Distortion free amp

[kach22i]

The amps Industrial Design is boring and understated but that is no crime.

The speakers are much sexier.

 

[JackD201]

Swedish high-end audio manufacturer Marten generally stick to speakers, but their latest work, the attractive M-Amp monoblock, is their first foray into the world of powered amplifiers. But are slick looks and distortion-free sound reproduction worth $45,000/pair to you?

According to Stereophile, this pair of amplifiers produces sound that one would expect from a more technically capable setup. And Ultimate AV Mag describes the M-Amp's internal workings as such:

The fully balanced M-Amp is based on Marten's proprietary Adaptive Modulation Servo (AMS) technology, which is described as an "inverting, globally self-oscillating, class-D stage" as well as a "load-independent, wide-bandwidth, first-order modulator, switching at 600kHz, with a constant loop gain of 30dB all the way up to the switching frequency."

gizmodo

What panels flank the loudspeakers? They look interesting but because they are clear, they are tough to make out. I can't tell if what's on the front wall are the same panels or sheer curtains either.

 

[DonH50]

I like this quote from the ML website: "The Nº53 is rated at an impressive 500 Watts into 8 Ohms, measures merely 21" H × 9" W × 21" D and weighs only 135 lbs — certainly playing up the strengths for which switching power amplifiers are known." Merely and only, hmmm...

Anyway, the amp's specs look like the efficiency is pretty good even with a linear supply (nearly 80% into 8 ohms); pretty impressive! I am struggling a bit with the 640 W input and 1000 W output into 4 ohms (New Math? No PFC?) It sounds like they interleave four amps to get the effective switching frequency to 2 MHz; I would expect spurs at the 500 kHz fundamentals but that's still pretty high.

 

[amirm]

I like this quote from the ML website: "The Nº53 is rated at an impressive 500 Watts into 8 Ohms, measures merely 21" H × 9" W × 21" D and weighs only 135 lbs — certainly playing up the strengths for which switching power amplifiers are known." Merely and only, hmmm...

As they say, everything is relative . I think they are comparing it to likes of Boulder at 400+ pounds.

Anyway, the amp's specs look like the efficiency is pretty good even with a linear supply (nearly 80% into 8 ohms); pretty impressive! I am struggling a bit with the 640 W input and 1000 W output into 4 ohms (New Math? No PFC?)

Unfortunately, their web site is full of mistake and lacking in information. Fortunately, there is a white paper with details of design: http://173.203.156.112/Uploads/File...Technology Background V5 04032010_5.17.10.pdf

    Power Output  ?53        | power consumption  | Equivalent Class AB amplifier 
       500 watts into 8 ohms | 785 watts          | 995 watts 
     1,000 watts into 4 ohms | 1,465 watts        | 2,085 watts

It sounds like they interleave four amps to get the effective switching frequency to 2 MHz; I would expect spurs at the 500 kHz fundamentals but that's still pretty high.

I can't explain why but the spurs are at 800 Khz:

And the spurs on a 100 Khz class-d amp:

 

[DonH50]

Ah, at least the power numbers make sense, thanks! I need to read the white paper as the concept seems interesting.

Reading the ad copy, it appears they may use a quasi-differential setup per side, suppressing the fundamental but leaving the half-rate subharmonics from the two sides. And. maybe 1.9 MHz is close to 2 MHz, thus the spurs and 0.8 MHz. Maybe -- speculating.

Those are pretty low spurs, at least relative to your other picture, with the implication that they have really fine-tuned and dialed in those amps. Still, somebody is bound to complain that -26 dB or so spurs at 800 kHz are causing some sort of audible distortion, assuming their tweeters can handle it! (I'd be more worried about the tweeters frying if that isn't filtered.)

Neat stuff, thanks Amir! - Don

Edit: Skimmed the paper: that is comparing a single 100 kHz amp to an interleaved-by-eight system, so the 800 kHz spur field makes sense. Comments about matching and suppressing the individual amps' switching noise still apply -- an impressive system!

 

[amirm]

Oh, I see. So the 800 KHz system was hypothetical, not that of No 53. The spurs of the 53 then must be at 2 MHz.

Did you read the clever part about building a model of the amp and then comparing it to the real one? I am going to find out more about that technique. Pretty clever.

 

[DonH50]

Well, macromodeling as well as detailed (transistor-level) simulation and comparison to lab tests have been a part of my world for ages so I guess I didn't pay much attention. Building a good macromodel is hard if you want a lot of detail and tight correlation to reality.

Interesting they show a magnitude voltage, not dB, vertical axis. Probably because people would freak seeing the low-in-dB suppression, not realizing it was pre-filter etc.

 

[Gregadd]

This might be a good point to bring in the technical brain amp and preamp. When paired with the TAD speakers by Pioneer indded represented something special. Transformer from tba.

JV talks about technical brain. http://www.avguide.com/blog/high-end-audio-japan-part-four-solid-state-technical-brainhttp://www.technicalbrain.co.jp/en/ssreview.pdf

 

[egidius]

I own an Audiopax M5 3A preamp, and it is a hell of an experience; not only because how good it can sound, but how long you need to suss it out!
Eduardo de Lima explained it to me, the desirable thing is not zero distortion in one single unit, but all the interrelated small distortions in all these highend high price gear adding up to something - which the M5 then can cancel out. Nice theory, but somehow it strikes me as more realistic approach. And it does indeed make a difference. Highly recommended reading http://www.audiopax.com/eng/tecnologia/tecnologia.htm

e