Hello Gang,
After a long development and testing process, I wanted to share my work on evolving the Sony TAN-8550 V-FET amplifier — a personal project aimed not at restoring the amp to factory condition, but at carefully reimagining it as a "Mark II" version, guided by 50 additional years of materials science, field data, and circuit evolution.
I hope this write-up is useful or at least thought-provoking for others who care about preserving and honoring these remarkable vintage designs.
This not a restoration.
This project is an evolutionary enhancement of the TAN-8550, and each change undertaken with the utmost respect for the original design.
Rather than restoring the amplifier to its original 1970s configuration, my goal was to thoughtfully reimagine the unit — as if Sony themselves had created a "Mark II" version, equipped with the benefit of 50 years of additional data and materials advancements.
At no point did I seek to "modernize" the TAN-8550 for its own sake. The changes are rooted in preserving its original soul while evolving key areas where reliability, stability, and audio performance could benefit from today's best practices.
1. Power Supply Redesign: Strengthening the Core
In the original design, Sony used dual-can electrolytic capacitors — a relatively common technique of the era — to manage parasitic elements such as series resistance and inductance. However, these capacitors are now largely unavailable, and newer design practices have rendered such compromises unnecessary.
For the redesign, the focus was primarily on current flow, ground stability and signal integrity:
Note the length of the old wire harness, between the filter cap and power supply, bridge rectifier and all. That is now eliminated :
Side Note: Why the Main Power Supply Matters in a Push-Pull Class AB Amplifier - How this was explained to me by very smart people and how I can best explain it forward:
In a conventional Class AB push-pull amplifier — particularly designs like the Sony TAN-8550 that share a common ground between the power supply and speaker output — the main filter capacitors are not isolated from the audio signal. They are, in fact, an integral part of the signal path.
Here’s why:
Thus:
By using audio-grade capacitors such as the Nichicon LKG, with their low, linear ESR, fast transient response, and high stability across frequencies, the amplifier’s rails remain stable and uncolored — allowing the V-FET output devices, in this case, to operate as intended, with minimal imposed distortion.
In essence:
We are not listening to the amplifier circuitry, we are listening through the power supply. Choosing the right capacitors and proper supply PSU circuit layout, ensures that what reaches the speakers is a faithful, clean modulation of the input signal, not a struggle against rail sag, ripple or imbalance.
2. DC Protection Upgrade / Addressing "Lack of Protection":
Protecting the Unobtanium not just the Speakers
The TAN-8550 originally employed a crowbar protection circuit that shorts the output to ground in the event of DC fault detection at the output.
While it protects the speakers, it can expose the unobtanium V-FET output devices to catastrophic fault currents.
For this upgrade I replaced the crowbar circuit with a more modern time-delay relay protection system adapted from Sony’s own later TAN-5550 design (with some required modifications). This circuit redesign:
In addition to the core improvements named above. I added modern features and revised legacy functions with modern techniques:
Frame of Reference and Iterative Testing
A critical part of this project was the use of a "mule unit" — a carefully preserved TAN-8550 with original configuration and components, including original-specification dual capacitors, thoroughly tested to meet Sony’s original data sheet specifications. This mule was not used for daily listening but served a singular purpose:
The evolution project itself took place over many months of iterative development and over two years from initial ideation to final completion.
Each change was carefully evaluated against the mule's sonic benchmark to ensure that no enhancement compromised the amplifier’s original strengths.
It is relatively easy to replace parts and claim an amplifier is "more reliable" — but in truth, the odds are heavily against achieving a sonic improvement unless every decision is grounded in disciplined referencing and technical care.
A poorly considered "recap" can easily degrade performance rather than enhance it.
Results:
The redesigned TAN-8550 exhibits:
These sonic improvements have been confirmed by seasoned listeners familiar with multiple examples of this same model — including some in excellent cosmetic condition but fitted with inferior-grade SMPS capacitors, original wire harnesses, or original PCBs whose limitations are well-documented compared to modern practices.
One particularly polite but revealing feedback quote from a fellow owner: "Wow - there is a difference. And not a small one"
Another Hi-Fi enthusiast visiting for an evaluation session simply did not want to leave, remaining in the listening seat for hours, deeply engaged by the sound.
Similar to my Yamaha projects, this is an evolutionary tribute, this time to a Sony Legend.
Every decision made throughout this project was guided by a simple philosophy:
Respect the original spirit. Improve only where experience, materials science, and circuit evolution offer a clear and meaningful advantage.
The result is not only a far more reliable amplifier than the original, but one that — in a very real sense — outperforms the original sonic expectations without betraying the intent and character that made the TAN-8550 a classic.
I know that purists do not agree with this philosophy but at least I hope this project resonates with those who believe that evolution, when pursued with discipline, and passion, can honor and extend the life of these extraordinary classic designs.
For reference, I am attaching a related youtube video. You have to excuse my mental acuity and coherence during the recording of this video. Any hiccups I will blame on Bacchus
Final assembly:
After a long development and testing process, I wanted to share my work on evolving the Sony TAN-8550 V-FET amplifier — a personal project aimed not at restoring the amp to factory condition, but at carefully reimagining it as a "Mark II" version, guided by 50 additional years of materials science, field data, and circuit evolution.
I hope this write-up is useful or at least thought-provoking for others who care about preserving and honoring these remarkable vintage designs.
This not a restoration.
This project is an evolutionary enhancement of the TAN-8550, and each change undertaken with the utmost respect for the original design.
Rather than restoring the amplifier to its original 1970s configuration, my goal was to thoughtfully reimagine the unit — as if Sony themselves had created a "Mark II" version, equipped with the benefit of 50 years of additional data and materials advancements.
At no point did I seek to "modernize" the TAN-8550 for its own sake. The changes are rooted in preserving its original soul while evolving key areas where reliability, stability, and audio performance could benefit from today's best practices.
1. Power Supply Redesign: Strengthening the Core
In the original design, Sony used dual-can electrolytic capacitors — a relatively common technique of the era — to manage parasitic elements such as series resistance and inductance. However, these capacitors are now largely unavailable, and newer design practices have rendered such compromises unnecessary.
For the redesign, the focus was primarily on current flow, ground stability and signal integrity:
- Replaced the original filter capacitors with audio-grade Nichicon LKG some of the best caps made for audio
- Upgraded the bridge rectifiers to soft recovery diodes, significantly reducing high-frequency noise and switching artifacts.
- Reworked the grounding topology with balanced ground potentials and symmetrical current paths, inspired by later designs from manufacturers such as Luxman and Accuphase.
Note the length of the old wire harness, between the filter cap and power supply, bridge rectifier and all. That is now eliminated :
Side Note: Why the Main Power Supply Matters in a Push-Pull Class AB Amplifier - How this was explained to me by very smart people and how I can best explain it forward:
In a conventional Class AB push-pull amplifier — particularly designs like the Sony TAN-8550 that share a common ground between the power supply and speaker output — the main filter capacitors are not isolated from the audio signal. They are, in fact, an integral part of the signal path.
Here’s why:
- During normal operation, the upper and lower output devices (in this case, N-channel and P-channel V-FETs) alternately draw current from the positive and negative supply rails.
- Each half of the waveform (positive and negative) relies on the stability and instantaneous energy availability of its corresponding rail.
- The ground potential, shared between the amplifier’s internal circuitry and the speaker return path, must remain balanced and undisturbed — even during rapid transients that cause heavy dynamic current flow.
- Any instability, asymmetry, or voltage ripple on either rail, or imbalance in the ground reference, directly modulates the audio output, adding distortion, smearing dynamics, and collapsing the soundstage.
Thus:
- The capacitors bridging each rail to ground act almost as extensions of the output stage itself.
- Their ability to charge and discharge linearly, rapidly, and symmetrically across a wide frequency band is critical to maintaining signal purity.
By using audio-grade capacitors such as the Nichicon LKG, with their low, linear ESR, fast transient response, and high stability across frequencies, the amplifier’s rails remain stable and uncolored — allowing the V-FET output devices, in this case, to operate as intended, with minimal imposed distortion.
In essence:
We are not listening to the amplifier circuitry, we are listening through the power supply. Choosing the right capacitors and proper supply PSU circuit layout, ensures that what reaches the speakers is a faithful, clean modulation of the input signal, not a struggle against rail sag, ripple or imbalance.
2. DC Protection Upgrade / Addressing "Lack of Protection":
Protecting the Unobtanium not just the Speakers
The TAN-8550 originally employed a crowbar protection circuit that shorts the output to ground in the event of DC fault detection at the output.
While it protects the speakers, it can expose the unobtanium V-FET output devices to catastrophic fault currents.
For this upgrade I replaced the crowbar circuit with a more modern time-delay relay protection system adapted from Sony’s own later TAN-5550 design (with some required modifications). This circuit redesign:
- Provides controlled isolation of the speakers during faults without risking damage to the amplifier.
- Introduces a startup delay, eliminating loud transients ("pops") on power-up — a notable improvement for high-sensitivity speaker systems.
3. Practical Enhancements for Future-Proofing Reliability
In addition to the core improvements named above. I added modern features and revised legacy functions with modern techniques:
- Speaker terminals were replaced with modern, high-quality connectors suitable for today's heavier-gauge speaker cables.
- AC mains cord was replaced with a modern IEC inlet, offering flexibility for appropriate power cords.
- Low-level signal switches, previously open-frame designs susceptible to oxidation, were replaced with sealed low-level signal relays to ensure clean switching and long-term reliability.
4. Methodology Matters:
Frame of Reference and Iterative Testing
A critical part of this project was the use of a "mule unit" — a carefully preserved TAN-8550 with original configuration and components, including original-specification dual capacitors, thoroughly tested to meet Sony’s original data sheet specifications. This mule was not used for daily listening but served a singular purpose:
- Establishing a true frame of reference for the original sonic signature and performance envelope of the TAN-8550.
The evolution project itself took place over many months of iterative development and over two years from initial ideation to final completion.
Each change was carefully evaluated against the mule's sonic benchmark to ensure that no enhancement compromised the amplifier’s original strengths.
It is relatively easy to replace parts and claim an amplifier is "more reliable" — but in truth, the odds are heavily against achieving a sonic improvement unless every decision is grounded in disciplined referencing and technical care.
A poorly considered "recap" can easily degrade performance rather than enhance it.
Results:
The redesigned TAN-8550 exhibits:
- Substantially improved imaging and soundstage precision.
- Tighter, more authoritative low-frequency control.
- Lower noise floor and improved microdynamic resolution.
These sonic improvements have been confirmed by seasoned listeners familiar with multiple examples of this same model — including some in excellent cosmetic condition but fitted with inferior-grade SMPS capacitors, original wire harnesses, or original PCBs whose limitations are well-documented compared to modern practices.
One particularly polite but revealing feedback quote from a fellow owner: "Wow - there is a difference. And not a small one"
Another Hi-Fi enthusiast visiting for an evaluation session simply did not want to leave, remaining in the listening seat for hours, deeply engaged by the sound.
Final Thoughts:
Similar to my Yamaha projects, this is an evolutionary tribute, this time to a Sony Legend.
Every decision made throughout this project was guided by a simple philosophy:
Respect the original spirit. Improve only where experience, materials science, and circuit evolution offer a clear and meaningful advantage.
The result is not only a far more reliable amplifier than the original, but one that — in a very real sense — outperforms the original sonic expectations without betraying the intent and character that made the TAN-8550 a classic.
I know that purists do not agree with this philosophy but at least I hope this project resonates with those who believe that evolution, when pursued with discipline, and passion, can honor and extend the life of these extraordinary classic designs.
For reference, I am attaching a related youtube video. You have to excuse my mental acuity and coherence during the recording of this video. Any hiccups I will blame on Bacchus
Final assembly:
Attachments
Last edited:
"likes" if they could.
