, in essence, how much of the energy that is passing through the cap, gets 'stuck' there and is released later over time, albeit a short time. These figures are pretty darn good. Most film caps, of either metalized film or thin metal foil construction will have numbers that range from 1.5 to 6 times what we seen here. This means the XOs will not create much audio smearing effect. These numbers are close to the values we more often see with really expensive parts. The Q factor is a little harder to define, as it as to do with the ratio of pure capacitance to the other characteristics of the part, such as series resistance and its total inductance. In essence, the larger the number, the more pure the character of the parts intended functional capacitance. Once again, these numbers are a little better than the average we see, and more typical of much more expensive, exotic materials caps.
I took other measurements as well, but did not bother to record them for this article, as they are rather technical in nature, and in the end, are simply supportive of the results we did report here.
I really like this new series of parts from Audience. They offer a refreshing update to an old reliable. Everything we care about in the music is better. The 3D recovery is more abundant, the clarity of fine details is improved. The bass is strong and smooth, not over-emphasized or woolly-warm like some other parts. The highs are sweet and not fatiguing. I like them, especially at the modest price.
Even in the face of all this newness, is the silhouette of an old familiar sound. The XO is, in my view, an evolution, not a revolution...and a nice evolution it is. Richard WOULD approve, were he still here to take a listen for himself.
I have some new speaker projects in the wings, and the XO will be THE specified parts. A pair of 1960s era AR-10s are arriving soon for 21st century updating... XOs will be installed all around, along with new wire and improved internal damping. We are going to be using a lot of these in the coming years...
The old Collie woke from his slumber, elongated his legs in a well-enjoyed stretch, and then struggled to his feet. Rear legs wobbling, but his tail a-wagging, he headed over to me and slowly placed his long white haired muzzle on my leg, awaiting my hand to rub his head. He had quite a dream there, revisiting his romping jumping running days of old... and I wondered if he could sense that I too had just had a memory of old times, of chatting with Richard about some other happy Collies, his "accidental" visit to the "Lassies."
It's been many years since I was a regular in the then-physical "pages" of Positive Feedback magazine, and it's been kinda' fun putting some thoughts together to once again share with the readers of what Positive Feedback has evolved into, the E-Zine of the modern internet world. I'll drop in again from time to time to make some observations, perhaps to give an opinion on a product, or to suggest some music or a film....
See you here again!