PrimaLuna EVO 400 preamplifier Measurements

Sidebar 3: Measurements

When I unpacked the PrimaLuna EVO 400 preamplifier, it looked very familiar. It appears identical to PrimaLuna's ProLogue Premium line preamplifier that Herb Reichert reviewed in our June 2017 issue (footnote 1). But on closer inspection, while the earlier preamp had single-ended inputs and outputs, the EVO 400 adds two pairs of balanced input jacks and a pair of balanced output jacks. And while the EVO 400 still uses two 5AR4 rectifier tubes, there are now six 12AU7 twin triodes compared with the ProLogue Premium's four 12AU7s.

I measured the PrimaLuna EVO 400 preamplifier with my Audio Precision SYS2722 system (see the January 2008 "As We See It"). The gain from the unbalanced inputs to the unbalanced output with the volume control set to its maximum was slightly lower than the specified 10dB, at 9.1dB. From the balanced inputs to the balanced output, it was 7dB. The preamplifier preserved absolute polarity (ie, was noninverting) from all the inputs and outputs. Specified as offering a very high input impedance of 220k ohms, the unbalanced inputs measured 175k ohms at 20Hz and 1kHz, dropping to a still very high 105k ohms at 20kHz. The balanced inputs had a much lower input impedance, at 4k ohms at 1kHz and 3.5k ohms at 20Hz and 20kHz. While this will not be a problem with sources having a direct-coupled, solid-state output stage, it will result in lightweight low frequencies with tubed sources that have a capacitor-coupled output.

The PrimaLuna EVO 400's output impedance is specified as 256 ohms, which is significantly lower than the ProLogue Premium's 2800 ohms. The EVO 400's balanced output impedance measured 1050 ohms at 20kHz, 590 ohms at 1kHz, and 2600 ohms at 20Hz. The unbalanced output impedances were 240 ohms at high and middle frequencies, rising to 2785 ohms at the bottom of the audioband. The EVO 400's unbalanced frequency response into the demanding 600-ohm load (fig.1, cyan and magenta traces) was therefore down by 3dB at 100Hz, compared to –3dB at 15Hz into 100k ohms (blue and red). The high frequencies were flat almost to 20kHz and reached –3dB at 90kHz This graph was taken with the volume control set to its maximum; the excellent matching between channels was preserved at lower settings, and, commendably, the frequency response didn't change. However, for balanced operation, the EVO 400's ultrasonic output peaked by 2.3dB at 40kHz (fig.2). This behavior is typical of a circuit using transformers to convert single-ended signals to balanced—looking at the preamplifier's interior revealed a small coupling transformer for each balanced input and output. Fig.2 was taken into 100k ohms, the specified input impedance of PrimaLuna's new EVO-series power amplifiers, and with the volume control set to its maximum. The peak was preserved at lower setting of the volume control but was reduced in amplitude as the load impedance was lowered.


Fig.1 PrimaLuna EVO 400, unbalanced frequency response at 1V into: 100k ohms (left channel blue, right red), 600 ohms (left cyan, right magenta) (1dB/vertical div.).


Fig.2 PrimaLuna EVO 400, balanced frequency response at 1V into 100k ohms (left channel blue, right red) (1dB/vertical div.).

Channel separation was the same as that of the earlier preamplifier for both balanced and unbalanced operation. It measured >77dB in both directions below 1kHz, but worsened to 50dB at 20kHz, due to capacitive coupling between channels (fig.3). The wideband, unweighted signal/noise ratio, taken with the inputs shorted to ground but the volume control set to its maximum, was 72.4dB ref. 1V (average of both channels). This improved to a superb 91.6dB with a 22Hz–22kHz measurement bandwidth, and to 100.9dB with an A-weighting filter in-circuit. The noise floor was dominated by spuriae at mainly the 60Hz AC line frequency and its odd-order harmonics (fig.4). These are probably due to magnetic interference being picked up by the tubes' steel pins, but these are all very low in level in absolute terms and were identical for balanced and unbalanced operation. Even-order harmonics of the power-line frequency, which stem from internal grounding issues, were even lower in level, which implies excellent circuit layout.


Fig.3 PrimaLuna EVO 400, channel separation.


Fig.4 PrimaLuna EVO 400, unbalanced output, spectrum of 1kHz sinewave, DC–1kHz, at 1V into 100k ohms (left channel blue, right red) (linear frequency scale).

As with the ProLogue Premium, the EVO 400 had relatively high levels of distortion once the output level rose to 1V, even into high impedances. Fig.5 shows how the percentage of THD+noise increases as the output voltage rises into 100k ohms. Although the THD+N is very low below 350mV, the percentage rises almost linearly as the output increases, reaching 0.6% at 2V and 1%—our definition of waveform clipping—at 3.05V. This graph was taken from the unbalanced output; the balanced output behaved similarly, though it clipped at 2.7V. Both these clipping points will probably be above the maximum voltage the EVO 400 will be required to deliver in practical use.


Fig.5 PrimaLuna EVO 400, unbalanced output, THD+N (%) vs 1kHz output voltage into 100k ohms.

As with the ProLogue Premium, the distortion remains constant with frequency (fig.6) as well as with load impedance. But also like the earlier PrimaLuna preamplifier, the EVO 400 is a second-harmonic generator. Fig.7 shows the spectrum of the EVO 400's balanced output—the unbalanced output behaved identically—while the preamplifier drove a 50Hz tone at 1V into 100k ohms. The second harmonic is the highest in level, at –50dB (0.3%), with the third harmonic at –60dB (0.1%). Higher harmonics, however, are all much lower in level. Significantly, the harmonic signature didn't change with frequency, and as the output level is reduced, while the second harmonic remains the highest in level, the third-harmonic content was reduced.


Fig.6 PrimaLuna EVO 400, balanced output, THD+N (%) vs frequency (Hz) at 1V (left channel blue, right red), 300mV (left cyan, right magenta), and 100mV (left green, right gray) into 100k ohms.


Fig.7 PrimaLuna EVO 400, balanced output, spectrum of 50Hz sinewave, DC–1kHz, at 1V into 100k ohms (left channel red; linear frequency scale).

When it came to intermodulation distortion, the "bent" transfer function responsible for the second-harmonic distortion results in a relatively high level of the second-order difference product at 1kHz with an equal mix of 19 and 20kHz tones—0.3% or –50dB (fig.8). However, only a few higher-order products are visible in this graph, and these lie close to –70dB (0.03%).


Fig.8 PrimaLuna EVO 400, unbalanced output, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 2V into 100k ohms (left channel blue, right red; linear frequency scale).

As with PrimaLuna's ProLogue Premium preamplifier, the EVO 400 measures well for a tubed design, other than having a second-harmonic distortion signature that will fatten up the sound.—John Atkinson

Footnote 1: Click here for the review and here for the measurements.
Durob Audio BV
US distributor: PrimaLuna USA
2058 Wright Avenue
La Verne, CA 91750
(909) 931-0219

ok's picture

..a poet and a great reviewer as well.

hrboucher's picture

Hold on, Mr. Reichert. Didn’t you compare the Rogue RP-7 to the Prima Luna with a different amplifier (Rogue Stereo 100) being driven by it? Doesn’t that violate a cardinal rule of equipment reviewing?

Bogolu Haranath's picture

Looking forward to the review of the new Magnepan LRS speakers :-) ........

Would be interesting to see the measurements of the LRS, if Magnepan allows it :-) .......

eriks's picture

Since it's using so many output transformers, I wonder if it has DC blocking capacitors in the output.

jmsent's picture driven by a transformer, but it looks like the single ended output is driven directly from a cathode follower through a coupling cap. My guess is that the transformer primary is directly in parallel with the unbalanced output. So yes, it's probably driven through a dc blocking cap.

Ortofan's picture

... autographing copies of his new book entitled "The Second Harmonic Sauce Diet".

If HR really wants to experience a super-scintillating sounding sense of a recording venue's space and atmosphere, then he should supplement his Harbeth 30.2 speakers with a REL T/7i subwoofer.

Bogolu Haranath's picture

Such a diet would 'fatten up' the midrange :-) ..........

Bogolu Haranath's picture

Music needs illumination :-) .......

mrkaic's picture

Primaluna makes some of the worst looking amplifiers in the industry.

They should learn visual design from Luxman or Leben. Or Quad. Or Line Magnetic. No shortage of good examples.

ok's picture

it certainly looks like a bread box.

Bogolu Haranath's picture

Music is much, much more than skin deep :-) ........

mrkaic's picture

...the amplifier here, not music in some platonic sense. :)

And this amplifier looks ridiculous, at least to me.

Bogolu Haranath's picture

I was just joking :-) .........

Bogolu Haranath's picture

Waiting for reviews and measurements of, PrimaLuna EVO tube DAC and clock, tube integrated, tube stereo and mono-block amps, in Stereophile :-) .........

Bogolu Haranath's picture

'You haven't heard notin' yet' :-) ........

mememe2's picture

"When I squeeze the tube, I notice differences in material density, color saturation". If you believe that you can squeeze a tube of paint and tell the colour saturation (without looking at the brand or price)then you need to go into rehab - but first stop drinking the kool-aid. What a load of horse droppings! And people believe what you write!

johnnythunder's picture

of a magazine. Especially Stereophile. What HR wrote and what he was implying in the context of an audio review was very clear to me. It must be sad and horribly disappointing to judge everything in the world through the closed lens of absolute objectivism.

Ortofan's picture

... drinking the kool-aid. Exposure to paint thinner vapors, in the absence of adequate ventilation, may induce vivid hallucinations.

mrkaic's picture

It’s Reichert and his ideas. Best to ignore and jump straight to measurements.

Bogolu Haranath's picture

It has about 8-bit SNR (resolution) ........ Look at Fig.7 :-) .........

Bogolu Haranath's picture

Also, look at the frequency response into different loads, Fig.1 :-) ...........

Bogolu Haranath's picture

For comparison look at the measurements of recently reviewed Bryston BP-17/3 :-) ...........

grantray's picture

I can do the same with papers and graphite. If you spend a lifetime mastering a medium through sensorial observation and use, most people who don't will call your learned abilities crazy, even though those abilities took decades to develop. For instance, those who are born blind are quite capable of hearing and smelling at levels of spacial detail the rest of us would consider superhuman, even though their noses and ears are nothing special.

Indydan's picture

I enjoyed reading this article. Well done Herb!

romath's picture

Herb - I find the meaning of the term "neutral" in audio depends on the user's perspective or tastes. For some, it's what I would consider "cool," while for others its a modest warmth like that I find comes from acoustic instruments. What I'm wondering in reading your Rogue comment at the end, is what you mean by it.

Other question: in your left-right brain metaphor, were you referring to the intellectual vs. emotional divide, however close these two were in that regard?

Jack L's picture

... the EVO 400 ($4499), which includes all the aforementioned luxuries plus input and output transformers for balanced XLR connectivity." quoted Herb Reichert.


As a classical music addict & a DIY audio designer/builder, my last thing to do is to use transformers in the music signal paths in my amps.

Why? Transformer is itself a non-linear device due to its non-linear magnetic characteristics. It rolls off both the low & high ends of the audio frequency bandwidth. It slows down the music: phase distortion, & distorts the audio signal with overloaded due to magnetic saturation.

It is noticeable when placed at the signal path front ends, e.g. input & output terminals of the preamp after magnification, e.g. balanced input/output termination with XLR connectors.

We have to know balanced signal transmission is a standard technical practical practice for recording studios, & outdoor/indoor sound performances where miles & miles long audio interconnects are used to connect the equipment. Balanced transmission using XLR connection is to eliminate noises induced into the signal transmission paths - noise-free
signal throughout.

But, but for home audios where long long run of audio signal cables does not exist, any balanced signal transmission using XLR connection therin is therefore redundant & overKILLING !!!!

As explained above, using transformers to provide balanced input/output termination only compromises the music quality due to the downside of any transformers besides spending extra big bucks to install such redundant devices.

So why BALANCED input/output termination using transformers for home audios ???????? Just to make it commercially more appealing & sellable ?

Listening is believing

Jack L