Rogue Audio M-150 Monoblock Tube Power Amplifier
An intelligent and reliable design.
Review By Dick Olsher
here to e-mail reviewer.
What do the following items
have in common: output transformer, ultralinear connection, triode mode,
KT-88, phase-splitter, fixed bias, feedback, and a Suzuki SV650 racebike?
Most of us would recognize and categorize all but the last item as tube
amplifier design elements. Then you might ask what a racebike has to do
with a high-end audio product such as the M-150? Rogue Audio's Mark O'Brien
has been an avid bike racer for most of his life, though he tells me that
Rogue Motorsports is in a bit of a transition spurred by injuries to a
couple of racing friends. He's giving up on bikes and plans to do some
(much safer) SCCA car racing next season. The point is that the mindset of
racing in terms of parts and build quality is very much akin to that of
high-end audio. And from where I sit it's clear that Mark has been able to
import excellence in design and execution into his audio products.
In my opinion, there are far too many 40 to 60 wpc KT-88
based stereo tube amplifiers out there. Most of them sound euphonic, lack
adequate bass control, and need to be carefully matched to a prospective
speaker load. What I was looking for was a tube power amplifier with
adequate power reserve that could partner any real-world speaker load,
extract maximum bass definition from a typical bass reflex speaker, and
reproduce harmonic colors and textures without significant editorializing.
And just as important, a price tag well below the cost of new automobile.
It's fair to say that the M-150 monoblocks easily fulfilled my wish list.
And I'm inclined to think that 150 wpc (UL mode) also fulfills the
Goldilocks principle — it is just right for all the speakers I have on
hand. The icing on the cake is superb cosmetics and build quality. How
about that 0.5-inch machined and engraved faceplate!
Mark O'Brien started designing his own tube gear
about 20 years ago. Having worked at Bell Labs in engineering physics he
had access to superb lab resources as well as lots of help from
electronics engineers who had received their degrees during the golden age
of tube electronics. It's obvious to me that Mark has learned his craft
well. It starts with the monoblock layout. It's nearly impossible for me
to accept an amplifier as Reference class unless it is of monoblock
construction. Two totally independent and isolated power supplies and
separate signal paths result in ideal channel separation and stable
imaging under dynamic drive conditions.
the single-ended (RCA) and balanced (XLR) inputs are connected to a
wide-bandwidth Jensen input transformer. Although it's strictly not
necessary in the case of the RCA input, Mark believes that it does a neat
job of eliminating potential ground loops and smoothing out the top end.
One leg of the secondary is grounded, so that only the positive going
portion of the balanced signal is used. The signal is then fed to a
long-tailed pair phase inverter consisting of a 12AX7 dual triode. The
driver stage is more sophisticated than what one typically sees in a 1950s
vintage amp. It's a series-regulated push-pull circuit, also known as a Mu
follower or totem pole, which uses a pair of 12AU7 dual triodes. The
output stage is comprised of two pairs of KT88s operating in parallel
push-pull configuration. The output transformers are quite beefy and
feature an ultra linear connection. They use E-I laminate cores of
grain-oriented silicon steel and are custom wound to a frequency response
spec of 10 Hz to 50 kHz (+/- 1dB). The windings are bifilar and are
interleaved to reduced leakage inductance. The output stage mode may
changed on the fly from UL to triode by the mere flick of a switch on the
back panel. Of course, maximum output power is reduced by about one half
in triode mode.
Instead of the more common cathode bias, a fixed bias
scheme is used. There are no cathode resistors, a dedicated –100 VDC
bias supply is provided. This is a far more efficient biasing scheme than
self bias since the entire B+ voltage is useable as plate voltage. A very
nice touch is the built-in current meter. The quiescent current of each
output tube may be adjusted individually using a set of pots located under
a plate near the meter. A special tool is provided for that purpose. The
entire process only takes a couple of minutes to complete if there are a
few pots that need tweaking. Note that the bias is set on the low side at
the factory as a precaution because it is a function of the AC main
voltage. The recommended bias current is 40 mA and it should be set
accurately only after each amp has been running for at least one hour. One
knock against self bias is that it is not self-regulating as is cathode
bias with tube aging. I found it necessary to tweak the bias a couple of
time during the first week of use. After that, there was little drift; I
would recommend a once-week bias check under heavy usage conditions and a
bimonthly frequency for the average user.
The power supply is solid-state rectified and features
voltage regulation for the front end preamp tubes. The other important
feature, mandatory really, for a fixed-bias amp is the soft start circuit
which ensures that high voltage is applied slowly while the bias supply
kicks in. Each output tube is individually fused as an added safety
feature to prevent runaway currents should the bias supply be interrupted.
The M-150 is normally shipped with Electro-Harmonix
KT88s and vintage US brand preamp tubes. My samples were outfitted with GE
12AX7 and 12AU7 tubes. Rogue Audio would prefer to use current
production preamp tubes, but have been dissatisfied recently with the
quality of the Russian-made 12AU7. The output stage plate voltage is said
to be near 600 VDC, which means that for a quiescent current of 40 mA,
plate dissipation is about 22 watts — well below the rated maximum. At
this operating point, KT88 lifetime is expected to be in the range of
2,000 to 3,000 hours.
Both 4 and 8-Ohm impedance taps are provided. These
ratings should not be confused with the amp's internal source impedance
which determines its damping factor. The source impedance is said to be
about 0.4 Ohm — quite low for a tube amplifier with moderate levels of
global feedback (about 18 dB) — and resulting in a damping factor of 20
into an 8-Ohm load. I'm so fed up with insulated EU style binding posts
that look fancy but are difficult to tighten. Hence, it was really nice to
see good old US style un-insulated posts of hexagonal profile that I can
really tighten down on with a nut driver.
My strategy for evaluating any power amp is to
audition it with several speaker loads in order to obtain a balanced
overview of its performance potential. Otherwise, the review process runs
the risk of merely becoming a hit or miss proposition. The M-150 was first
teamed up with the Esoteric MG-20, a speaker of fantastic mini-monitor
like imaging capability. In fact, it really raises the bar in this regard.
First order of business was to experiment with impedance tap options and
UL vs. Triode modes. A valuable lesson I re-learn periodically is how
critical it is to try all of a tube amplifier's impedance taps with a
given loudspeaker. Most audiophiles simply select impedance taps that
match the nominal impedance of their speakers. That could be the optimum
choice, but not always. And in the case of the MG-20 with its 6-Ohm
nominal rating, the optimum choice isn't clear. My finding was that
performance off the 8-Ohm taps was surprisingly more dynamic vs. the 4-Ohm
taps with sweeter sounding mids and a fuller bodied tonal balance.
It was a closer call in selecting the operating mode.
Off the 8-Ohm taps, Triode mode gave a more relaxed presentation and
timbre accuracy was judged to be a bit better relative to UL. It would
appear that in Triode mode the distortion spectrum becomes a bit richer in
second order harmonic distortion and would explain the observation that
midrange textures were sweeter sounding. But, and this is significant, I
missed the microdynamic intensity and drive of UL. The music's raw energy
was slightly tamed, more polite if you will in Triode mode. Ultimately it
seemed to be a choice between pretty vs. exciting sound, though I admit
that my preference for a particular mode was also influenced by the choice
of music. Therefore, you should feel free to experiment in this regard as
it boils down to flicking a single switch on the back panel.
The M-150 needed about 30 minutes of warm-up to sound its
best, after which its essential characteristics of clarity, image focus,
and transient speed shone through in spades. Teamed up with the Prima Luna
model Eight CD player, which erects a 3-D soundstage like no other player
out there, the resultant soundstage was not only cavernous but featured a
nicely layered depth perspective. Image outlines were fleshed out with
palpable extension, but did not quite equal the
reach-out-and-touch-someone realism of the much more expensive ($19,000 to
be precise) Esoteric model A-100.
My first impression, which held up during subsequent
auditions, was that at least with the stock tube complement the M-150 was
not particularly tubey sounding. It didn't overly liquefy midrange
textures, by virtue of its extended bandwidth, and it didn't romanticize
the lower midrange. I could do without the euphonics, but I did wish for a
more vivid harmonic color palette and a smoother presence range. There was
no tube glare or gratuitous brightness, but a bit of textural grain broke
through occasionally. It was time for tube rolling. In my experience, tube
substitution is a necessity rather than a luxury with most tube
amplifiers. My first avenue of attack was the 12AX7 in the phase splitter
circuit because there was only one tube in this circuit and it was the
first one in the signal path. My two favorites in this spot turned out to
be the Telefunken “smooth plate” and the Sylvania Gold-pin 5751, which
Mark O'Brien assured me would work fine in this circuit. The Telefunken
extended the imaging virtues of the M-150 to a new level. It engendered a
wonderful sense of space, a great feel for hall reverb, creamy smooth
midrange textures, and expressive microdynamics. It was clear that the
Telefunken was a truly great choice in this circuit. The Sylvania 5751 was
also a winner, by virtue of its airy treble range, pure textures, and
tight image outlines. Overall, I preferred the Sylvania by a slight margin
due to its more neutral balance vs. the darker presentation of the
Telefunken. But I did not achieve sonic nirvana until I replaced the
driver circuit 12AU7s with NOS Mullard CV4003 box anode types. It took me
a while to accumulate two pairs of the Mullard, but it was worth the wait.
This is probably the finest 12AU7 type available today and it took care of
the remaining items on my wish list; namely, more vivid harmonic colors
and a smoother presence range. Now that's the way to make the KT88 output
stage sing! The M-150 now delivered not only speed and stress-free
dynamics, but also plenty of finesse. Complex passages were fully resolved
and the music boogied forward with plenty of verve.
I've waited till now to reveal one truly amazing aspect
of the M-150's performance. The MG-20, being a bass reflex design, had
previously evinced fully satisfactory bass control and pitch definition
only when driven by solid-state power amps such as the Silicon Arts ZL-120
monoblocks and James Bongiorno's Son of Ampzilla 2000. The M-150,
especially in UL mode, grabbed hold of bass lines with an iron fist and
dug deep with excellent extension. Its high damping factor was responsible
for convincing bass control not only with the MG-20 but also with the two
other speaker loads I threw at it: my own DIY Basszilla Platinum Edition
Mk. II, also a bass reflex design, and Final Sound's 1000i ESL.
With the Basszilla Platinum Edition, I preferred the
8-Ohm impedance taps. The Triode mode sounded more tubey, more mellow, and
less crisp than the UL mode. Most of the time I preferred the UL setting
because of its enhanced speed and immediacy. With the 1000i ESL, I
preferred the 4-Ohm taps and Triode mode for its smoother treble and
sensuous mids. At no time did the M-150 have any difficulty driving the
1000i to satisfying levels. However, the resultant tonal balance was a bit
dark, lacking in treble energy. I suspected that the M-150 was rolling off
the 1000i's treble, due to the interaction of the amp's source impedance
with the low load impedance, which touches 1.2 Ohm at 20 kHz. The measured
frequency response of the 1000i driven by the M-150, with the measurement
microphone centered on the tweeter panel, showed a treble rolloff starting
at 10 kHz. The roll off reached a level of 3 to 4 dB at 20 kHz and implies
an effective source impedance closer to 1 Ohm. To be fair, other tube
amplifiers and even some solid-state amplifiers also exhibit similar load
interactions with the 1000i ESL.
M-150 is an intelligent and reliable design, which I'm very fond of from a
technical perspective. But that's not all. Outfitted with the right preamp
tubes, it is a high-resolution muscle amplifier combining finesse and
brawn under one chassis. Its bass performance is astounding for a tube
amplifier. And it has the power reserve to cope with most real-world
speaker loads. The M-150 is absolutely a world-class amplifier, and
considering its performance and build quality, its asking price is all the
more remarkable. Highly recommended!
Type: Monoblock tube amplifier
P.O. Box 1076
Brodheadsville, PA 18322