Western Electric Type No.91E integrated amplifier

During my 30-odd years inhabiting New York City's Greenwich Village, I've seen many things come and go. Today's Village buzzes, blasts, and bellows in every direction, change itself the only constant.

Once the province of printers, factories, piers, and the maritime trade, the Far West Village, from Varick to Greenwich, plays host to a looming Disney megaplex. Concrete office blocks, empty in the 1990s, are choked with condos offering wraparound frontage, their storefronts touting Botox ("Wrinkle Prevention Studio"), cosmetic dentistry ("secretly straighten your teeth"), and bootcamp gyms ("smash your fitness goals"). At Barrow Street stands the old film-noirish Keller Hotel, its Hopperesque HOTEL sign weathered by time and grime. Farther uptown, former slaughterhouse buildings host Apple and Google offices. The nearby Ear Inn, a classic watering hole and neighborhood haunt, stands strong.

At 463 West Street, facing the Hudson, a massive, Neoclassical, 12-story building dominates its block. More than a century old, its copper-clad roof has turned a jaundiced shade of green. Proud and mysterious, its colossal Gothic gates are chained. Traffic rushes by, unaware of its historical significance.

In the 1990s, drummer Ken Micallef rented a subterranean rehearsal space in the building. (Then, as now, it is home to Westbeth Artists' Housing.) Ken—I—was unaware then that the quiet man who took my cash was Paul Bley, the jazz pianist who influenced Bill Evans, Keith Jarrett, and drummer Paul Motian. (Motian eventually left Evans to work with Bley.) The building was designated a National Historic Landmark in 2011, but I've never found a plaque there; unlike many large European cities, New York City is distinctly sparing of plaques. It's as if the city is worried they would hamper development.

463 West Street is a national monument not because jazz musicians worked there—and not because Ken Micallef rented rehearsal space there—but because between 1897 and 1966 it was home to Western Electric, the applied-research division of AT&T, parent company of Bell Laboratories. AT&T/Bell Labs was the source of many hi-fi–related inventions including the vacuum-tube amplifier, the transistor, negative feedback, the world's first wedge-based anechoic chamber, and digital audio. (Harry Nyquist and Claude Shannon, who were involved in the development of the eponymous fundamental sampling theorem, did their work at Bell Labs.) Indeed, the first low-distortion, wide-bandwidth recording apparatus—hence, hi-fi recording—was developed at Bell Labs in 1925 (footnote 1).

The Western Electric division's own creations include the Orthophonic phonograph, the Westrex cutting-lathe system, the 300B vacuum tube, considered by many the greatest triode power tube ever made, and the 1936 WE 91A single-ended triode (SET) integrated amplifier, which powered cinema systems across the country.


Following the 1974 antitrust suit against AT&T and the consequent 1982 dissolution of the Bell telecommunication system, Western Electric spun down. In 1995, the company's tooling, schematics, intellectual property, and name rights were sold to entrepreneur Charles G. Whitener. In 1997, Whitener's Western Electric resumed manufacture of the Western Electric 300B vacuum tube, first at Western Electric's Kansas City works, then in Huntsville, Alabama, and finally to a new factory in Rossville, Georgia, equipped with hi-tech hydrogen-reduction ovens, automatic cathode-cleaning lines, flashy laser-welding systems, a modern water-deionization system, and an updated testing system. Earlier this year, he announced that his company would start producing several other vacuum tubes: 12AX7s, 12AT7s, 12AU7s, 6550s, and 6L6s (footnote 2). If, as seems true, Whitener is determined to elevate his Western Electric to the glory of the original brand, he's off to a hell of a start.

Whitener's 15-strong engineering team works in a state-of-the-art facility where, in addition to 300B vacuum tubes, they manufacture the 91E integrated amplifier ($14,999), modeled after the company's early WE 91A amplifier.

"The goal, from the beginning, was to break through with a SET amp that allowed the 300B to perform at a never-before [heard] level," Whitener wrote in an email. "There are new 300B amps every year from other manufacturers. Our contribution needed to be something significant, something powerful that only Western Electric could pull off."

Design considerations
Whitener claims that the 91E, which was developed by a team of engineers over a period of 10 years, retains what's best about the early WE 91A—its delicacy, bloom, openness, and tonal magic—and adds modern improvements. "You'll find a solid low end and extended highs. We hoped to retain the clarity of the original 91A's midrange as a SET 300B amp. The 91E has zero negative feedback, both local and global. The design of a SET amp without negative feedback is a difficult challenge and requires selection of only the finest components in the audio chain. These factors account for the incredible soundstage performance of the 91E. THD and damping factor may measure better in push-pull configurations but not without compromising musicality and soundstage. The 91E has proprietary topology that all but eliminates audible noise." It also has remote control and Bluetooth, which Western Electric founder Elisha Gray could hardly have imagined.


"The 91E uses Western Electric's patented Steered Current Source (SCS) technology," says a description on the company website. "The unique parallel feed topology contributes half of the AC current to the plate (anode) of the 300B by modulating quiescent current. As a result, half of the power dissipation occurs in the tube, allowing for greater power output." (The 91E is rated at 20Wpc into 8 ohm with 10% THD.)

"Part of [the reason for] the increase in power is [that] we're running the 300B in Class A2, grid drive positive," Whitener explained. "The SCS circuit always keeps the power dissipation of the tube within the safe operating area (SOA), which of course extends the life of the tubes."

The 91E uses a "stepped logarithmic attenuator," the product's webpage explains, "to mimic the response of human hearing. The attenuator consists of discrete resistors in a ladder configuration that are switched in and out of circuit to produce a logarithmic attenuation of the output signal."

"It's similar to an audio taper in potentiometers," Whitener added. "Stepped attenuators have less noise and superior channel-to-channel tracking ability compared to motorized potentiometers. In essence, this microprocessor-controlled stepped attenuator uses relays to engage each discrete resistor in circuit and is accomplished with software that emulates a DAC. It complements human hearing with an algorithm for audio taper."


When the 91E is powered, a gray image of Western Electric's winged Golden Boy (formerly, Genius of Telegraphy), clutching his thunderbolts and electrical cables, appears momentarily in the amplifier's control display. A 30-second warm-up period follows, and then 30 seconds of microprocessor-controlled auto-bias. After 20 minutes of inactivity, the 91E goes into standby mode.

A novel feature of the 91E is its interchangeable transformer block, which is used instead of a multitap transformer. A 4 ohm transformer as well as a 16 ohm one are available as add-ons ($999 each) to replace the included 8 ohm transformer. "We use separate transformers for different impedances because the additional windings in a traditional, tapped transformer cause the signal to deteriorate," Whitener said.

The 91E uses circuit boards. "The circuitry is much too complex for point-to-point wiring," Whitener explained. Brands and specifications of the 91E's capacitors, resistors, transformers, internal wires, RCA jacks, switches, and binding posts are not disclosed.

Setup and first impressions
Unpacking the 91E was like removing a massive jewel from a custom case. Unboxing the pair of 300Bs, each in its own padded container, was a treat. Full documentation accompanied both.

The 91E is massive and required careful maneuvering to fit into my Salamander rack, but two nickel-plated handlebars make moving it a cinch. I plugged in my DAC and turntables and wired up the speakers, but before I could set Golden Boy free to dance around in my room, I had to load my cartridges.


To use the 91E's onboard phono stage, Whitener and Western Electric's Marketing Communications Lead Cobi Boykin supplied me with six pairs of phono termination plugs ($74/pair). Each pair was marked with capacitance for MM (100pF, 220pF, 330pF) or resistive values for MC (100 ohms, 330 ohms, 1000 ohms). Whitener recommended the 100 ohm pair for my EMT TSD15 N Super Fineline. The 330 ohm plug worked best for my Ortofon SPU Classic GE MkII MC cartridge (footnote 3).

Footnote 1: See Bell Labs timeline.

Footnote 2: See Re-Tales #22. Also see Herb Reichert's May 2021 Gramophone Dreams column, "The Venus Tube" and Peter van Willenswaard's "In Search of the Perfect 300B Tube."

Footnote 3: Western Electric offers advice to owners on choosing the proper phono termination plug.

Western Electric
201 West Gordon Ave.
GA 30741
(404) 352-2000

JRT's picture

I think that this should be mentioned as eventual replacement of the tubes is part of the expenses in using the tube amplifier (and the life might be shortened by operation in class A2).

The prices direct from Western Electric for new 300B triodes are currently, $699 each for a single (1) tube, $1499 for a matched pair (2), and $3099 for a matched quad (4). This amplifier uses either two single tubes (unmatched) or a matched pair. The latter would cost $101 more, but might offer better measured performance, perhaps also audible.

JRT's picture

When the preamplifier portion eventually fails, and if repair is impracticable (perhaps some future capacitor leakage destroys an internal layer on the circuit board, or a monolythic integrated circuit fails after it has gone obsolete, out of production, out of support, becoming too difficult to source a replacement component to feasably effect a good repair), then the separate preamplifier could become dumpster fodder while the tube amplifier(s) continue a very long useful life needing occasional replacements of tubes and perhaps also capacitors.

Integrating the components saves build costs (reduces number of separate enclosures, reduces assembly labor, reduces packaging, perhaps also reduces number of separate internal power supplies, etc.). I would argue that it does not save any footprint because a matching separate stereo preamplifier could be designed to stack under the separate stereo tube amplifier. But these are not separated, and I would suggest that the integration is almost all about cost reduction, and in this $15k price level, there should be more than enough budget to avoid the real compromise of excessive integration and the reduction in system flexibility and the reduction in system reliability which both result from that compromise. There should be more than enough in that budget to afford separate components.

JHL's picture

Simply breathtaking. Any passing knowledge of super-triode amps shows this item to be absolutely top-shelf. Thank you for covering it.

(But tubes wear out omg and it just must be a two-piece sez JRT, lol)

Jack L's picture


"Super triode" topology is a totally 'revolutionary' audio amp design using a triode as active negative feedback device from the plate (P) of the output power tube to its grid, claimed to improve the linearity of the power tube, hence better sound. This topology was invented by Shin-ichi Kamijo, a DIYer of Japan some 30 years back.

I doubt very much such "super triode connection" (STC) topology ever employed by any brandname amps.

Or you mean something else ?

Jack L

JRT's picture

Some decades ago, while designing toroid output transformers for Plitron, Menno van der Veen published at least one whitepaper that included information on his "Super Triode", and he trademarked the phrase. And his is different from Shinichi Kamijo's "Super Triode Connection".

It is described in the paper at the link at the bottom of this post. In that paper, see circuit 7 in figure 1, which is similar to Ultralinear, except with the addition of separate cathode feedback windings (aka cathode feedback tapes) connected to the cathodes with phase such that the cathodes receive negative feedback from the output transformer.


Jack L's picture


What are Plitron toroidal transformers to do with "Super triode"???

Triode is a glass tube & transformer is a solid metal impedance converter ???

What is the logics there ?

Jack L

ejlif's picture

whenever I read about tube amps that have all the slam and bass of solid state my thought is why not just get solid state then? I like a tube amp because it sounds that way. I would be interested to compare this amp to the Feliks Arioso 300B. I have that amp and the Decware Zen triode. The Feliks is no wimp and I would say it probably sounds a little bit like solid state in some ways even. To me if you want a tube amp you want zero listening fatigue. To me that is what the hassle of tubes is all about. What you say about the Shindo stuff that sounds very intriguing.

Jack L's picture


It depends on the design, my friend.

My design/built power amps sound fast, punchy & see thru-transparent like a quality SS amp, yet retain tube's melodic & emotional sonics that SS lacks !

Listening to triodes is believing

Jack L

remlab's picture

..would probably match nicely with these

JRT's picture

... Musical Fidelity 750K Supercharger monoblock power amplifiers (circa 2008). Those booster amplifiers are designed to provide a benign load impedance while also boosting power output to something capable of driving conventional loudspeakers to usable SPLs at usable propagation distances with unclipped signal crests (clipping does not sound good, regardless the technology used to provide signal gain), and with low enough nonlinearity such that the nonlinearities of the tube amplifier dominate the sonic character of the signal gain.

JA1 reviewed a pair:

Jack L's picture


Yes, that why trioide sounds better than another other active devcies like pentodes, tetrodes, & all bipolar junction devices, e.g. transistors, FETs, op-amps. Only triodes get truly linear signal transfer curves vs kinked/kneed transfer curves of all other devices.

My skeptical ears hear the sonic difference particularly at high climax volumes !!!

Listening to triode is believing

Jack L

Jack L's picture


Nope, it is not the "booster amps" "to provide benign load impedance" for driving louspeakers.

It is the "booster amps" should be designed enough to handle the roller-coasting 'hostile' load impedance (R+C+L) of any loudspeakers.

The bench test data provided by most, if not all, brandname amps has been based on resistive dummy loads instead of a loudspeakers since day one decades back.

The Sterophile lab bench-tests any amps under review using simulated loudspeaker loads instead of resistive dummy load, providing measureed data more realistic to the realworld situation.

Jack L

JRT's picture

The SET amplifier is not load invariant, operates with better linearity if driving a benign constant resistive load impedance. The input to the booster amplifier provides a benign constant resistive load impedance to the SET amplifier driving it.

The output of the booster amplifier drives the loudspeaker and is much more nearly invariant to the load presented by the loudspeaker, if operated within the design limits of the booster amplifier.

Jack L's picture


First off, why NEED a "booster amp" for a SET amp ????

I doubt very much you have auditioned in depth the sonic quality of a SET - melodic & emotional music quality that noooo other tube & SS output power stage designs can match !!!!!

We are talking music QUALITY not quantity, OK ???

So why add a booster amp to ruin the musicality of a SET ? What is your logics ?????

Adding a booster amp to the loudspeaker output of the SET just amplify the SET/programme sources harmonic, phase distortion & noise level. You are opening a hugh can of worms, pal !

So why bother to go for a SET in the first place? Just get a tube power amp with pentode push-pull power stage to deliver much higher power to push your loudspeakers instead.

Sorry, I don't think you appreciate the distinct musicality of a SET at all !!!!

Technically, you think the 8-ohm loudspeaker output of the WE300B SET would be happy to load a typical 50K-100KR input of a booster amp ???? FYI, the best coupling impedance ratio should be 1:10, OK !

Back to the booster power amp, tube or SS respective, it still has to work with the ever-changing complex impedance (C+L+R) of any loudspeakers hooked up to it. Same destiny of a SET! So what if it were operated "within the bench-tested design limits of the booser amp" ????

Good fascination ! Please get real !

Listening to SET is believing

Jack L

Jack L's picture


Yes, "loop negative feedback" should be not be used in any amps, including phono RIAA EQ, & power amps for best sound.

K. Micallef quoted "91E has zero negative feedback both local & global".

So is there actually any loop NFB in this amp at all per yr captioned quote ?

Jack L

directdriver's picture

"The Western Electric division's own creations include the Orthophonic phonograph, the Westrex cutting-lathe system, the 300B vacuum tube, considered by many the greatest triode power tube ever made, and the 1936 WE 91A single-ended triode (SET) integrated amplifier, which powered cinema systems across the country."

The original WE 91A, I believe, was used as a monitor amp for the projectionist to hear what's playing in the theatre while he's in the projectionist booth. The 91A was not intended to drive speakers for the movie audience.

Jack L's picture



In 1940 cinema era in USA, the cinema/auditorium loudspeakers started to use the then newly invented efficient permanent magnet: "Alneco" in the speaker unit drivers.

The tube amps also started to use the then newly developed power tubes, e.g. KT66 & 6L6 in 4-6 parallel push-pull configuration to provide up to 40-60 watt "hugh" power for all cinema/auditorium sound systems across the country.

No way WE91A low output power was powerful enough to drive the cinema sound systems back then.

Jack L