Luxman M-10X power amplifier

As I started to write this review, the news broke that Sound United, the owner of Boston Acoustics, Bowers & Wilkins, Classé, Definitive Technology, Denon, Marantz, and Polk, was going to be purchased by a corporation that makes medical instruments. Such consolidation is not new. China- and UK-based International Audio Group (IAG) was one of the first organizations to acquire iconic audio brands. IAG owns Audiolab, Castle, Quad, Leak, Mission, and Wharfedale. In 2009, they purchased Luxman.

My most recent listening session with a Luxman amplifier was at the end of 2010. I had been auditioning the Japanese company's 80th-anniversary B-1000F solid state monoblocks driving Vivid G1 Giya speakers in the late Wes Phillips's system. I drove the amplifiers to my place to be measured (footnote 1). Listening to some of my hi-rez live piano recordings, I was blown away by the sheer force the massive monoblocks endowed the instrument's left-hand register with. The midrange and high frequencies sounded unforced and natural.

When Luxman America's PR rep suggested that I review Luxman's new flagship solid state stereo power amplifier, the M-10X, I initially declined because the only speakers I had to hand were KEF LS50 and GoldenEar BRX standmounts. While both of these have uncolored midrange and treble regions and offer superbly stable, accurate stereo imaging, neither delivers low frequencies extending much below 50Hz in my room. Luxman was not concerned about the lack of low bass, however, and I took on the review (footnote 2).

The M-10X power amplifier
Priced at $19,995, the M-10X has a luxurious appearance and looks broadly similar to the Luxman L-509X integrated amplifier that Ken Micallef reviewed in May 2018. However, at 106.5lb, it is almost twice as heavy. The beveled front panel features two large, illuminated analog power meters. These are offset to the right, so when the amplifier is used in bridged-mono mode, the active meter will be the one in the center. The Operation button (standby/power) is on the bottom left of the front panel, with a small button to its right selecting between Line (single-ended) and Balanced Line inputs. A second small button turns off the meters' illumination. Two rows of mesh-covered openings run from front to back at the sides of the polished aluminum chassis' top panel, acting as vents for the internal heatsinks.


The rear panel is dominated by two pairs of heavy-duty, five-way binding posts for the speaker outputs. Pairs of single-ended and balanced inputs, on RCA and XLR jacks, respectively, occupy the space in between. The 15A IEC AC power jack is below the inputs; the small pushbutton beside it serves as the main power switch. The AC jack doesn't have a ground connection, but there is a grounding post that can be used in case of hum—although even with no chassis ground, the amplifier was very quiet in my system. One slide switch selects stereo or bridged-mono operation; another allows the balanced input's polarity to be inverted. (I determined in my measurements that the balanced inputs inverted polarity with this switch set to Normal.)

Power inside
The M-10X is specified as delivering continuous power up to 150Wpc into 8 ohms and 300Wpc into 4 ohms (both equivalent to 21.76dBW), the first 12W in class-A. As a bridged monoblock, the Luxman will deliver 600W into 8 ohms (27.8dBW). The maximum instantaneous power is said to be 1.2kWpc into 1 ohm in stereo mode, 2.4kW into 2 ohms in bridged-mono mode.

Each channel's output stage uses eight complementary pairs of Darlington-connected transistors. The power supply features low-loss Kyocera Schottky rectifier diodes, a low-loss EI-type power transformer with flat copper windings, and 80,000µF of filter capacitance. Current is delivered to each channel's output transistors with a thick copper bus bar; another bus bar connects the devices' push-pull output to the speaker binding posts via a high-quality relay. This relay has four switch elements connected in parallel to reduce contact resistance. The power supply rails for the driver stage use high-performance regulator chips and selected Zener diodes, the latter manufactured by Vishay in the United States. The 0.1mm-thick traces on the printed circuit boards are gold-plated.


The older L-509X's output stage featured a version of Luxman's Only Distortion Negative Feedback (ODNF) system. First utilized in 1999, ODNF is said to apply negative feedback only to the distortion products detected at the speaker outputs, thus minimizing the problematic aspects of negative feedback. (Used indiscriminately, negative feedback reduces the level of the subjectively innocuous second and third harmonics but increases the levels of the sonically harmful high-order harmonics, footnote 3.)

According to the new M-10X's design brief, the M-10X uses the Luxman Integrated Feedback Engine System (LIFES1.0) rather than ODNF. LIFES1.0 was developed using circuit simulation software to experiment with alternate layouts and individual circuit components before the design team committed to an actual hardware circuit. This was then optimized with extensive listening tests.

It is implemented with a dual FET (field effect transistor) with an "unusually high transconductance" in the primary stage. This device is used to detect the distortion generated by the output stage. LIFES1.0 is claimed to provide the sound quality of a nonfeedback design.

Footnote 1: Wes's review, the last he wrote for Stereophile, was published in the February 2011 issue.

Footnote 2: I did use Roon's DSP to apply a 3dB boost below 80Hz with both the GoldenEar and KEF speakers for this review. After I have submitted this review to editor Jim Austin, I will be unpacking a pair of full-range floorstanding loudspeakers, and will be using the M-10X to drive them. If I have more thoughts on the Luxman's low-frequency performance with these speakers, I will report them in a followup.

Footnote 3: See here and here.

Luxman Corporation
US distributor: Luxman America Inc.
27 Kent St., Unit 122
Ballston Spa, NY 12020
(518) 261-6464

Axiom05's picture

Is there any way to judge the amount of negative feedback that is used based on any particular measurement?

orosie68's picture

Each JC1+ has 198,000uF of OUTPUT capacitance. The 44,800uF is the INPUT power supply capacitance. Lets get our facts strait.

Tony P's picture

You beat me to it!

Jack L's picture


Sorry, what do you mean by "OUTPUT capacitance" ????

Jack L

Jack L's picture


Please get your "facts strait" when you said "198,000uF of OUTPUT capacitance.

You asked what you meant by it. You ignore my question ! You know or don't know the right answer at all ?

Jack L

orosie68's picture

Do you not understand filter capacitance in power supplies? Why do you keep asking the same idiotic question over and over?

Have another....I am done here.

Jack L's picture


YOU were 100% drunk when you quoted above stupid statement. YOU need a coffee to wake you up.

If YOU were not drunk, your would have stated: "198,000uF FILTER capacitance used in amp's output power section." You expect every consumer readers here knows your drunken use of English ????

I started desigh/builing solidstate power amps some 4 decades back.

FYI, If you knew enough filter capacitors, you should know huge huge huge capacitance is only part of the filter design for better SOUND !
Do you know the AC internal design of a capacitor ??

MattJ's picture

Very impressive, but I think I would still take a Benchmark AHB2!

SteveR1's picture

That was enjoyably well-written and I thought it was one of the better ones from the most recent Stereophile.