Lamm Industries ML3 Signature monoblock power amplifier Measurements

Sidebar 3: Measurements

Before performing any measurements, I ran one of the Lamm ML3 Signature amplifiers (serial no. G10029) for an hour at 1W into 8 ohms from its 4 ohm tap. At the end of that period, I muted the input signal and checked the plate current of the GM70 output tube using the two test points on the top panel and Lamm's recommended Fluke 87 multimeter. The wall voltage was 119.3V; the test-point reading was 1.172V DC, which is out of the recommended range of 1.078–1.122V. Using the multi-turn potentiometer that's visible through a circular cutout in the top panel, I reduced the plate current until the reading was 1.103V.

That adjustment made, I performed full sets of measurements from each of the ML3's three output-transformer taps, using Stereophile's loan sample of the top-of-the-line Audio Precision SYS2722 system (see www.ap.com and the January 2008 "As We See It"). I then repeated some of the tests with the output stage's negative feedback (NFB) switches set to their two positions. As the XLR input is provided for convenience but simply parallels the single-ended RCA jack with pin 3 connected to ground, I used the single-ended input exclusively for the tests.

The ML3's input impedance was the specified 41k ohms at 20Hz and 1kHz, dropping to 17k ohms at 20kHz. This drop is most likely inconsequential. The amplifier preserved absolute polarity (ie, was non-inverting) from all three transformer taps. The voltage gain at 1kHz, measured into 8 ohms, was 23.6dB from the 4 ohm tap, 25.3dB from the 8 ohm tap, and 26.2dB from the 16 ohm tap. Looking at the effect of the NFB switches, NFB2 reduced the gain into 8 ohms from the 4 ohm tap to 21.6dB, NFB1 to 20.4dB. With the input shorted and assessed from the 8 ohm tap, the ML3's wideband signal/noise ratio (ref. 2.83V into 8 ohms) was okay at 63.8dB. Restricting the measurement bandwidth to the audioband increased the ratio to a good 86.1dB, while A-weighting increased it further, to 94.9dB.

Even without any negative feedback, the 4 ohm tap's output impedance was moderately low for a single-ended design, at 1.55 ohms at low and middle frequencies, rising to 1.9 ohms at the top of the audioband. As a result, the modification of the amplifier's frequency response by the Ohm's law interaction between that output impedance and the impedance of our standard simulated loudspeaker remained within ±0.9dB (fig.1, gray trace). Applying 1.2dB of NFB reduced the output impedance from this tap to 1.25 ohms at 20Hz and 1kHz, 1.4 ohms at 20kHz; 2.4dB of NFB further reduced the impedance to 1.1 and 1.2 ohms, respectively.

913Lammfig01.jpg

Fig.1 Lamm ML3 Signature, 4 ohm tap, frequency response at 1V into: simulated loudspeaker load (gray), 16 ohms (green), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red) (1dB/vertical div.).

Without NFB, the output impedance was much higher from the other taps, the 8 ohm tap measuring 2.9–3.9 ohms, the 16 ohm tap 6–8.5 ohms, both depending on frequency. Consequently, the modification of the frequency response with our simulated loudspeaker was more severe than from the 4 ohm tap, at ±1.5dB from the 8 ohm tap (fig.2, gray trace) and ±2.4dB from the 16 ohm tap (fig.3, gray trace). The response rolls off above the audioband from all taps, causing an increase in the risetime of a 10kHz squarewave (fig.4). But the commendably extended extension at low frequencies seen in figs.1–3 leads to an impressively accurate reproduction of a 1kHz squarewave, with flat waveform tops and bottoms (fig.5).

913Lammfig02.jpg

Fig.2 Lamm ML3 Signature, 8 ohm tap, frequency response at 1V into: simulated loudspeaker load (gray), 16 ohms (green), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red) (1dB/vertical div.).

913Lammfig03.jpg

Fig.3 Lamm ML3 Signature, 16 ohm tap, frequency response at 1V into: simulated loudspeaker load (gray), 16 ohms (green), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red) (1dB/vertical div.).

913Lammfig04.jpg

Fig.4 Lamm ML3 Signature, 4 ohm tap, small-signal 10kHz squarewave into 8 ohms.

913Lammfig05.jpg

Fig.5 Lamm ML3 Signature, 4 ohm tap, small-signal 1kHz squarewave into 8 ohms.

The GM70 tube is run at 1200V on its plate to get it to operate in the most linear region of its transfer function. Even so, it is not that linear, especially into low impedances. Although I measured how the percentage of THD+noise changed with output power from all three output-transformer taps, I have shown only the results from the 8 ohm tap, into 16, 8, 4, and 2 ohms (figs.6–9); the overall picture was the same from the 4 and 16 ohm taps. With the load impedance twice the nominal value of the transformer tap, the THD starts to rise above the noise floor at a few hundred milliwatts, then rises linearly with power until the amplifier starts to clip at around 10W, reaching 3% THD at 25W. As the load impedance drops, the THD+N is higher at lower powers, and the power at which the THD reaches 1% drops. With the load half the transformer-tap value, the power at 3% is around 10W; and with a further halving of the load impedance, the ML3 is nonlinear even at low powers. Fig.10 shows the ML3 at its linear best, with the 4 ohm tap driving 16 ohms. The small-signal distortion is impressively low for a single-ended amplifier operating without the palliative effect of negative feedback. The THD+N remains below 0.05% until the output power reaches 4.5W, with the THD reaching 1% at 12W (13.7dBW), and 3% just below 20W (16dBW).

913Lammfig06.jpg

Fig.6 Lamm ML3 Signature, 8 ohm tap, distortion (%) vs 1kHz continuous output power into 16 ohms.

913Lammfig07.jpg

Fig.7 Lamm ML3 Signature, 8 ohm tap, distortion (%) vs 1kHz continuous output power into 8 ohms.

913Lammfig08.jpg

Fig.8 Lamm ML3 Signature, 8 ohm tap, distortion (%) vs 1kHz continuous output power into 4 ohms.

913Lammfig09.jpg

Fig.9 Lamm ML3 Signature, 8 ohm tap, distortion (%) vs 1kHz continuous output power into 2 ohms.

913Lammfig10.jpg

Fig.10 Lamm ML3 Signature, 4 ohm tap, distortion (%) vs 1kHz continuous output power into 16 ohms.

Fig.11 shows how the THD+N percentage changes with frequency at low power with the 4 ohm tap driving loads from 2 to 16 ohms. (The behavior from the 8 and 16 ohm taps, figs.12 and 13,, is very similar but with increasingly higher distortion overall.) Again, the amplifier is most linear when the load impedance is at or above the nominal transformer-tap value. The distortion starts to rise in the low bass, due to the transformer core starting to saturate, but the rise is low in absolute terms. This is a good transformer. The effect of the two NFB settings is shown in fig.14: each increase in the amount of NFB reduces the level of midband distortion by about 20%.

913Lammfig11.jpg

Fig.11 Lamm ML3 Signature, 4 ohm tap, THD+N (%) vs frequency at 1V into: 16 ohms (green), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red).

913Lammfig12.jpg

Fig.12 Lamm ML3 Signature, 8 ohm tap, THD+N (%) vs frequency at 1V into: 16 ohms (green), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red).

913Lammfig13.jpg

Fig.13 Lamm ML3 Signature, 16 ohm tap, THD+N (%) vs frequency at 1V into: 16 ohms (green), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red).

913Lammfig14.jpg

Fig.14 Lamm ML3 Signature, 4 ohm tap, THD+N (%) vs frequency at 1V into 8 ohms with 0dB NFB (red), 1.2dB NFB (magenta), 2.4dB NFB (blue).

As is typical of single-ended tube amplifiers, the ML3's distortion signature is heavily second harmonic (fig.15), which will work against the audibility of the circuit's fundamental nonlinearity. However, even at low powers, the spectrum of the ML3's output has spurious components visible at 60Hz and at both its even- and odd-order harmonics (fig.16), with the odd-order components (other than the third harmonic) each lying close to –80dB (0.01%). Even so, the third harmonic of the AC supply lies at –77dB (0.014%). The same behavior can be seen in fig.17, which shows the spectrum of the ML3's output while it reproduces an equal mix of 19 and 20kHz tones at 1 W into 8 ohms from the 8 ohm output tap.

913Lammfig15.jpg

Fig.15 Lamm ML3 Signature, 4 ohm tap, 1kHz waveform at 1W into 8 ohms, 0.199% THD+N (top); distortion and noise waveform with fundamental notched out (bottom, not to scale).

913Lammfig16.jpg

Fig.16 Lamm ML3 Signature, 8 ohm tap, spectrum of 50Hz sinewave, DC-1kHz, at 1W into 8 ohms (linear frequency scale).

913Lammfig17.jpg

Fig.17 Lamm ML3 Signature, 8 ohm tap, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 1W peak into 8 ohms (linear frequency scale).

The measured performance of Lamm's ML3 Signature may well raise eyebrows, but it actually measures well for an amplifier with a single-ended output stage. The measurements do show that it is important to use the lowest-possible output-transformer tap with any particular loudspeaker. But when that is done, the ML3 sounds more powerful than it has any right to. When I drove up to Michael Fremer's place to pick up the Lamms, and before we packed them up, I listened to my rough mixes of an Attention Screen concert with the ML3s driving his Wilson Alexandria XLF speakers, recordings that I know intimately well. I have listened to these speakers in previous visits to Michael's, when they were driven by Dan D'Agostino Momentums and darTZeel NHB-458s; as good as those solid-state monoblocks sounded with the Wilsons, I must say that the Lamm ML3s driving the XLFs was magical.—John Atkinson

COMMENTS
popluhv's picture

For some reason these amps really piqued my interest when they came out (always been a fan of the laboratory aesthetic), but I could find little in the way of a review other than show reports.  What really made my day was to find Jansch and Renbourn were used for the review; always nice to match the calibre of the music with the system! 

MVBC's picture

$140,000 in order to get stock metal boxes, no significantly special rare component, basic cabling inside, that ultimately will bring "more-than-adequate" bass and yet will be able to ruthlessly discriminate between  speakers cables, speed light % or whatever.

And we're warned it's nowhere near the real thing... despite its muscular 32w feeding $200,000 average 92dB/w/m sensitivity speakers! Who would have guessed?

At this rate, it'll be cheaper to buy a lifetime of live concert tickets and live forever after with the memory of these performances! crying

John Atkinson's picture

MVBC wrote:
And we're warned it's nowhere near the real thing... despite its muscular 32w feeding $200,000 average 92dB/w/m sensitivity speakers! Who would have guessed?

It's all too easy for someone to carp from the sidelines, and yes, these amplifiers are silly-priced. But until you have actually listened to them driving the Wilson speakers, as Michael and I have done, your opinion is meaningless.

John Atkinson

Editor, Stereophile

MVBC's picture

these amplifiers are silly-priced

You said it. So why don't you provide information such as the real cost of designing and building one unit and a thorough investigation to what should be the price as a result since you acknowledge the present price is silly?

Nothing in this open chassis suggests anything special in manufacturing or parts. A set of bulbs at about $120 is rather mundane... That this amp would sound good is the least this stuff can do!

But what is also too easy is to flog outrageous priced product and what amounts to a large slack on the technical level (euphemism such as more-than-adequate, in spite of bloomier, less than that attacks etc... when elsewhere you are counting the 0.0001%) given to this product as an argument of authority. Information over promote.

John Atkinson's picture

MVBC wrote:
John Atkinson wrote:
these amplifiers are silly-priced

You said it. So why don't you provide information such as the real cost of designing and building one unit and a thorough investigation to what should be the price as a result since you acknowledge the present price is silly?

I meant "silly" as in super-expensive, not that the price was out of proportion to its manufacturing cost or its sound quality. If you think this amplifier is over-priced, then please, do not buy it.

John Atkinson

Editor, Stereophile

MVBC's picture

I perfectly understand, "super-expensive", so please justify why with information instead of cult peddling. Are the transformers made of pure gold, the capacitances in platinum, since you kindly offered that the bulbs are about $120 bucks... Enlighten us

enlightened

P.S.: For instance, some expensive speaker manufacturer is billeting their enclosures from solid aluminum. I may not buy their product but I indeed appreciate the information and do understand the process would make their product more expensive than particule board kitchen cabinets. Do I make myself clear?  

 

ChrisS's picture

You know the difference between this power amplifier and a kitchen cabinet, but you're asking someone who's not at all involved with the design, manufacture, and marketing of this product to explain and apologize for its cost?

Does that make sense?

MVBC's picture

No I am asking someone who supposedly offers a critical assessment of these products to offer information instead of arguments of authority. If it's to rehash the manufacturer's website, then what's the point? Where is the value to a potential buyer reading that stuff?

gozoogle's picture

"Are the transformers made of pure gold, the capacitances in platinum"

What's with the obsession over the cost of the components?  Who cares what the cost of the components are?  Are you planning to melt it down for scrap cost?  It's like saying you wouldn't pay more for a camera lens that performs better (because of the optics design) if it wasn't made out of expensive fancy materials.  Or, all chipsets/DSPs are just made out of the same materials, right, so who cares about firmware/software?

The reviewer provided a critical, albeit subjective, review, along with objective measurements of the product performance.  It's up to the buyer to decide whether that level of performance is worth the price.  End of story.  If you think another product is better at a lower price, don't buy it.

John Atkinson's picture

MVBC wrote:
I am asking someone who supposedly offers a critical assessment of these products to offer information instead of arguments of authority.

I think you are reading a different review from what Stereophile published. We described this amplifier, informed readers what it cost, discussed how it sounded, and offered a full set of measurements, along with its specifications and details of the system used to audition it.

No, we didn't condemn it because of its very high price. While that will restict total sales of the Lamm amplifier, that alone should not preclude recommendation.

John Atkinson

Editor, Stereophile

MVBC's picture

On the contrary... I never asked you to condemn anything, on price or on whatever else but to explain what in the components, manufacturing etc... can justify such price. I even took an example to make myself clear. Yet, you keep misrepresenting my position. Fine. [flame deleted by John Atkinson]

Doctor Fine's picture

Michael sets up the typical audio ignorant straw man arguement at the outset of the article:  "Unlike Video which has standards such as grey scale, tracking accuracy blah blah blah---Audio has NO standards and is simply whatever the mastering engineer prefers."

Well, no.  Actually this is totally wrong and part of the reason the High End is so full of horse puckey.  Baloney about audio seems to permeate the Domestic "experts" and I for one would like for it stop and for domestic audio to start advancing once again leaving behind the snake oil and mystery which is so easily sold to the misinformed.

There are in fact many standards employed by mastering engineers.  Simply peruse the lastest coffe table marvel from Recording Architecture, a british audio firm which is leading the charge to standardise mix, mastering and post audio playback rooms. 

The list of audio qualities that they enclude in every project is as long as your arm.  It is all designed to avoid the misery of old thinking---where every listening room had a different take on the balance and thus the hapless professional would go mad trying to find consistency in their work from one room to another.  And from one playback type (Pro Audio) to another (Domestic Audio).

Even a superficial investigation of the big shots of the mix world would immediately reveal their dedication to standardising domestic playback results.  The playback accuracy attained by the world renown mix engineers is a direct result of the "correctness" of their studio playback equipment and rooms. It is in fact part of the reason these guys make the big bucks.  Their stuff WORKS.

Please go look on line at this subject.  These guys work their butts off to make consistent standards of excellence which translate across the audio medium.  There is a TON of stuff which audiophiles should be stealing in ideas from the masters in the recording industry.  The pro guys use OUR stuff all the time.  But thats just where they BEGIN.  Their attention to setup is astonishing compared to the slap dash way most domestic audio is built as systems.

Most top pros use better gear the YOU do Michael.  Wilsons are nice but I have seen lots of pro rooms where just the room itself cost twenty times what your entire system costs.

Apparently Michael has also been left out of the loop when it comes to the elaborate reference standards imposed on theater audio by THX.  Mike should visit some of the elaborate designs that this industry builds in the determination to standardise playback fidelity to a high standard of excellence.  Mix rooms may in some cases require the construction of actual movie theaters simply to make no mistake about what the consumer is supposed to hear.

I really do not appreciate this level of mis information coming from a professional audio publication.  Please cut it out.  Michael please go to school and learn what you are talking about.  I love your writing and you have a ton of street level cred and knowledge but this subject has apparently been completely left out of your education with glaring omission.

X