User Tools

Site Tools


audio:audio_by_van_alstine_work

Audio by Van Alstine work

I have done quite a bit of consulting work for Audio by Van Alstine over the years. I'm summarizing below some of it. While I completed a number of smaller projects for AVA before 2002, I decided to pick up the story with the Transcendence Five as that is the first product that was my complete circuit design. The list below isn’t exhaustive.

The focus of my work with AVA has always been to bring the highest standards of audio engineering to the company while helping to foster a more considered and inclusive view of the industry and its consumers.

Transcendence Five

c. 2002

New vacuum tube line stage to replace AVA’s existing design originally derived from the Dyna PAS 3.

Contribution: Concept origination, development, and implementation. I’m pretty sure this was the first commercial application of my vacuum tube computer optimization methods based on my SPICE model optimization work (i.e., there were two layers of computer-based optimization: one to model the tubes and the other to optimize the performance of the circuit).

Discussion: I thought the time had come for AVA to have a new vacuum tube line-level stage to replace their aging 12AX7A-based design, and they agreed. I based the initial work around two stages of 12AT7A gain using heaps of the SPICE simulation to optimize the circuit behavior for audio reproduction. I originally wanted to apply my design optimization methods to a number of different tube triodes and subjectively evaluate the results, but after hearing the results with the 12AT7A, AVA decided to put the design into production immediately. A variant of this stage remains in production today in the form of AVA’s Fet Valve CFR and FetValve CF RB preamps.

Transcendence Six

c. 2002

Low-cost hybrid line stage using a computer-optimized 12AT7A gain stage and integrated circuit buffer output.

Contribution: Concept origination, development, and implementation.

Discussion: This was a continuation of the work I started with the Transcendence Five. Historically, AVA had demonstrated some success with two different hybrid approaches. In the first, high-voltage MOSFET buffers were placed inside the main feedback loop of vacuum tube gain stages. In the second, a low-cost opamp stage was used to provide high impedance buffering of a conventional closed-loop tube preamp design. My idea centered around placing a high-speed solid-state buffer IC stage inside the global feedback loop. This would allow AVA to offer a preamp with performance that exceeded their tube/MOSFET hybrid at a considerable cost savings. This design was so good that it obsoleted their tube/MOSFET hybrid preamp until I was able to complete work on the Transcendence Seven.

Transcendence Seven

c. 2002

12AT7A-based premium hybrid line stage using a computer-optimized 12AT7A gain stage and optimized power-MOSFET buffer output stage.

Contribution: Development and implementation. This was a major rethink of the work covered by AVA's “Fet Valve” patent and embodied in their first-generation Fet Valve products.

Discussion: One of the reasons my Transcendence Six design outperformed AVA’s more costly discrete MOSFET hybrid approach was that it used the optimized 12AT7A-based gain stage core initially developed for the Transcendence Five. So a natural progression was to see whether using that gain stage in one of the discrete MOSFET output buffer architectures covered in their patent provided an advantage over the IC-based buffer I designed into the Transcendence Six. In the process of designing the new MOSFET stage, I revisited a number of assumptions made in the original AVA implementation and produced a circuit optimized for this application that was vastly simpler than the one it replaced. AVA thought its performance warranted it being offering as a premium hybrid product.

"T7 CF"

c. 2002

All-tube buffered line stage using a computer-optimized 12AT7A gain stage and 12AU7 buffer stage.

Contribution: Concept origination and development.

Discussion: While working on the Transcendence Seven, I was able to take a deep dive into the tube/MOSFET architecture covered in the AVA Fet Valve patent. One result of this is that I could not find a convincing reason why a power-MOSFET was better suited to buffering the gain stages than a cathode follower, but I was able to come up some reasons a MOSFET might be worse. So I developed an iteration of the Transcendence Seven design using a 12AU7A-based circuit to replace the power-MOSFET stage.

At the time I wasn’t given the go-ahead to prototype and test this design. My understanding is the AVA wanted to maintain a commitment to their patented architecture as they felt this gave them marketing distinction. However, this circuit was eventually put into production around 2014 as the Fet Valve CF.

Active power supplies for tube and hybrid electronics

c. 2004

All-new power high-voltage supply regulation circuits for use with all AVA tube and hybrid products.

Contribution: Concept origination, development, and implementation.

Discussion: To address a concern AVA raised regarding the audibility of B+ power supply shifts caused by line voltage variations and other factors, I developed an active regulation architecture using voltage references and high-voltage power-MOSFETs. The approach was simple and cost-effective enough that several independent regulation stages could be be used to better isolate the individual stages that comprised an entire design. AVA felt that the subjective improvements resulting from the regulated power supplies warranted the adoption of the scheme in all tube and hybrid products. This resulted in the Transcendence Five SLR, Transcendence Six SLR, and Transcendence Seven SLR and ECR. The same work was also applied to the Fet Valve EXR power amplifiers.

The Ultra preamplifier (c. 2006) was essentially a Transcendence Seven SLR with a greater number of regulation stages. Similarly, the first version of the Transcendence Eight (c. 2006) was essentially a Transcendence Five SLR with more regulators. I wasn’t involved in the development of these particular applications of my regulator design as my ongoing work for AVA went on hiatus just before these products went into development.

Remote control

c. 2004

Remote controlled volume circuit adaptable to all existing AVA preamplifiers.

Contribution: Concept origination, development, and implementation.

Discussion: I was aware that AVA was losing preamplifier sales because they didn’t support any kind of remote control and that AVA regarded even the simplest of remote control functionality an unsolvable problem given their manufacturing quantities and methods. So I decided to solve it. The solution hinged on working with a turnkey supplier of remote control transmitters and pre-programmed receiver ICs and sourcing a motorized potentiometer that was as good as the potentiometers they were using. The result involved minimal changes to the chassis design and a build that was easier than the non-remote approach they had been using.

My original expectation was that AVA would offer the remote as standard equipment on all “big body” preamplifiers. Instead it was offered as a relatively costly option, which they justified because the remote control transmitters were costly, which they were because of the small quantities they ordered them in, which was the case because remote control was offered only as an option.

Ultimate 70 and Ultravalve power amplifiers

c. 2005

Revision of AVA’s low-power tube amplifier circuits for the Dyna ST70 chassis.

Contribution: Strategization and development. Changes centered on improving loop behavior and power supply regulation, producing a circuit design that has remained in the AVA catalog for over 13 years.

Discussion: AVA felt they needed to produce a revision of their Dyna 70 rebuild offering and asked me to design it. I was proscribed from changing the tube complement or making similar large changes. So I opted to apply some of the insights gleaned from the application of regulated power supplies in tube and hybrid applications to the existing circuit and to revisit some assumptions about pole and zero locations that the original Dyna engineers had made, likely based on the kinds of parts that were available in 1959 and definitely based on the slightly different tube complement. The revisions I developed resulted in the Ultimate 70, and the circuit continues today in different packaging as the AVA Ultravalve amplifier.

DAC core for Vision and Vision Hybrid DACs

c. 2009

Next-generation S/PDIF decoder and DAC board

Contribution: Design and implementation.

Discussion: In the mid-1990s, I built AVA a prototype stand-alone DAC that demonstrated to them that producing such a product was within their manufacturing abilities. This ushered in an era of their being able to offer a product that they had previously considered impossible. So it was probably natural that they approached me to design the core circuits that would be used in their next-generation of stand-alone DACs. They had preselected the the Wolfson WM8740 as the DAC IC they wanted the design based around, but by the time the design was completed we had moved to the WM8742.

AVA originally wanted to support only 44.1kHz PCM formats and a single coaxial S/PDIF input. Additionally, their manufacturing capabilities precluded the use of microcontrollers, which would have allowed for advanced use of the receiver and converter ICs. In spite of these constraints, I designed in automatic support for up to 96kHz PCM streams, jumper-configurable support for 192kHz PCM, and support for multiple inputs. Because of this, AVA was able to use this design to meet subsequent customer demands for high-rate audio support and multiple inputs; this board was used unchanged for all AVA DAC production for seven years until it was replaced by my DAC MK5 design.

This would be the first time AVA would use predominantly SMD parts in a PCB design. The previous generation DAC had used a single hand-soldered SOIC package. This design used SMD components exclusively.

I was not involved in the development of the analog stages for the first DAC products using this board. However, I was responsible for the analog circuits found in the 2015 revision.

Fet Valve CF

c. 2014

All-tube, next-generation premium preamp line stage.

Contribution: Concept origination and development.

Discussion: Around 2014, finding that hybrid topologies were becoming a marketing liability, AVA revisited my “T7 CF” design work without my knowledge. To my original circuit design of 2002, they added grid stopper resistors ahead of the 12AU7A stages (which my original simulations didn’t point to the need for). They released it as the Fet Valve CF, and it remains in production today as the Fet Valve CFR and Fet Valve CF RB.

Vision RIAA Phono Preamplifer

c. 2014

Component value selection for best RIAA conformance.

Contribution: Development.

Discussion: AVA decided to pursue a stronger presence in the growing phono market by developing a stand-alone phono preamp. I was consulted on the project near the end of its development to determine the best passive component values to be used to yield the best RIAA conformance. I provided some subjective evaluations of the unit while it was being developed as well. I had nothing to do with the preamp’s topology or other design decisions.

Vision DAC improved

c. 2015

New reconstruction and output stage based on a discrete, class-A, solid-state gain stage I developed as an independent project.

Contribution: Concept origination, development, and implementation.

Discussion: I had been developing a discrete, class-A, solid-state gain stage as an independent project for a number of years. My goal was to create a design that was free of the issues I kept hearing in opamp ICs. Late in 2014 I thought I succeeded, and so I built a single-stage differential-to-single-ended anti-imaging filter/output stage using the module for the Vision DAC board I had designed earlier for AVA. The result impressed AVA enough that it was immediately adopted for their Vision DAC offerings.

Vision SL improved

c. 2015

New line stage based on a discrete, class-A, solid-state gain stage I developed as an independent project.

Contribution: Development and implementation.

Discussion: Based on the performance gains my discrete gain-cell module provided the Vision DAC, AVA became interested in using it for their solid-state preamp offering.

DAC MK5

c. 2017 or this

All-new DAC product with USB and S/PDIF inputs supporting high-rate PCM and DSD formats.

Contribution: Development and implementation. All-new DAC implementation based on the AK4490EQ. All-new reconstruction filter and output stage circuitry based on my discrete, class-A, solid-state gain stage. Fully isolated S/PDIF receiver based on the AK4118EQ. Galvanic isolation and switching for the Amanero Combo384 USB receiver. Microcontroller programming for input decoding, DAC operation, and user interface. Industrial design subject to AVA’s production and styling constraints. Production workflow refinements to accommodate large-scale SMD production.

Discussion: With the Vision DAC incapable of effectively supporting newer in-demand digital audio formats, AVA asked me to design for them a product that would. The original design brief specified support for up to 192kHz PCM sample rates and DSD64. I insisted on a design that provided support for DSD128 as well and that had the infrastructure to support DSD256 once that format matured. Full support for DSD256 was added in 2018.

Working with application engineers at AKM, I developed support circuitry for the AK4490EQ whose low-frequency reference voltage stability was significantly greater than the circuits used in AK4490EQ reference designs. This in turn reduced the low frequency nonlinearity of the DAC appreciably over more conventional approaches.

This would be the first design that required AVA to use microcontrollers they would burn in-house — and it did so in a big way: I ended up using three independent microcontrollers in the final product, one for each of the DAC’s core functions: input selection and control, DAC control, and the user interface. Also notable in this design is the scale to which SMD technology was adopted and the high level of build integration. These required introducing significant changes to AVA’s manufacturing workflow, which in turn has brought AVA’s production methods more in line with modern approaches.

Vision SLR, Fet Valve CFR

c. 2017 or this

Fully remote-controllable chassis and supporting circuitry.

Contribution: Concept origination, development, and implementation. New for AVA is minimal-path relay-controlled input selection, custom user interface design and microcontroller programming, and improved manufacturability over the previous generation's less capable chassis. Industrial design subject to AVA’s production and styling constraints.

Discussion: I had two main goals with this project. The first was to bring the remote controllability of AVA preamplifiers up to feature parity with the competition as quickly as possible, and the second was to move AVA away from the costly turnkey remote control solution I adopted for their initial remote control offering. An additional goal was to simplify the assembly of preamplifiers. I originally planned to work on the first of these goals immediately after completing the volume-only remote control design back in 2004. I had even gotten as far as producing prototype PCBs and a prototype build using relay-controlled input switching. But AVA opted for different priorities at the time, and this work was tabled. In 2017, the case for developing a fully remote-controllable platform was far more undeniable.

RB-RC

c. 2018

Low-cost remote controlled volume circuit adaptable to AVA's existing RB chassis.

Contribution: Concept origination, development, and implementation.

Discussion: Given the availability of a less costly remote control transmitter that I secured for the SLR and CFR chassis as well as AVA's new ability to burn microcontrollers in-house, it was natural to apply these toward a limited-feature remote control design for their RB (Real Basic) preamplifer chassis to replace the existing design developed 14 years earlier. As was the case 14 years before, I expected AVA to adopt the remote control as standard equipment on the RB chassis, but again they decided to offer it as an option, albeit not as costly as the solution it replaced.

The present

I continue to provide ongoing consulting services to Audio by Van Alstine, particularly in the areas of digital audio, small-signal analog audio, low-power amplifier design, and user interface design. For reasons that should be obvious, I'm not at liberty to discuss projects under development.

The future

Within audio, my work focuses on developing solutions that represent meaningful audio progress rather than chasing coloration signatures that happen to be fashionable. On a more general level, my design interests include a desire to bring greater equity to technological realms and to improve user experiences in every stage of product use. I try to bring all this to bear as much as possible in my work for AVA.

audio/audio_by_van_alstine_work.txt · Last modified: 2019/08/03 00:03 by mithat