Pioneer SM-83 Restoration, Repair and Complete EL84 Rebuild

The Pioneer SM-83 is a fairly common Japanese tube amplifier that can still be found today at vintage electronics markets, such as Marzaglia, or through second-hand listings. Units requiring restoration are often sold for between €100 and €150, although the actual price naturally depends on their overall condition.

In this article I present two very different projects. The first part covers the repair of a Pioneer SM-83 and explains the issue with its original 7189A output tubes, which should not be replaced directly with standard EL84s. The second part describes a completely new amplifier built by reusing the chassis and a few surviving parts from an SM-83 that had been reduced to little more than scrap.

Repairing and Retubing the Pioneer SM-83

From a circuit standpoint, the Pioneer SM-83 is not a particularly complicated amplifier and, in most cases, can be repaired without major difficulties. The main concern involves the original output tubes: four 7189As.

For many years the 7189A was available only as a NOS tube, with prices often exceeding the value of the entire amplifier. Today the situation has improved thanks to the new production Tung-Sol 7189, which finally offers a practical solution for anyone wishing to keep the amplifier in its original configuration without relying exclusively on expensive vintage tubes.

The 7189A belongs to the EL84 family but is designed to withstand significantly higher plate and screen grid voltages. For a more detailed explanation of the differences between these tubes, see the article “Exploring the EL84 Tube Family“.

The Pioneer SM-83 was advertised as a 28-watt amplifier, although I have never personally measured more than about 20 watts RMS before clipping. It is likely that the published specification included output levels measured at considerably higher distortion.

I have repaired several Pioneer amplifiers and some Scott models originally designed for 7189A tubes where someone had simply installed standard EL84s without making any circuit changes. The amplifier may appear to work for some time, but the EL84s are exposed to voltages well beyond their ratings and are eventually destined to fail.

The important point to remember is that there is no true modern direct replacement for the 7189A. Anyone wishing to keep a Pioneer SM-83 completely original should use genuine 7189A tubes and accept their higher cost.

Restoration and Repair of the Pioneer SM-83

The amplifier shown in these photographs arrived in my workshop with the typical age-related problems. After more than fifty years of service, it is perfectly normal to find dried-out electrolytic capacitors, out-of-tolerance components and solder joints requiring careful inspection.

The restoration included a complete inspection of the amplifier, replacement of genuinely deteriorated components, verification of the power supply voltages and tube bias, followed by final testing. The goal was to restore the amplifier to its original operating condition while preserving as much authenticity as possible and avoiding unnecessary modifications.


Complete Rebuild of a Pioneer SM-83

The second part of this article is not about a conventional repair. In this case, the original Pioneer contributed mainly its chassis, power transformer and a few other reusable parts. Everything else was designed and built from scratch.

The customer owned an SM-83 with a burned-out output transformer, several circuit faults and no output tubes. In practical terms it was little more than scrap, but the chassis still retained the distinctive appearance of the original Pioneer amplifier. I was therefore commissioned to develop a Premium custom project that would reuse the chassis, the original power transformer and any remaining components still suitable for service.

The objective was to preserve the appearance and basic structure of the original SM-83 while creating a completely new, straightforward and modern amplifier inside. No tone controls, complicated input selectors or unnecessary features—just one or two inputs, a volume control and little else.

The output stage uses four EL84 tubes operating with fixed bias. Each channel employs an ECC82 as the phase splitter, while the original 6AN8 input stage was retained. The original output transformers were replaced with a pair of my own 8KPP84A transformers. The original Pioneer power transformer was retained, and several original capacitors were also reused after being tested and found to be in good condition.

The power supply does not use filter chokes. Instead, the high-voltage filtering is achieved through a double CRCRC network, a simple yet entirely suitable solution for this design.

Naturally, the same amplifier can also be built entirely from scratch without starting from an old SM-83. In that case, a suitable power transformer would also be required. Below is a censored version of the Premium schematic:

Once construction was completed, the amplifier was brought to me for final adjustment, bias setting and laboratory measurements. The workmanship was excellent. A new aluminum front panel was fabricated, the chassis was repainted and the amplifier was assembled using standard surplus electronic components. No boutique capacitors or so-called audiophile components were used—only properly specified, carefully selected and thoroughly tested parts.

Laboratory Measurements

Measured output power is approximately 12 watts RMS per channel before obvious clipping occurs. Frequency response extends from about 8 Hz to 90 kHz within ±1 dB. Total harmonic distortion at 1 watt measures 0.15%. Negative feedback has been deliberately kept moderate. The measured damping factor is approximately 5.5, corresponding to an output impedance of around 1.4 ohms.

Spectrum Analysis at 1 Watt

Frequency Response with Resistive Load

Frequency Response with Reactive Load

8 Hz Sine Wave

The following image shows the 8 Hz sine wave. I wanted to include this measurement because a frequency response graph alone does not fully describe the low-frequency behavior of an output transformer.

A transformer may show very little amplitude loss while already introducing visible waveform distortion. I have observed this even with well-known, expensive transformers whose sine wave started to deform below about 70 Hz.

In the graph, the yellow trace represents the signal from the function generator, while the blue trace is the amplifier output. At 8 Hz the sine wave remains well preserved. Noticeable distortion only began to appear around 6–7 Hz.

At such low frequencies, phase shift is naturally present, and performance of this kind may have little practical importance for normal music reproduction. Nevertheless, it remains an interesting measurement, especially considering how often discussions on internet forums focus on huge primary inductances, miraculous core materials and almost mythical transformer designs without verifying their actual performance on the test bench.

Square Waves at 100 Hz, 1 kHz and 10 kHz

1 kHz Triangle Wave

Continue reading...

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Ancestrum Lux, the Light of the Ancestors

Ancestrum Lux, when something serious is born from scrap! The story of Ancestrum Lux begins the way many SB-LAB adventures begin: with a customer arriving carrying a pile of junk passed off as handcrafted Hi-End equipment by some misunderstood genius. After opening them up, the term “scrap” seemed considerably more appropriate than “Hi-End”.

They were the usual devices described as “super natural”, “transparent” and “musical”. A description that, in the world of creative audiophilia, often successfully replaces any measurement or engineering competence. Listening tests, however, suggested a slightly different conclusion. And a far less flattering one.

One characteristic shared by many of these unpresentables is the presence of large mirror-polished copper or brass panels. To be fair, aesthetics are often the only thing their builders are capable of conceiving. When electronic knowledge approaches zero, there is always the polishing machine. And if one really wants to complete the masterpiece, the classic motor-run capacitors carefully spray-painted black cannot be missing, because in basement-level Hi-End even an industrial component worth a few euros can magically become “audio grade”.

Among the various landfill relics there were two monoblocks based on 801A tubes. Enormous transformers, completely oversized, probably unsuitable for the tube itself, to the point that they could barely deliver 1 watt. Naturally, they were monoblocks. Why? Because in the audiophile world there is a myth that separating the two channels into two distinct chassis magically improves stereo separation.

Monoblocks have only one real advantage: they allow the weight to be distributed. If we are talking about a pair of KT88 power amplifiers with large transformers and generous power supplies, where a stereo amplifier could easily reach 35 or 40 kilograms, splitting it into two chassis makes perfect sense. When instead we are discussing amplifiers delivering barely one watt per channel that, assembled into a single stereo chassis, would perhaps weigh five kilograms, the whole story about “channel separation” belongs more to audiophile folklore than engineering. If two channels interfere with each other inside an amplifier of this type, the problem is not that they share the same enclosure. The problem is that whoever designed or built it did not know how to do the job properly.

Fortunately, the two chassis were identical and, most importantly, built from reusable materials. Perfect candidates for a classic SB-LAB recycling operation. I therefore completely dismantled both amplifiers, recovering all reusable materials together with numerous components taken from other devices delivered by the same person. From this scrap emerged a stereo amplifier divided into two separate chassis: one dedicated to the power supply and one dedicated to the audio section. The choice was not dictated by fanciful audiophile theories about channel separation, but by the need to keep the power transformer as far away as possible from the interstage transformers. The two chassis are therefore connected through a dedicated multipolar cable.

From this demolition came Ancestrum Lux. The name derives from two elements. The first is the presence of 801A tubes, thoriated-tungsten filament triodes known for their characteristic intense glow. The second is the circuit philosophy adopted, strongly inspired by amplifiers from the 1920s and 1930s.

Ancestrum Lux is in fact a completely zero-feedback design, based on interstage transformer coupling and with all bias voltages derived from a single resistive divider inserted in series with the center tap of the high-voltage secondary winding. A solution that would probably make many modern designers faint today, but which was perfectly normal at the time.

The circuit uses a 5R4GY rectifier, an unavoidable choice considering the plate voltages exceed 500V. The input stage is entrusted to 76 tubes. The 76 is a triode introduced in the early 1940s as a universal tube for amplification, oscillators and detectors. It is a simple, linear tube and very pleasant to use in low-level voltage stages.

The 76 tubes drive a pair of 1626s. The 1626 was originally developed as a high-power RF oscillator for military applications. It was not designed for audio use but, as often happens, some tubes created for other purposes proved to be excellent in completely different fields. In Ancestrum Lux, the 1626s operate as an intermediate power stage, providing all the energy required to properly drive the interstage transformers.

Finally, we reach the undisputed stars of the show: the 801A tubes. The 801A is the evolution of the famous 10Y and probably represents one of the finest thoriated-tungsten filament triodes ever produced. Thanks to its extremely linear plate curves, it can operate with very low distortion levels even without any assistance from feedback. Compared to the original 10Y, it offers a higher plate dissipation rating of 20 watts.

Like all directly heated filament tubes, it requires special attention to filament supply design, because any shortcut immediately becomes audible. In the photograph below I was fine-tuning the filament power supply circuit.

All transformers used are custom SB-LAB models. After dismantling the original equipment, the longest part of the work began. Hidden beneath a generous layer of black spray paint was a beautiful walnut veneer that deserved to be preserved. The paint was completely removed, and the many unnecessary holes present on the rear panel were closed.

Custom decorations were then created through laser engraving and wood staining. Once preparation was completed, the chassis received a traditional shellac finish. For final assembly, I built new metal panels which were subsequently professionally painted.

Some mechanical adaptations were produced using 3D resin printing.

In the end, a large quantity of components from the original equipment found a new life inside a project finally worthy of the name.

The final measured performance is:

  • Maximum power: 3.6W
  • Undistorted power: 2.8W
  • Damping factor: 1.7
  • THD at 1W: 0.9%

Observing the frequency response reveals an interesting detail. The output transformer deliberately exhibits approximately 1dB attenuation at 50Hz. No, this is not a mistake. It is entirely intentional. Zero-feedback amplifiers with a low damping factor naturally tend to produce bloated, slow and poorly controlled bass. Many Hi-End manufacturers fight the problem in exactly the same way, deliberately limiting low-frequency response. The difference is that almost nobody will ever admit it.

Having repaired, modified and measured hundreds of amplifiers over the years, I can say that I have encountered this solution far more often than many people imagine. In Ancestrum Lux this choice was made openly and consciously. A slight controlled attenuation is preferable to artificially inflated and uncontrolled bass. Even the shape of the 10kHz square wave perfectly reflects the nature of the design. After all, the signal passes through no fewer than two transformers along the audio path. Expecting solid-state amplifier performance would simply mean not understanding what is being observed.

From a sonic perspective, Ancestrum Lux possesses all the classic characteristics of zero-feedback amplifiers. The sound is open, airy and endowed with a remarkable sense of depth and three-dimensionality. These are qualities many enthusiasts deliberately seek and which contributed to the legendary reputation of classic triode circuits.

These characteristics do not appear out of thin air. Like every engineering decision, the absence of feedback carries a precise sonic signature. Ancestrum Lux was not designed to pursue absolute fidelity or a reproduction rigorously identical to the original signal, but rather to enhance those particular tonal nuances that make classic triode amplifiers so fascinating.

This does not mean the result is less enjoyable, quite the opposite. It simply means that the goal of the project is not to disappear completely from the audio chain, but to offer a musical presentation with its own personality. It is precisely this personality, made up of harmonics, spaciousness and perceived naturalness, that continues to attract many enthusiasts to zero-feedback circuits even today.

For this reason, it finds its ideal match with full-range drivers, acoustic-suspension systems and high-efficiency horn loudspeakers. Like all zero-feedback amplifiers with a low damping factor, it is not the ideal choice for bass-reflex loudspeakers that depend heavily on electrical control of the driver at low frequencies.

The nature of the project should also be considered when choosing musical material. Ancestrum Lux is built around a historic circuit topology, free of feedback and characterized by a very low damping factor. For this reason, it is not the most suitable amplifier for reproducing particularly complex, densely layered music with large dynamic swings and numerous overlapping instruments. Expecting this kind of performance from a few-watt single-ended zero-feedback amplifier would simply mean asking it to perform a task it was never designed for.

Where Ancestrum Lux truly excels is with small acoustic ensembles, jazz, vocal music, blues, chamber music and, more generally, all genres in which tonal naturalness, soundstage reconstruction and microdynamics are more important than sheer impact and sound pressure. In these contexts it delivers an engaging and highly captivating listening experience, perfectly consistent with the philosophy of the great triode amplifiers of the past.

Ancestrum Lux was born from a pile of junk equipment. And perhaps that is its most interesting aspect. It demonstrates that even the worst audiophile scrap can become something worthwhile, provided there is a serious design behind it, a few decades of experience, and above all the determination to place engineering ahead of legends.

Continue reading...

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Aurelion, the 1.6W single-ended amplifier born from the forgotten tubes of the 1920s

When an amplifier becomes something truly unique. Some projects begin as simple commissions, while others slowly evolve into something far more personal. Aurelion unquestionably belongs to the latter category.

This project was born from a very specific request: to create an unconventional single ended amplifier built around extremely rare historical tubes, with intentionally minimal output power. An amplifier designed for late-night listening sessions, bedroom systems, and low-volume enjoyment, while still preserving all the sonic magic that only certain minimalist circuits can deliver. And honestly, these are exactly the kinds of projects that truly excite me.

The Meaning Behind the Name Aurelion

The name Aurelion was conceived as an epic and fantasy-inspired reinterpretation of the Latin name “Aurelio”, itself derived from “aureus”, meaning “golden” or “shining”. It does not refer to any real star, yet it evokes something luminous, celestial, majestic, and almost cosmic. A name intended to suggest light, elegance, energy, and visual presence, while still maintaining a connection with classical Latin roots.

Even the symbol engraved on the front panel follows the same philosophy. A central eye surrounded by a radiant star and astronomical references. The eye symbolizes vision, perception, and awareness, while the star represents energy and light. Together they create a deliberately mysterious and elegant image, perfectly aligned with the character of the amplifier itself.

A Single Ended Amplifier from Another Era

Aurelion uses 843 tubes as output devices, indirectly heated triodes originally developed for telephone applications during the first half of the 1920s. These are tubes that are virtually unknown in the modern audio world, yet absolutely fascinating from both historical and technical perspectives.

For the driver stage I selected 24A tubes, generally considered the first true modern tetrodes. These also date back to the first half of the 1920s and represent one of the earliest evolutionary steps beyond the classic triode. I also discussed the 24 tube family in this article. To stabilize the screen grid supply of the 24A tubes, I used an 85A2 neon gas regulator, while rectification is handled by a 5U4GB.

An Amperite 6N030 thermal delay was also included to postpone the application of high voltage without relying on modern timer circuits.

Despite the extremely low current requirements of the 843 tubes, a 5U4GB rectifier still proved necessary in order to keep operating voltages within manageable limits. The 843s operate at roughly 500V into a load of approximately 12kohm. Final output power is intentionally very low, around 1.6W RMS nominal.

That may sound ridiculously low by modern standards, and in many ways it is. But this amplifier was never designed to “play loud”. It was specifically created for intimate and relaxed listening sessions with high-efficiency loudspeakers, prioritizing delicacy, microdynamics, and tonal presence above sheer volume.

Output Transformer Design

As often happens with my projects, everything started from the output transformers. In this particular case, the combination of extremely low power and very high primary impedance made double C-core construction especially suitable. For this reason I decided to adopt the “2C” technique used in many classic McIntosh transformers.

To achieve this, I designed and 3D printed a dedicated resin bobbin developed specifically for this amplifier. The transformers were then hand wound using carefully selected materials and highly meticulous construction techniques.

Once completed, the transformers were resin impregnated and installed inside soft iron enclosures. These housings are not merely mechanical supports, they also help shield stray magnetic fields. Details like these are rarely visible from the outside, yet they are an integral part of the SB-LAB construction philosophy.

Chassis, Mechanics and Construction

For the external appearance, the amplifier was built using a wooden chassis designed to match Neutrino, the preamplifier intended to partner this power amplifier. Aurelion is in fact a pure power amp, with no volume control and no input selector.

I then produced the aluminum panels, handling drilling and finishing personally, machined the bakelite mounting panel, and completed the full assembly of the amplifier. As always, every part was handcrafted individually, without relying on standardized industrial solutions.

From a circuit design perspective, I deliberately chose not to use global negative feedback. Aurelion operates in an extremely direct and minimalist way, keeping the signal path as clean and simple as possible.

The only feedback present is a very light type-1 shade feedback applied locally, totaling an almost irrelevant 1.25dB. Such a low amount does not perceptibly alter the sonic character of the amplifier, nor does it compromise its dynamics or tonal naturalness. What it does provide is a small degree of compensation for the inevitable construction tolerances found in tubes manufactured nearly a century ago.

This was therefore more of a technical refinement than a corrective measure, implemented not to chase laboratory numbers, but to achieve greater operational consistency while preserving the original sonic personality of the project.

Measured Performance

Final measurements are as follows:

  • Clean output power: 1.7W RMS
  • Maximum full clipping power: 2.5W
  • Damping factor: 2.2
  • Excellent square wave response
  • Frequency response: 10Hz – 70kHz (-1dB)

THD

Bandwidth

Square Waves 100Hz / 1khz / 10khz

On paper, these numbers may appear almost insignificant compared to modern standards, yet they reveal very little about the true personality of this amplifier. Aurelion was never created to impress through brute power. It was created to build atmosphere.

It is one of those projects that would make very little sense as a mass-produced product, and precisely for that reason it perfectly represents the spirit of SB-LAB: building unique, timeless devices without compromise and without chasing commercial trends.

Continue reading...

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.