What's the best EL84 amp?

Not any longer. I lived in SoCal (Glendale, Sherman Oaks, Burbank, Tujunga, and finally Palm Desert) from 1979 until February of 2016, when I moved up here to Vancouver Washington, just across the state line (the Columbia River) from Portland Oregon.

Ya know, the 868 is available both as just a line stage (868L), and as a full function (line and phono stages) pre-amp. I too would love an EAR 912 (Art Dudley said it was the only pre-amp he had heard which competed with his Shindo), but unless someone dies and leaves my some dough that ain't gonna happen! 

Exactly @fleschler. Roger and Tim were the only members of a mutual admiration society ;-) .

About ten years ago I entered a room at the hi-fi show held by the airport in Los Angeles, and was surprised to see Tim de Paravicini sitting in a chair, no one around him. I sat down and had a nice conversation with him, during which he told me a lot about tubes. Like Modjeski, he was an expert in the subject, and considered the tube amp designers job to be to maximize the quality of the signal coming out of a tube (or set of them). The loss of both Roger and Tim is a real tragedy for hi-fi music reproduction.

I don’t own an EAR amplifier, but I do own (and love) an 868L pre-amp, best one I’ve ever had.

@pinthrift: For Roger’s thoughts, advice, and warnings about "hi-fi" fuses, make sure to read what he posted about that subject on his AudioCircle Forum. In fact, do NOT install any in your RM-10 until you have done so. Roger is not around to repair your amp if you don’t.

It was Roger’s electrostatic loudspeaker that I thought was the last product he introduced as a "stock" (at least by Modjeski standards ;-) Music Reference model. And I understood it to be available either with or without the direct-drive OTL amps. Perhaps that too is a myth.

I DO know his ESL included dynamic woofers (for 100Hz down). Roger told me he considered the bass of the Quad ESL to be inadequate, that the bass panels had a resonance which could not be eliminated. As I said, he advised me to use a symmetrical 24dB/octave filter when employing a sub (I’m inclined to consider a driver coming in at 100 Hz for use with a planar as a woofer, not a sub, but that’’s just a matter of semantics I suppose) with the Quad, necessary to make the resonance inaudible. As his own ESL included a "sub", and as he told me he considered the Quad’s bass reproduction to be inadequate, and was adamant that the x/o filter be steep, I guess I just came away with the assumption that he used a dynamic woofer with his Quads. I am in fact surprised he didn’t. For Quad owners looking for a woofer, the Rythmik plate amps include a 24dB/octave settings on their crossovers.

Roger stated in his Music Reference AudioCircle postings that of the MR amps returned to him for repair, it was most often as a consequence of botched modifications, usually for nothing more than the replacement of his stock parts with "magic" parts of the same value (including "magic" fuses ;-). When looking for MR amps I therefore always made sure the ones I bought were unmolested, 100% stock. I got my RM-9 from Sheila Berdan, and as Brooks was a MR dealer the amp may have originally been bought from him. Kismet? ;-)

As to the "lightweight" sound of the RM-10: when mated with a typical high-sensitivity loudspeaker, that may actually make for a synergistic pairing. Many of them have bass I would characterize as "woolly" (not a fan of the bass-reflex design).

Thanks for the corrections, clarification, and additional information @clio09 (in spite of your subtly churlish tone ;-) .

May I ask you to elaborate on your "Not by a long shot" response to my statement that the ESL/OTL was Roger’s final product? I know he had also introduced a number of single-ended amps in his final years. Or was "Not by a long shot" said in relation to something other than whether or not the ESL/OTL was the last product Roger introduced?

My first hi-fi amplifier was the Fisher X-100-A integrated (bought used in 1968), which used the 7189 tube, an EL-84 variant. A nice place to start ;-) . My first guitar amp was a 1966 Fender Deluxe Reverb, also a sweet little amp.

@pdreher

Let me begin (oh no, here he goes again ;-) by recommending you head over to the AudioCircle site, where you will find a Music Reference section in the Forums. Roger was eventually kicked off it, but before that happened he posted for a number of years, providing his thoughts on a lot of technical issues, including those related to the design of all his amplifier models. Really fascination stuff.

As I previously posted in a thread here on Audiogon, Roger told the story of Tim de Paravicini (one of the few hi-fi amplifier engineers/designers with whom he felt "aligned") telling Roger he could predict the bass response of a power amp by looking at it’s output transformers. It is well known that the more "iron" there is in a transformer, the better will be it’s reproduction of bass frequencies. Conversely, the worse will be the sound of the reproduction of high frequencies. Everything in hi-fi design is a matter of design choices, trade-offs that must be made. Every design choice comes at a cost.

Roger Modjeski was not a "normal" hi-fi company owner. He freely admitted he didn’t enjoy running a business, or even building "product". What he liked was designing, finding solutions to engineering challenges. He stated every amp he offered was an answer to a design challenge, not just another amp. He didn’t design or build an amplifier to have a certain sound, but rather to be stable, linear, and distortion-free. To have no identifiable sound of it’s own. Of course, that is the oldest mantra in hi-fi, but imo he took it more seriously than do most amplifier designers. To hear Roger discussing amplifier design in the flesh, search for the YouTube videos of the talks he gave (three of them) at the annual Burning Amp Festival in San Francisco.

In the RM-10, Roger set out to prove that 35 watts could be created by a pair of EL-84 tubes, twice what anyone else had ever been able to coax out of that tube. Damned if he didn’t do it! And without sacrificing tube life. You can learn how he did it by watching the Burning Amp videos and reading the Music Reference sections in the AudioCircle Forums.

The RM-200 was Roger’s answer to the challenge of designing a tube amp that 1- was optimized for powering low-impedance loads, and 2- produced a hundred watts out of a pair of KT-88 (or 6550) tubes. Again, twice the normal power created by those tubes. I got myself an RM-200 specifically to drive the midrange/tweeter panels of my Magneplanar Tympani T-IVa, a notoriously inefficient planar-magnetic speaker. 100 watts is not nearly enough power for a T-IVa run full range, but the speaker is very easy to bi-amp, something Magnepan urged Tympani owners to consider. Used with a brute amp on the bass panels, the RM-200 is enough for the m/t panels when used in a moderately-sized room.

I got myself an RM-10 for a specific application, one for which the amp is perfect: the Quad ESL. Roger used that loudspeaker and the 16 ohm LS3/5a as his speaker load in the design phase of creating the RM-10, the Quad a notoriously difficult speaker to drive. If an amp can remain stable, linear, and distortion-free when driving the Quad, powering just about any other speaker is a breeze. Not an inefficient, low-impedance speaker, of course (for that he offered the RM-200).

As for the bass characteristics of the RM-10, that was immaterial to me. I have always used subs with my Quads, as did Roger. Removing the low frequencies from the bass panels of the Quad and the amplifier driving them results in a significant improvement in the sound of the combo, PROVIDING one has a sub/woofer of sufficient quality. I have the best: the Rythmik/GR Research OB/Dipole Sub, THE woofer for any and all dipole loudspeakers.

The RM-10 and RM-200 came years after the RM-9, Roger’s first and most "conventional" amplifier. Roger started his hi-fi electronic career the same way Bill Johnson of ARC did: by repairing broken stuff. Johnson owned a repair shop, Roger worked in one while attending college. Roger made a point of investigating and analyzing the design of every amp he repaired, taking note of what caused them to fail. Building a dependable, trouble-free amplifier was a very high priority for Roger. In the RM-9, Roger’s goal was to create the most stable, linear, distortion-free push-pull/ultra-linear tube amp he could. One that provided long tube life, and of course superior sound. In his review of the RM-9 Mk.2, Dick Olsher asked the rhetorical question: "When we have the RM-9, who needs the McIntosh MC-275?" I haven’t owned a lot of amps, but I can tell you the RM-9 is a better amp than the ARC’s I owned in the 1970’s (of course that was a long time ago, and ARC designs have greatly evolved).

In the 1980’s (I believe it was) Roger was hired by Harold Beveridge to design and build the tube amps included in the Beveridge ESL. Roger had previously discovered and become intrigued by the Futterman OTL amplifier, and he worked hard at solving the engineering problems that amp design presented. Roger’s final product was not just his own ESL loudspeaker, but one with no input transformers (the cause of a lot of non-linearity). Already an unheard of idea! But Roger went further, designing and building an OTL amp specifically to drive his hand-built, transformer-free ESL panels.

Think about it: the output tubes of an OTL amp directly driving the ESL panels themselves. Talk about transparency! By the time the ESL/OTL had gone into production, Roger had moved from Santa Barbara up to the Bay Area. I was out in Palm Desert, and was unable to make it up to his new operation to hear his masterpiece before Roger’s illness made that impossible. The biggest regret of my hi-fi life.

Nor will I @car123. Or my RM-9 and RM-200, for that matter ;-) . Brooks Berdan had a fairly wealthy clientele, and he was more than happy to sell them VTL and Jadis amps (cha-ching ;-). But he recommended the Music Reference amps to his clients who wanted not just great sound, but great value as well.

Music Reference owners tend to keep their amps a long time; buyers of many other brands seem to always be looking for something better, constantly changing in hopes of finding whatever it is they are looking for. Roger Modjeski was a musician, and his priorities in reproduced sound closely aligned with mine. As did his taste in music.

Roger felt that the electrostatic loudspeaker design was the over-all best reproducer of recorded music. He was a longtime QUAD ESL owner, and used that speaker as his load in the development of the RM-10. His last product was his own ESL, with his own woofer design for 100Hz and below. He advised me to use symmetrical 24dB/octave filters with my QUAD ESL/sub setup.

I couldn’t agree more @dhcod. Brooks was also an automobile enthusiast, and had a nice little stable of ’em. Riding with him in his souped up Mercedes Benz to Vegas for the CES was an experience I’ll never forget!

In his younger days (high school) Brooks played drums, and when the band I assembled to play at his 50th birthday party (at the behest of his wife Sheila), he got up on my set and played "Wipe Out" with the band. Happiest I ever saw him! I feel like part of me died with him.

While Dick Olsher preferred the original RM-9 fitted with KT88’s in place of the standard EL-34’s, in the Mk.2 version he preferred the stock EL-34’s, finding the KT-88’s to imbue the Mk.2 with "a subtle, bright tinge through the upper mids and lower treble, and to reproduce the lower mids and upper bass less fully fleshed out." Now THAT sounds like the description of a typical solid state amp!

By the way, SkyFi Audio currently has an RM-9 (original) in stock, priced at $3499. For those who found tube amps to go through tubes too quickly (ARC, anyone? ;-), in the RM-9 the EL-34’s are biased at only 30 milliamps, resulting in low plate dissipation, therefore cool running tubes with a typical lifespan of about 10,000 hours. 10,000!

But Roger ended up, I believe, liking the RM-10 even more than the RM-9. At the talk he gave at Brooks Berdan Ltd to introduce the RM-10, I asked Roger if he recommended a single RM-9 over a pair of RM-10’s for use with stacked Quads (my speakers at the time), and he told me a single RM-9, as it’s greater power output would better prevent overload/clipping with the demanding load stacked Quads present an amp.

The next time I saw him he told me he had been thinking about my question, and had changed his mind. I know he had a pair of Quads himself, and after living with the Quad/RM-10 combination had come to REALLY like their synergy. In my humble opinion, the RM-10 is THE amp for the Quad ESL.

Roger was an enthusiastic proponent of using a pair of subs with the Quad, crossed over at 100Hz with a steep filter (symmetrical 24dB/octave). If you’re going to do it, look into the Rythmik/GR Research OB/Dipole Sub. In my opinion, THE sub for use with the Quad, or any dipole loudspeaker, for that matter.

@dhcod: If you want to experience a hot amp, get a pair of mono Atma-Spheres. A pair of M-60's run 16 output tubes (in Class A), for a mere 60 watts//ch!

I’d be interested to know why a person would prefer the original RM-9 to the Mk.2 version. Dick Olsher (a favorite audio critic of mine) reviewed both, greatly preferring the Mk.2. Both reviews are available for reading in the Stereophile archives, I believe (I have a complete collection of the digest-size issues, my first subscription issue being that of Autumn 1971. It was J. Gordon Holt who changed my hi-fi life, not Harry Pearson ;-) .

In his Mk.2 versions of the RM-9, RM-10, and RM-200, Roger Modjeski successfully endeavored to lower the distortion and increase the linearity of the circuits of all three amps, resulting in improved sound quality. I know some listeners actually like the sound of high-distortion tube amps (Herb Reichert? ;-) , hence the popularity of single-ended designs in some circles.

Congratulations @dhcod! You are in a select group of lucky music lovers who have the pleasure of owning a Music Reference amplifier.

Music Reference RM-10 Mk.2.