Copper reader Adrian Wu lives in Hong Kong and has spent time in the UK and elsewhere, as you will see. He is a contributor to the Asia Audio Society website, dedicated to reference-quality sound and reproduction. In Part One (Issue 120) he shared his early audio experiences, his introduction to high-end audio and recording, and his enthusiasm for vintage gear. His journey continues here.
In my quest for ever-better sound I had purchased a pair of vintage Brook 12A amplifiers – but they came with a story. These amps use 2A3 output tubes in push-pull configuration, and were supposed to be Paul Klipsch’s favorite to drive his Klipschorns. The seller told me that the amplifiers he was selling were defective, the reason for the low price. He had bought them in a non-functioning state and had hired a technician to restore them. However, they sounded distorted even after restoration.
The original wiring of the Brook was a rat’s nest, unlike the British and German amps I had encountered. After I received the amps, I confirmed that the sound was indeed distorted. I downloaded the schematic and checked the wiring. Everything appeared to be in order. I studied the schematic carefully and noticed that the polarity of one bypass cap was reversed. The cathodes of the 2A3s are directly connected to ground via the heater transformers, which means the control grids are at negative potential. The cap that bypasses the grids to ground should therefore have the positive terminal connected to ground, instead of the negative terminal as shown on the schematic. The technician must have used the same schematic (there was only one schematic on line as far as I could find), and made the same mistake. It would have been fine if he had used non-polar electrolytics. After reversing the polarity of the capacitors, the amps sounded like magic. I wrote to the seller to give him the news, and he was not pleased!
The next question was, which components should be used to restore the Brook? I tried Jupiter wax capacitors (the original version), which was a mistake, as they didn’t do well with the heat generated by the amplifiers. There was a temptation to use American components of the same vintage as the amps, such as the Sprague Black Beauties, but I worry about the reliability of these ancient components. I ended up putting in antique Siemens paper caps, since I had the right values on hand. For resistors, I used my favorite Kiwame carbon films. These retain the tone of the carbon composition resistors while maintaining stability, and they don’t burst into flame either. For chassis wire, I removed the cheap wire the technician put in, and used new production cotton-sheathed copper wire to maintain the antique look.
The frequency response of the amps was way off. It turned out that the insulation of the input transformers has broken down after all those years. Teflon wasn’t available for transformers in those days, and they used paper insulation. The dampness in the basement had destroyed the insulation. I contacted Telefunken USA and they were very kind to agree to rewind the transformers for me using the original specifications, but with Teflon insulation. The measured performance of the amps went back to the original spec with the restored input trannies.
The amps came with the standard battleship gray faceplates of all the Telefunken studio gear of the time. I had the plates replicated but with a chrome finish, which looked more in place in a domestic setting.
I have compared the three vintage amps I own for driving my restored Quad ESL electrostatic loudspeakers and a pair of Tannoy SRM10B studio monitors. The Leak TL12.1 has a lovely midrange; the bass is a bit soft, but it sounds very engaging and musical. The Brook 12A has a more detailed sound, with more clarity and transparency, especially in the upper registers. It is really lovely for strings and vocal. As for the Telefunken V69a, these amps use pentode output tubes, whereas the Leak has triode-connected KT66 and the Brook has directly-heated triodes. The V69a have better bass, with more impact and a more solid foundation. Music takes on a larger scale. This amp is also very detailed and transparent, but sounds a bit lean when compared to the Leak. I feel they are tonally more neutral though. I would prefer the V69a for rock and large scale orchestral music, the Brook for chamber music and female vocalists, and the Leak if a warmer sound is desired.
Going back to my day to day system: after moving back to Hong Kong from the US, the conrad-johnson PV10a preamp and Aragon amplifier now needed step-up transformers. Having upgraded my front end with a Michell Orbe turntable, Graham 2.2 arm and Lyra Helikon cartridge, I wanted to improve the other parts of the system as well.
Michell Orbe turntable.I first met Tim de Paravicini at the Heathrow Penta Hi Fi show in the mid-1980s. He demonstrated his amplifiers with a Revox reel to reel tape player and stacked Quad ESLs, which was certainly unusual at the time and hard to forget. I’d kept in touch with him from time to time, and had always wanted to own his designs. During a trip to London, I went to Walrus Systems in London (now closed) and auditioned some of the EAR amplifiers. I ended up buying the 834P phono stage and the V12 integrated amplifier. The V12 was interesting as it used parallel push-pull ECC83 small-signal tubes for output! The current incarnation uses EL84s, and I have not heard it, but the original version had a rather distinct sonic signature. They actually worked rather well with the ESL, giving a very transparent, involving presentation.
As I got more into vintage audio, I found out that I could buy non-functioning or poorly functioning audio components, restore them to original specifications, and sell them for profit. Since I have more fun restoring and optimizing them than keeping them, this was a good way to sustain the hobby without the headache of finding somewhere to store the equipment. Hong Kong has a very vibrant audio scene and it is very easy to sell vintage and good-quality audio gear. For popular items, I could usually sell them within a day of placing an ad on the popular Review33 website. I was able to source equipment through classified ads abroad and this allowed me to gain knowledge through experimentation to find the best components for restoration.
I also got to know various local artisans such as transformer makers that few people knew existed. Various Leak and Pye amplifiers came and went, and my hobby was financially self-sustaining. I also started to look into turntables, specifically Garrards. My first 301 came about after I spied a Schedule 2 machine on a slate plinth (made by the now defunct Slate Audio) with an SME 3012/II arm and Clearaudio cartridge at a second hand shop in London for the grand price of 1000 pounds. The turntable had already been serviced and the whole thing was plug and play. I sold the cartridge and mounted my Lyra Helikon. It had the drive and the solidity that I felt was lacking in the Orbe. The music had more presence due to the improved dynamics. I decided this one was a keeper and sold the Orbe instead. I subsequently upgraded to a late grease-bearing model, selling the Schedule 2 to a Japanese enthusiast for a good price. As for the SME, as it had a plastic knife-edge bearing, it was a good excuse to upgrade to a bronze knife, and I rewired the arm with silver wire and added a bronze base for good measure. I stayed with this table for 15 years, only recently exchanging some of its parts for a Classic Turntable Company 301. The superior main bearing, sturdier chassis and perfectly balanced platter brought a huge improvement. The improved speed stability results in better dynamics and tonal stability, the lower noise floor manifests as better transparency, and the frequency response also became more extended. This turntable is probably the greatest bargain on the market, especially when compared to the reissue 301 from SME.
The SME reissue of the legendary Garrard 301 turntable.
I left academia after six years, having experienced the SARS epidemic while working at a public hospital. It was an experience I thought at the time I would never see again, but how wrong was I! I was also totally fed up with the politics of academia. After a few years, with the kids getting older, I thought it would be a good idea to move to a larger apartment closer to work. It was a perfect excuse to realize my long-planned project: horn loudspeakers.
As the apartment needed to be gutted and totally renovated, I engaged a friend who was an acoustical architect to design the lounge (my wife is an architect, but she was very tolerant!). This friend worked for a number of years at the Arup Group in the UK and was responsible for the design of a number of concert halls and performance venues before returning to Hong Kong to set up his design firm. He also advised the late Mr. Winston Ma (former owner of First Impression Music) during the construction of his listening room near Seattle in the late 1990s. I got to know my friend when I wanted to use a concert hall he designed for a recording session. The hall quickly gained a reputation as having the best acoustics in the territory.
After the first site visit of the new apartment, he liked what he saw, since the room was irregular, and had no parallel walls but had the correct dimensions. He designed the air conditioning system, a subject of great importance to me as AC noise has always been a problem in many recording venues, so much so that we try to avoid doing recordings in the summer. Four-inch acoustical foam was placed strategically inside the walls and the ceiling. He also designed a ceiling to break up the standing waves, a design that sent my wife into a tizzy, and which the contractor declared was impossible to build. We ended up with a compromise, a ceiling that slopes at different angles in four directions. It actually does not look so weird once we got used to it, but it always elicits some reaction from new visitors. He even designed the LP shelves on one wall to control the first reflections from the loudspeakers, with the records stored at a precise angle of 23 degrees.
The built in bass trap doubles as a storage unit (or is it the other way around?). The idea was to make it a normal living room with only subtle hints of acoustical treatment. When I check the RTA (real time analysis, a measurement of the frequency spectrum of an audio signal) from time to time, I am still amazed at the smoothness of the frequency response. And the AC is completely silent. Most importantly, being in an apartment, the room is soundproofed with a subfloor floating on rubber insulation (to isolate it from the walls, which transmit noise to other floors of the building), and the same type of doors that are used in recording studios were installed for the front entrance and the corridor leading to the bedrooms.
In the meantime, a pair of new horn speakers were planned. I had heard various horn iterations over the years in friends’ systems. Trips to Tokyo also presented opportunities to visit horn builders as well as antique audio dealers. I visited Jean Hiraga in Paris while he was still the editor of the magazine La Nouvelle Revue Du Son. He had set up at the time a pair of Altec A5 Voice of the Theatre loudspeakers, with his own crossovers, driven by Hafler solid state amps (no 300Bs!). The source was the original Philips CD player (probably modified, but my French at the time was not good enough to ask for details). Not exactly how I imagined it would be given his reputation. However, the sound was quite a revelation. Very dynamic, life-like and musical. He gave me a tube data manual (in French) as a gift. I also visited La Maison de L’Audiophile during that trip. Jean subsequently visited Hong Kong and gave me advice while I was setting up my horn system.
During a trip to LA, I visited Dr. Bruce Edgar, creator of Edgarhorn loudspeakers. He was a very kind man and full of knowledge and experience. Unfortunately, he was in poor health at the time, having just got out of hospital after a leg infection, but he still spent an afternoon with me. I finally settled on the combination of an EV (ElectroVoice) T350 tweeter, a JBL2450 midrange (picked up at a good price in Hollywood) and Altec 515C bass drivers. I had trouble finding a good pair of vintage 515C. I then found out that an outfit called Great Plains Audio was servicing Altec drivers, and they had just started to produce some new drivers. I called the owner and explained to him what I wanted, but he was a bit bemused to learn that I wanted the Alnico version. He could not understand why since he did not believe it was in any way better than the ferrite version, and it loses magnetism over time. Anyway, I convinced him that there was a market for it, and he agreed finally to produce a prototype. Months passed, and he finally contacted me, telling me that he had a pair of prototypes that performed to the original spec. I bought them, and the 515C has been a regular item on his catalogue ever since.
I had considered various high-frequency drivers such as the JBL 077 and various Fostex models, but I had been impressed with the sound of the EV T350 at a friend’s place. Pretty much the only thing that can go wrong with these drivers is the voice coil, and amazingly, EV still produces these phenolic diaphragms. I bought a pair of 400 Hz rectangular exponential horns and reflex bass cabinets from a builder (Tatematu Onko, no longer in business) in Japan, the latter designed specifically for the 515C. I decided to use active crossovers, which allows for easy adjustments, maximizes sensitivity of the speakers and avoids adding reactance. I built a three-way crossover from Marchand Electronics initially, and subsequently switched to an Accuphase F-25 analogue frequency divider.
To backtrack a bit, a few years before I started the horn project, I got to know Allen Wright. In my quest to learn more about amplifier circuits, I came across his writing on the internet. Allen was an Australian guru who started his career as a technician at Tektronix. He designed the amplifier used in oscilloscopes, which were all tube-based in those days. These amplifiers needed to be extremely quiet and linear up to the megahertz range. He then decided to tackle audio and set up his own company, Vacuum State Electronics. He was very well respected within the DIY circle, and was in high demand for doing modifications and upgrades, as well as consulting for manufacturers.
He was developing his Realtime Preamplifier at the time, and wanted beta testers to iron out problems. I became one of his 10 beta testers, and was sent the components and the chassis to build the preamp. The design is very complicated, with a phono section that has an input sensitivity of 0.1mV. It is a dual mono, balanced differential design based on the E88CC tube. It has a shunt regulated power supply. Allen had a set of principles that he steadfastly adhered to. These included:
- A fully differential circuit.
- Zero negative feedback.
- Internal wiring with the thinnest conductors (solid wire and foil), preferably in pure silver
- He advocated using the cheapest RCA plugs and sockets (the conductors in audiophile connectors are too thick), or preferably, the Lemo Redel connectors (non-magnetic connectors normally used in MRI scanners and the defense industry).
- Teflon dielectric.
- All electrical “anchor points” are tightly regulated, which means the extensive use of current sinks and current sources to achieve the highest impedance possible.
- Choke-filtered power supply with solid state rectification. Fast-recovery diodes for high voltages, Schottky diodes for low voltages.
The whole experience was an excellent learning exercise, not only in the theory of audio electronics, but also in the art of point-to-point wiring construction. It took a good two years to finalize the prototype. The preamp is very quiet, dynamic and tonally neutral. I used it as my phono preamp for several years before buying a factory built RTP-3D version. Last year, I modified the prototype to serve as my tape head preamplifier.
For power amplification, I am using two pairs of Allen’s DPA300B tube amps. These amps have a differential input stage using a pair of the Russian 6H30Pi in cascode to drive a pair of 300B power tubes in push-pull configuration. These amps serve the tweeters and mid-range drivers, while a Mark Levinson No. 27.5 amplifier drives the bass.
Back to the loudspeaker discussion: a few years ago, I was introduced to a friend who had built a speaker system using field coil drivers made by G.I.P. Laboratory in Japan. These are recreations of the ancient Western Electric drivers, and are extremely expensive. The system had very impressive dynamics, but he was still working on integrating the various drivers, and the system was not very coherent as a whole. However, what it did do well got me interested in field coils. [A field coil loudspeaker uses an electromagnet which needs to be powered by DC, as opposed to more conventional speakers that use permanent magnets. – Ed.] The GIP drivers were out of my price range, but Line Magnetic in China also produces similar drivers at a somewhat lower price level.
However, I don’t believe 80 years of advances in science and engineering could not improve upon these ancient designs. Classic Audio Loudspeakers in Brighton, Michigan produces a line of modern field coil drivers using state of the art materials such as beryllium diaphragms, and the designs are based on Altec and JBL drivers. This meant I could get drop-in replacements to use in my current set up with minimal adjustments required. After detailed discussion with John Wolff, the designer of these drivers, I bought a pair of 6475, which is based on the JBL475, the consumer version of the 2450, and a pair of 1501, based on the Altec 515.
The higher breakup frequency of the beryllium diaphragms allows me to operate the midrange drivers up to a higher frequency, and I moved the crossover frequency up an octave to 7kHz. There was an immediate and marked improvement with the new drivers. There is more detail and the dynamics, both at the micro and macro level, are greatly improved. The bass notes are faster and more tuneful. One can perceive the vibrations of the membrane of the tympani after each strike of the mallet. The tonal color of the instruments seems more natural and real.
I also changed from the T350s to the Acapella ion tweeters, something I had been very interested in doing ever since I heard them several years before. These tweeters use high-energy electrical plasma to vary air pressure and create sound. What I liked about the T350 is that the phenolic diaphragms avoid the hard edge that metal diaphragms can impart on the high frequencies. Yet the ion tweeters go further and impart a totally natural, ethereal quality to string tone, female voice and percussive instruments. The better high-frequency extension also gives an enhanced perception of space and depth. I estimate that the ion drivers made the greatest difference to my system, even though the frequency response of my ears rolls off above 12 kHz!
As I feel I have finally accomplished a satisfactory result with my speakers, I have moved on to deal with the recordings that I have made over the years. I needed a master recorder for editing and playback of the tapes. The Nagra IV-S is good for neither task; its playback electronics are more of an afterthought, and it does not allow for precise positioning of the tape for editing. The Nagra T-Audio recorder was initially developed as a scientific instrument, and later adapted for the television and film industry. Due to its substantial cost when it was introduced, it was too expensive for most music studios. The listed price in 1983 was £26,000, enough to buy a modest house in London!
Luckily, by the mid-2000s, analogue had fallen out of favor, and I was able to pick one up, fully refurbished with new heads, from Nagra for 8,000 CHF (about $8,800 US). I was attracted to its small footprint and the amazingly precise mechanical function, which makes tape editing very easy. However, the playback electronics of the machine, while competent, are not up to audiophile standard. The extensive use of 1980s-vintage op-amps and complicated compensation networks give the sound an unnatural, electronic character, although the dynamics and scale of the sound are outstanding.
Nagra T-Audio tape recorder.There is now a trend for audiophiles to bypass the native electronics of professional recorders, and this was what I did. I wired the playback head with solid core pure silver wire directly to my prototype preamp, after modifying the RIAA EQ network for IEC and Nagramaster equalizations. (Most commercially available 15-ips tapes nowadays use IEC EQ, and all my own recordings use Nagramaster EQ.) It took some experimentation to optimize the frequency response, but the end result is highly satisfactory. The writeup about the modifications has been published at my group’s website (the Asia Audio Society) for those who are interested in the technical details. Tape playback avoids the pitfalls of LPs, such as distortion, noise and dynamic compression. It also sounds more natural and musical than the majority of digital recordings.
Commercial recordings are becoming available on 15-ips reel tapes from companies such as Tape Project, Analogue Productions and others. I also have a collection of master tape copies of some of my favorite music, which I have obtained from a couple of recording engineers in Europe.
I feel I have finally arrived at a sound that I find quite satisfactory, and I can just enjoy the music without worrying about what I need to do next (for a while).
However, I have been neglecting my turntable for quite a while! And I have not even started looking at digital…
