In 2013, not long before our daughter was born, my wife Sabina and I went out to search for a pair of loudspeakers for us to restart our journey in the appreciation of music. When it came to audiophile pleasures, we were both out of practice and felt unsure of what to aim for. Visiting some local HiFi showrooms around Frankfurt, we finally decided that we did not understand enough about loudspeakers to make a permanent decision. However, we did feel some attraction towards Canton speakers that outperformed the competition in terms of clarity. We finally settled on a pair of used Canton Vento 890 DC that we bought from a vendor in Günsburg, Bavaria, unaware that the company was actually located very close to our home in Frankfurt am Main.

Canton was our entry ticket to audiophile listening, even before we fully grasped what this might mean. The 890 DC were revealing speakers that could show the merits and flaws of a system at an instance. They led to us gradually improving our front-end from the music source, via the preamp and amplifier, to the cables connecting the devices, until we noticed a flaw in the Ventos themselves: No matter how hard we tried, we could not get these speakers to sound tonally balanced. The treble, and perhaps also the upper midrange, simply sounded too bright to be natural and did not match the otherwise commendable sophistication of the speakers. Listening to the Ventos could easily become tiring because of this effect and we ended up selling them.

Exactly ten years had passed since our original purchase, and I found myself listening to a pair of Canton speakers once again. Instead of being large floor-standers, however, the GLX 100 were small and inconspicuous bookshelf speakers. They were finished in modest anthracite rather than our Vento’s shiny silver, and the soft fabric dome tweeters made me hopeful that the resulting sound would be less harsh than I had previously experienced with or 890 DCs. And I was sure to give the GLX 100 bookshelf speakers the royal treatment by putting them on our MDF loudspeaker stands and taking good care when positioning them in the room to get the tonal balance right. The GLX 100 had been a gift to my daughter from her great aunt, and I was glad to have the opportunity of auditioning them here first.

As usual, the Cantons opened with a wide soundstage and even gave the impression of proper stage depth when placed a little deeper into the room. Their depiction of female vocals was penetratingly transparent, thus creating an instantly engaging experience that kept the ears at high alert. I also noticed a raspiness in the upper midrange that I had trouble identifying. It could be that where was some competition between the midrange and the tweeter that was causing it. Just to be sure, I removed the front metal grille which, to my ears, made matters worse. These speakers had obviously been designed for the grille to be in place. To achieve greater tonal balance, I moved the GLX 100 closer to the front wall. This gave them improved bass foundation, however, it did not lessen the energy of the treble which proved to be too much for my ears.

I reluctantly turned on the tone controls on our preamplifier and dialled back the treble by 2-3 dB. This was my last resort when speakers were obviously off balance. The silvery and over-accentuated treble subsided and revealed a pretty decent speaker underneath. Without this measure, the GLX 100 were obvious candidates for ear fatigue, if they were driven with a semi-revealing front end. Engineered for the point of sale, the Cantons would have dazzled shoppers in the 1980s and have enjoyed the ride home more often than they deserved. There was no question that Canton knew how to build great speakers for those who had the money and interest in learning about them, but I could easily see that they also knew how to sell speakers to those still new in this territory.

Specifications

• Type: 3-way bookshelf speaker
• Principle: closed, with dynamic drivers
• Frequency range: 28 - 30,000 Hz
• Crossover frequencies: 800 / 5000 Hz
• Low frequency slope: 12 dB per octave 
• Midrange slope: 6 dB per octave
• Tweeter: 20 mm, fabric dome tweeter
• Midrange driver: 28 mm, fabric dome tweeter
• Woofer: 220 mm, paper cone
• Nominal impedance: 8 Ohm
• Power handling (RMS): 65 Watt
• Resonant frequency: 69 Hz
• Terminal: spring-secured for 2mm cable
• Number of terminals: single wire
• Dimensions: (W) 245mm x (H) 370mm x (D) 215 mm
• Country of manufacture: Germany
• Weight: 5.5 kg
• Year(s): 1983

This rather pretty looking Dual CV 1260 integrated amplifier was handed down to my daughter from her recently deceased great aunt. It came along with a Dual CS 630Q direct drive turntable, a Dual CT 1260 tuner that was connected via 5-DIN plug, and a Denon DCD 660 CD player. There was even a Dual C 816 tape deck that I had some trouble finding worthwhile cassettes for in these days. Two Canton GLX 100 bookshelf speakers served to wrap up the mid-Fi ensemble. Before surrendering these new arrivals at our house to our 9-year-old daughter, I had to test them and make sure the were safe to use, of course. Any publication of these test results is purely incidental, of course.

I must confess that there was something exciting and liberating about reviewing gear at this price level. After all, unless developers we asked to add intentional flaws to their designs, there was no reason why lower-priced gear could not sound as good or even outperform higher priced contenders. Proper acoustics had very little to do with the price of the components and was more reliant on the ingenuity of its designers. The speaker manufacturers Elac and KEF had demonstrated numerous times that audiophile listening pleasures were possible even at modest price points, and examples of purposefully poor designs were not difficult to be found in the industry. Rotel, for instance, had given their RC-9608X preamplifier a plastic floor board that sacrificed the pharadeic cage that the full metal enclosure would otherwise have offered. Manufacturers need to make sure that their devices perform according to the relative position that is reserved for them. 

Therefore, I did not know what to expect of Dual’s 1984 amplifier. For one thing, it had a larger brother, the CV 1460 Class-A amplifier of the same year that offered 95 watts per channel into an 8 Ohm load. And then, both devices were no longer built by Dual in Germany, but rather by the Japanese manufacturer Denon. As I unwrapped the 1260 from its protective bubble foil, I noticed how positively sturdy and heavy the unit felt. At just over seven kilos, the moderate-sized amplifier did not feel puny at all. Turning the Dual on for the first time, I was positively surprised by the elegant look of the softly illuminated output level meters. I made made a mental note to capture these in my photos of the unit. I did not want to test the CV 1260 in combination with the Canton speakers, and so I hooked up our Epicure 3.0, half expecting the little amp to fail miserably.

I began my listening explorations with Carla Bruni’s 2014 live album A L’Olympia that I played on our trusted Denon DCD 1420 CD player. Our Epicure 3.0 speakers were connected via 2m runs of Belden 9497 as single wire. As both the speakers and the Denon amplifier had spring-loaded binding posts for small diameter speaker bare wire, I had to clip off my usual banana plug terminations. The idea that speaker cables need to be terminated to be of higher quality is relatively modern, of course. In reality, fewer material transitions and less mass in the signal path should have a positive effect on sound. The tinned Belden snapped into place, and, as these cables had already seen many hours of use, the resulting sound was pleasant and balanced from the very first minute. It was the instantly familiar sound signature of excellent receivers, such as the Harman Kardon 730, that surprised me most about the Dual. 

In combination with the Epicure 3.0, the midrange sounded warm and alluring. The soundstage was clean and well-ordered, and there was a surprising amount of space around each music event, especially when considering the relatively moderate power of approximately 2x 90 watts into 4 Ohms. Rated at 44 watts power consumption at idle and labelled as Class-A amplifier, it was safe to assume that average household listening volumes were served by pure Class-A transistor power. This certainly showed in the music performance. While the Dual presented a realistic amount of detail, the top-end never felt detached or edgy. There was wonderful cohesion and flow in the music, although the dynamics did not quite reach the level of a Luxman L-10. 

Bass performance was in balance with the rest of the music. It was not as tight-fisted as on the Luxman L-10 and also not as boomy as on the Pioneer SX 850. Complex bass sounds were better layered than on the Harman 730. The more I listened, the more I fell in love with the abilities of this integrate amplifier, and as the sun set, I started to enjoy the illuminated front with the level meters waving consistently at under 1 watt. The piano keys on Carla Bruni’s album sounded deliberate and simmered naturally, and I was once again reminded how good the Epicure speakers really were, even when powered by an inconspicuous mid-Fi Dual amplifier from the early 1980s such as this one. What a delightful discovery this was. My daughter would be spoiled from the very start by beginning her journey into music with this Dual.

Specifications

• Type: integrated solid-state amplifier
• Principle: Class-A amplification
• Output power (RMS): 2x 60 watts
• Frequency range: 10 - 45,000 Hz
• Signal-to-noise ratio (line): 88 dB
• Signal-to-noise ratio (phono MM): 68 dB
• Channel separation: input 70 dB / output 55 dB
• Signal damping: >70
• Number of audio inputs: 5
• Phono (cinch): 2.5 mV / 47 kOhm (MM)
• Tuner (cinch): 200 mV / 47 kOhm
• Tape1 (DIN): 200 mV / 47 kOhm
• Tape2 (cinch): 200 mV / 47 kOhm
• Monitor (cinch): 200 mV / 22 kOhm
• Number of audio outputs: 3
• Rec-Out Tape 1 (DIN) / Tape 2 (Cinch, Ri 470 Ohm)
• Headphone socket (jack-type): 6.3 mm
• Tone control: bass and treble (+/- 6 dB), loudness
• Bypass of the tone controls: not available
• Filter: subsonic
• Loudspeaker terminals: two (8-16 Ohm)
• Power consumption: 680 watts (max.), 44 watts idling
• Features: two combined VU meters, illuminated
• Dimensions: (W) 440mm x (H) 97mm x (D) 265mm
• Weight: 7.1 kg
• Country of manufacture: Japan (Denon)
• Year(s): 1983 - 1985

The CDP 502 ES was one of the early Sony CD players. It stemmed from a time in which engineers were given enormous budgets to derive what was humanly possible from their new digital audio medium. And, although Sony’s CDP 101 had been the first CD player to enter the world market, in early 1982, the company knew that they were not alone in the race. To share R&D costs and spread the risk, Sony had developed the technology in a joint venture with Philips, who were the first to launch their CD-100 in Europe in that same year. Marantz, Denon, and other manufacturers were soon to follow, and Sony had to innovate quickly in order to stay abreast of the competition. Launched in 1984, just two years after the CDP 101, the CDP 502 ES already showed many improvements over the company's first designs.

I received my test specimen from Luigi, who had purchased it from a collector along with some other audio devices. At first glance, there were no clues to betray the CDP 502 ES’s exceptionally high quality. To me, it looked like yet another CD player of the early 1980s. Black-box design with front-loading tray, lots of square buttons, an inconspicuous display, new infrared remote sensor, and headphone/output volume control. There really was nothing special in the design, until I tried lifting the Sony off the table where Luigi had placed it. Not only was it much heavier than the standard gear sold, it also seemed more rigid in its construction. I wondered, if Sony had also glued led sheets under the unit’s top plate, as I had previously seen on the Denon DCD 1500 II, however, a swift look under the hood revealed more substantial reasons for CDP 502 ES’s weight.

The front-loading drive mechanism was constructed out of machined metal blocks, instead of the already more common plastic. There was a large transformer with separate outputs for the digital and analog audio sections, as well as for the front display, and the drive and loading section. The stabilising power supply capacitors looked large enough even to serve a small amplifier. There was a copper-plated metal sheet to separate the drive unit and power supply from the audio signal section, and another sheet to separate the digital processing unit from the analog output stage. Although this was not instantly visible from the outside, the Sony engineers had taken many precautions to take audio performance to the next level. Due to its massive components and quality loading mechanism, the CDP 502 ES had the touch-and-feel of professional studio gear.

Although it was difficult for any CD player to compete with the cast-iron transport of the Philips CD 104 with sophisticated swing-arm laser paired with Rodenstock glass lenses, I must say that I preferred the smoother and more elegant feel of this Sony tray. The simple act of placing the disc on the tray felt very reassuring. I could also see how the understated design might have even highlighted this effect. There was no detectable time lag in closing the tray, and the unit offered ultra-fast pre-scan and disc access. Button commands were prompted by instant relay-switching feedback that was still audible from as far as 4 meters (!) distance. When the CDP 502 ES was in operation, one of three back-lit buttons served to indicate the play mode (Program, Shuffle, and Continuous).

The Sony offered two chinch/RCA outputs, of which one was variable via the Line Out/Headphone volume control. This feature allowed for the CD player to be integrated in a system with components of various audio output levels. I would have liked to see a digital S/PDIF output socket for added versatility and also welcomed a display-off function for additional noise reduction. However, as I had not been given the remote control along with the player, I remained unsure, if the display-off function had at all been available there. The headphone jack was of the larger 3/4-inch type, which was also my personal preference. Playback accuracy was said to be excellent due to the use of a new unilinear converter system that performed all digital signal processing in synchronisation with a single clock.

Sony proclaimed the CDP 502 ES to stand for a new generation of reference machines, and audiophiles have commented on the player’s ability to reproduce lots of musical detail while preserving the tonal warmth that was often associated with analogue equipment. Sony also pioneered in the development of phase-correct analog and digital filters, as well as in advanced error correction that were all incorporated in the 502 design. The analog stage and output amplifiers were of a double-mono design and used Sony’s newly developed OP-amps. In many ways, the CDP 502 ES incorporated the best of CD player technology that was available at the time, and I was eager to find out how the 40-year-old unit would perform.

I hooked the CDP 502 ES up to our main system consisting of a Dynaco PAS-4 tube preamplifier and Hafler 280 XL Linear MOSFET power amplifier. The speakers were a pair of Martin Logan SL-3. The Martin Logans’s electrostatic panels were capable of reproducing even the most subtle nuances, and the Dynaco preamplifier served to assure that the overall impression was tonally rich and not overly analytical. The Hafler amp offered approximately 250 watts per channel into a 4 Ohms load and showed great low-resistance capabilities down to 1-Ohm loads. It was also known for its ultra-flat amplification well beyond the audible spectrum. Our usual CD player in this position was a Rega Planet 2000, of which it was also said that it offered an analogue-like sound.

For my listening tests, I used an assortment of Jazz and Vocal Jazz CDs, Rock, and even some Hip Hop. I wanted to find out how the CDP 502 ES would perform over a wide range of styles. From what I had seen on its circuit boards, I could imagine that the Sony would pack a punch, and I was not disappointed: The music was fast and loud with slight upper-bass emphasis. Using the fixed output cinch/RCA sockets, there was massive gain right from the very first second, and the music seemed so punchy that it bordered on overly potent. And yet, there was also an element of tight-fisted control that somehow reminded me of Denon’s DCD 1500 II, a CD player that had come out two years later. 

Both the Denon and the Sony were High End players in their time, and both weighed around 9kg. A lot had been done to reign in high-frequency distortion on both of them. As a result, both machines seemed to march through the songs rather than swing though them. The resulting sound felt a little lifeless at times, and I would have wished for more freedom for the music to unfold and for individual notes to be allowed to simmer longer. The strong upper-bass presence could serve to give vocals a little more substance, which was a welcome effect. On the other hand, faster passages could at times sound a bit cluttered or congested because of it. 

Piano keys did not simmer as long or resonate as low as they did on our 30-years younger Rega CD player. Low-bass extension was generally good on the Sony but not great. Despite Sony’s lead in error correction, the CDP 502 ES did not perform too well on dirty or scratched CDs. For the music to sound at its best, the discs needed to be in excellent condition. I could confirm that the 1984 Sony provided an analog and non-analytical sound experience on our system. The CDP 502 ES sounded fuller, richer, and darker than many of its competitors. Due to excessive filtering, however, some of the intrinsic flow and musicality of the performance was lost, which also took away from the believability of the music event. 

Finally, I could not help but wonder, if there were available modifications for this player that would remove some of the analog and digital filter stages to uncover the beauty and power of the diamond underneath. Considering what Roman Groß ‘New Perspectives on Sound’ and ‘KR High End Laboratory’ had accomplished with the Philips CD 104, it was only fair to assume that the CDP 502 ES also had the potential of blowing most modern CD players out of the water, simply because of the impressive components that were used inside.

Specifications

• Type: Compact Disk player
• Laser principle: GaAlAs double hetero diode
• Disk revolutions per minute: 500 ~ 200 (CLV)
• Paying speed: 1.2 - 1.4 m/s
• Error correction: Sony Super Strategy
• DAC compound: 16-bit, straight line
• Frequency range: 2 - 20,000 Hz (+/- 0.3 dB)
• Harmonic distortion: < 0.0025% (1 kHz)
• Dynamic range: < 96 dB
• Wow and flutter: below measurable limits
• Line output: 2 Volts
• Headphone output: 28 mw (32 Ohms) 
• Power consumption: 16 watts
• Dimensions: (W) 430mm x (H) 80mm x (D) 335mm
• Weight: 8.5 kg
• Accessories: Wireless Remote, RM-D502
• Country of manufacture: Japan
• Year(s): 1984-1987

Sintron Distribution GmbH sells a whole range of affordable gear under the name of Dynavox to target entry-level audiophiles. Among the many worthwhile components, there is the VR-70E II tube amplifier that, with some minor amount of tweaking, performed very well in our listening setup. The Dynavox brand name can also be found on audio accessories such as cables, plugs, and power distributors both with and without filter technology. The Dynavox X7000, which lent its components to the construction of the High-End network filter that is the subject of discussion here, had most likely been a capable power conditioner to begin with. However, the aim of this construction project was to take filtration technology to the next level.

My first exposure to the High-End network filter project was on one of my routine visits to my trusted audio technician. As I walked into his shack, I could see him applying a string of ferrite beads to a freshly laid power cable inside the Dynavox X7000’s housing. I was surprised by the amount of free space inside the large X7000 cabinet, and Winfried explained that the original filter design had only taken up about a hand’s breadth of room inside the 43cm wide enclosure. And yet, the unit was large enough to hold far more sophisticated filters and even allowed for the use of isolating transformers. Winfried further explained that Dynavox’s use of high-quality materials on the cabinet, sockets, and switches made it the ideal basis for his new construction.

This must have been in early 2020, and it was not until two years later that I asked Winfried what had become of his High-End network filter design. He seemed pleased about my asking and offered to lend it to me for a few days so that I could find out for myself. He explained that he had galvanically separated the audio circuit from the household grid through isolating transformers allowing for 300 watts of throughput power. When seen from the rear of the unit, looking from left to right, there were two power sockets filtered for analog music sources, two power sockets intended for digital music sources, and two sockets without filters. On the power input side, a number of ferrite beads served to prevent high frequency noise from passing through the interior circuits.

Winfried’s new power filter stages were designed to target the anharmonic frequencies of 50 Hz. These were at 150 Hz, 250 Hz, 350 Hz, 450 Hz, etc. The filters became more effective the higher the anharmonic frequency was. Distortion measurements conducted by an independent laboratory found the filter strength to be 60 dB at 1,000 Hz and 90 dB at 10,000 Hz. These were very good results for any power conditioner. In addition to this, Winfried’s High-End network filter provided DC protection through a combination of bridge rectifier and capacitor and also maintained the original voltage control display that could be switched off for maximum noise protection. The touch and feel of the finished unit was great. It felt as if it had been carved out of a solid piece of metal, an effect that was supported by its weight.

Back in our upstairs listening room, I connected the filter to the same wall outlet that had already proven itself as being relatively free of distortion on our existing system. As we lived in an apartment building that was a home to four families and located in a mixed residential and industrial area, all observations on the electrical noise floor were relative and subject to change, depending on the day of the week, the time of day, and the number of other household devices that were active on the grid. In addition to this, our house was located within a five kilometre radius from the airport and about 800m beeline from the local tram and train lines. On clear days, we also had unobstructed view of the radio tower on Feldberg and the airport’s revolving radar with visual signal light—not to mention the many cellphone antennas that were perched on nearby buildings. 

To filter out the cacophony of electric noises on our power grid was the proclaimed job of Winfried’s network filter. Where, if not here, could a device of this kind prove its merits? — I connected our Rotel RC-960BX preamplifier to the power output that was designed for analog devices and our Denon DCD-1420 CD player to the power output for digital devices. Both of the sources were considered solid entry-level components. I then connected my custom-built Echle LF-3519 solid state amplifier directly to the wall outlet and warmed up the system. The first thing I heard was a mild buzzing sound emanating from the filter’s large transformers, followed by the by now familiar noise floor from the high-gain amplifier. I was a little disappointed that the amplifier’s noise floor did not drop significantly with the filter attached. Sadly, this did not change even when I connected the amplifier directly to the unfiltered output of Winfried’s power conditioner.

I chose a-ha’s 2017 MTV Unplugged album “Summer Solstice” as listening sample, as it was a solid live recording with natural instruments, spoken and sung male and female vocals, and a small venue audience that could sound believable in our listening room. Songs like ‘This is our home’ provided a bass drum impression that allowed for a good sense of proportion, attack, and decay. Listening with the filter in place, I noticed that stereo imaging was much improved. ‘This is our home’ begins with a spoken introduction in which the first speaker is sitting further to the left. Without the filter, the two musicians appeared to be speaking from the same position. There also seemed to be more space around the instruments with the sound stage panning further and more seamlessly from left to right. The position of the audience seemed more realistic with fewer holes in the sound stage.

Power plug polarity had a major effect on the overall sound signature. This came as a surprise to me, as I would have expected the network filter to have a harmonising effect between the units. Without the filter in place, it was immediately clear to me which plug polarity was my preference. However, this was not so easy for me to decide with the filter in place. In one plug combination the sound became lighter and more airy and in another darker and tonally richer. I would normally have preferred the second, as it offered a stronger mid bass and a weightier and fuller sound, if it had not been for an increased amount of grain in the midrange that made vocals slightly less believable. No matter how much I tried different plug combinations, I was always left between those two choices: darker and more voluptuous with slight grain on the vocals or light and airy without the grain. But which one was correct?

Listening to music without the filter again, I noticed that tonality remained slightly shifted upward towards light, airy, and slightly sharp in the treble. Bass notes were lacking a little energy and could not free themselves so much. My preference would have been to keep the filter in place, and to set plug polarity to produce the darker tonality but without the slight sense of grain. This, however, was a result that I failed to reach on our system. I decided to report my findings back to Winfried, hoping that he could further help me in understanding the unit and the effect I had found.  — Audiophiles are a tough species to satisfy, I suppose. 

< 4. Power Filtering | Dynavox Tube Amp (Part II) >

Specifications

• Type: 240 volts power isolator and filter
• Isolation principle: galvanic separation
• Features: soft-start (to offset transformer magnetisation)
• Filter principles: low-frequency harmonics and HF noise
• Filter frequencies: 150 Hz / 250 Hz / 350 Hz / 450 Hz (anharmonic)
• Filter strength: 60 dB (@1,000 Hz); 90 dB (@10,000 Hz)
• DC-filtration principle: bridge rectifier and capacitor
• Throughput power (max.): 300 VA
• Power sockets (switched): 2x analog, 2x digital, 2x direct
• Switchable display (on/off): mains voltage input, LCD (amber)
• Cabinet basis: Dynavox X7000 Power Conditioner
• Dimensions: (W) 430mm x (H) 100mm x (D) 305mm
• Country of manufacture: Germany (custom-made)
• Weight: 18.5 kg
• Year(s): 2020 - 2023

I listened to my first pair of Epicure speakers at Luigi’s house about two years ago. As Luigi was a seasoned audiophile who had provided me with many valuable insights over the years, I was a little sceptical at first to find the small and unpretentious-looking Epicure 20+ set up in the same position where his majestic Snell A3 had been before. However, I need not have worried, because these speakers sounded great despite their awkward and smallish look, or perhaps even because of this.

I kept the name 'Epicure' in positive memory, and, in early 2022, I picked up my very own pair of Epicure speakers from Würzburg in Bavaria. I did not want to buy the same model that Luigi had found for himself and was more attracted to the larger EPI 500 speakers that were also about ten years younger than the 20+ (from the 1980s rather than the 70s). Once I had the EPI 500 at our house, it took me a few weeks to learn how to position them properly. They were particularly fussy about their coupling towards the floor. In the end, I prevailed and they sounded great.

In the very beginning of listening to the EPI 500, I noticed a slight scraping of one of the midrange drivers. It was only on some frequencies and so short that I was never sure if the noise had been signal clipping or really mechanical scraping. In the end, it did not matter much, because I knew that the ferrofluid-cooled midrange drivers were still the originals and around forty years old. The driver that I suspected of scraping even smelled of burnt resin when I held my nose really close to the cone. Before each longer listening session, I would move this particular driver by hand a few times, carefully pressing down on the lower end of the cone surround. The midrange driver, being slightly corrected in trajectory, usually worked without noise for a few hours after this.

I loved the sound of these speakers, and, from personal experience, I knew that they could compete with some very pricy modern designs. Therefore, when Luigi called me up to tell me about another Epicure set on offer, I did not have to think about it very long. A pair of Epicure 3.0 was being sold in Berlin, and I knew that Luigi had initially been planning to purchase them for himself, until, finally, his wife revealed that she did not like the strange pyramid shape at all. I consequently showed the speakers to my wife, Sabina, who gave me instant approval. With all the preliminaries being settled, I only had to work out a plan on how to arrange for the transport from Berlin to Frankfurt am Main. Luckily, I had some highly supportive friends in Berlin who helped me make this possible. Enrico did a fabulous job, and I was indeed grateful.

Despite everyone’s sincere effort, one of the Epicure 3.0 tweeter’s inverse paper domes had a slight tear down its centre that was most likely from material fatigue. While the tear was never audible to me when listening to music from my usual 3 meters listening distance, I could see it expanding with each new session. This meant I now had two Epicure speakers that I enjoyed listening to a lot, but that each had a flaw which would make my pleasure short-lived and selling them later a nearly impossible task. After all, to knowingly sell broken speakers might work for some people, but it did not work for me.

Ever since my first research on Epicure speakers, I had come across the Human Speakers company in New Hampshire. At this time, it had just been an address in the USA for me, albeit, not too far from my former US home in White Plains, New York. On the Human Speakers website, I had read that Huw Powell offered parts and services for Epicure speakers. However, I did not know how old and accurate this information was, how old Huw himself was, if he could provide shipment to Germany, and if the parts were of sufficient quality and would manage to be integrated seamlessly in the otherwise pretty flawless Epicure designs. — On the other hand, I had little choice but to take the plunge and hope to end up in good hands.

I first sent Huw an e-mail to explain about my problems with the EPI 500 midrange drivers and Epicure 3.0 tweeters and ask him for advice. He informed me that he had designed replacement drivers that would blend in with the existing fixtures and crossovers in a way as to complement the original Epicure design. If there was any difference at all, he wrote, the new components sounded ‘better’. This made me a little apprehencious of what to expect, nevertheless I placed my order using my credit card and saw that there was a reservation in the amount of my purchase soon after. The drivers were built following my order and arrived at our doorstep just two weeks later. I was relieved to hold the little box of speaker parts in my hands.

I first unpacked and installed the midrange drivers on the EPI 500. Removing the original drivers was simple. I unscrewed each driver, taking care not to damage the foam surround and foam rubber insulator underneath. The wires were simply plugged in rather than soldered in place which required minimal effort from side. When screwing in the new PRO 025 driver, I noticed that the outer ring was not reinforced with a rim, as had been the case on the originals. As a consequence, the outer edges folded downward when fastening the screws. Huw suggested to stop tightening the screws right where they started to bend the chassis. However, this position gave the drivers' chassis a metallic ringing when touched. And so I tightened the screws until the ringing was deadened by the speaker cabinet.

It is possible that later EPI 500 versions had a slightly larger cutout in which the PRO 025 would have fit even better. I guess, I could have sanded the rim of the cutout before placing the new driver, but I decided that the slight bending would not matter and that I should give the EPI 500 a good listen first. I am glad that I did, because the new midrange drivers not only blended in with the original design and sound signature, they also enhanced midrange imaging and brought the whole listening experience to a new level. Reading about sonic improvements on Huw's website was one thing, but hearing this with my own ears was another. The EPI 500 still sounded tonally rich and painted wonderful natural instruments, but there now was a new sense of order in the mix that—once heard—I would not want to miss again.

Very pleased with my first repair project, I unpacked the Epicure 3.0 tweeters. The first tweeter was in good shape and was easily installed, involving only a limited amount of soldering. The original paper inverse dome tweeters had been round in shape and had stood out like protruding eyeballs. Their new replacements were semi-circles forming an arch on top of each pyramid. Although I liked the original design and would have preferred to preserve it for nostalgia sake, I could also see that the new design was a visual improvement.

To install the new tweeter, I carefully took off the plastic ring around the midrange, made note of the wiring and unplugged the driver. This gave me free access to the underside of the tweeter mount, where I could then easily loosen the screws that held both the tweeter and the o-rings for the wiring. I cut off the o-rings and soldered the replacement wires that Huw had sent into place after leading them through the two small cabinet holes that had been left empty by the original tweeter. The new PRO 002 tweeter was placed on top of the cabinet with the back of the driver plate about 40mm from the front baffle. 

I could have easily replaced both tweeters within a single repair session, however, I was sad to see the second tweeter rolling out of the bubble film packaging in two parts. Without obvious damage to the packaging and also without any damage to the other parts inside, the face plate had come off on this tweeter. Looking at the parts, I first believed that we could fix the tweeter and even discussed this option with my trusted technician, but we soon learned that the face plate had been bumping against the voice coil during shipment which had led to the dome being dented inward.

I called Huw to explain what had happened, and he offered to send me a replacement for the broken tweeter. Two weeks later, a fully functional PRO 002 tweeter arrived in the mail, and the second Epicure 3.0 was finally repaired. Without having given the speakers a good listening test, yet, I can only say that the sound is a little brighter than it was before, especially with the treble adjustment on the speakers themselves set on flat. While the preferred tweeter level (0, -3, -6 dB) will depend on placement and listening distance, I have achieved good results dialling the treble down by 3 dB. The harder dome material has made the sound slightly more modern and less forgiving of poor recordings, but I will need to set up a proper test before I can report on final results.

[Listening test in progress...]

< Epicure EPI 500 | Epicure 3.0 >