Ayre Acoustics KX-5 Twenty line preamplifier Measurements

Sidebar 3: Measurements

I measured the Ayre Acoustics KX-5 Twenty's electrical performance with my Audio Precision SYS2722 system (see the January 2008 "As We See It"). The volume control operated in 1.5dB steps, the maximum gain obtained when it was set to "46." The maximum gain for both balanced and unbalanced inputs to balanced outputs was just 0.2dB. This is lower than the norm, but it's something I welcome because preamplifiers in general are used to attenuate, not amplify, the levels of source signals. The maximum gain from unbalanced input to unbalanced input was –5.8dB; ie, an input of 1V results in an output of 511mV.

The Ayre's input impedance was very high, at >400k ohms unbalanced and 1 megohm balanced. I can't give more specific values, as the voltage-drop method I use to measure input impedance becomes increasingly inaccurate with the increase in impedance of the device under test. The KX-5 Twenty preserved absolute polarity (ie, was non-inverting) with both balanced and unbalanced inputs. Its XLR jacks are wired with pin 2 hot, the AES convention. The preamplifier's unbalanced output impedance was a low 78 ohms from 20Hz to 20kHz; the balanced impedance was twice that figure, as expected.

The KX-5 Twenty's frequency range is specified as "DC–250kHz," and the preamplifier's response was indeed flat from 10Hz to 200kHz for balanced and unbalanced inputs and outputs at all settings of the volume control (fig.1). Channel separation was superb, at >120dB in both directions below 2kHz, and still 100dB at 20kHz. The KX-5 Twenty is also a very quiet preamplifier: the audioband signal/noise ratio ref. 1V output, measured with the input shorted to ground but the volume control set to its maximum, was 92.5dB (average of both channels). Switching an A-weighting filter into circuit slightly improved this ratio, to 94.6dB, and the wideband, unweighted ratio was a still-good 77.5dB. Spectral analysis of the KX-5's low-frequency noise floor revealed a complete absence of power-supply–related spuriae, other than a minuscule blip at 60Hz (fig.2).


Fig.1 Ayre KX-5 Twenty, balanced frequency response with volume control set to "46" at 1V, into: 100k ohms (left channel blue, right red), 600 ohms (left cyan, right magenta) (1dB/vertical div.).


Fig.2 Ayre KX-5 Twenty, balanced spectrum of 1kHz sinewave, DC–1kHz, at: 2V (left channel blue, right red), 0V into 100k ohms (left cyan, right magenta) (linear frequency scale).

Fig.3 plots the percentage of THD+noise in the KX-5 Twenty's balanced output into 100k ohms. The THD+N rises below 2V output due to the fixed level of noise becoming an increasing percentage of the signal level. The actual distortion rises above the noise floor above 2V, when it is just 0.0018%, and is still only <0.03% at 15.45V output, the maximum I could obtain with the Audio Precision's signal generator. The picture hardly changed when I reduced the load impedance to the current-hungry 600 ohms. For unbalanced input to unbalanced output (fig.4), which is how Art Dudley auditioned the preamp (in my own auditioning I used balanced operation), the distortion was a little higher, and the preamplifier clipped (defined as when the THD+N reaches 1%) at just under 5V, confirming the specified maximum unbalanced input of 8V.


Fig.3 Ayre KX-5 Twenty, balanced distortion (%) vs 1kHz output voltage into 100k ohms.


Fig.4 Ayre KX-5 Twenty, unbalanced distortion (%) vs 1kHz output voltage into 100k ohms.

Distortion remained constant throughout the audioband (fig.5), which is remarkable considering the circuit's lack of loop negative feedback. The spectrum of the distortion comprised only the second and third harmonics (fig.6), which would still be subjectively innocuous at 100 times the level they are actually present in the Ayre's output. As expected, these harmonics were around 20dB higher in the unbalanced output (fig.7), but this is still nothing to be concerned about. Intermodulation distortion was very low, even into 600 ohms (fig.8).


Fig.5 Ayre KX-5 Twenty, balanced THD+N (%) vs frequency at 2V into 100k ohms (left channel blue, right red) and 600 ohms (left cyan, right magenta).


Fig.6 Ayre KX-5 Twenty, balanced spectrum of 50Hz sinewave, DC–1kHz, at 2V into 100k ohms (left channel blue, right red; linear frequency scale).


Fig.7 Ayre KX-5 Twenty, unbalanced spectrum of 50Hz sinewave, DC–1kHz, at 2V into 100k ohms (left channel blue, right red; linear frequency scale).


Fig.8 Ayre KX-5 Twenty, balanced HF intermodulation spectrum, DC–30kHz, 19+20kHz at 2V into 600 ohms (left channel blue, right red; linear frequency scale).

Ayre's KX-5 Twenty is a superbly well-engineered line preamplifier. I just wish I could have shared these results with its designer, the late Charley Hansen.—John Atkinson

Ayre Acoustics, Inc.
2300-B Central Avenue
Boulder, CO 80301
(303) 442-7300

mrkaic's picture

...what this means: "Instead of adjusting volume in the traditional manner—with a potentiometer that discards varying amounts of the product's signal gain—each phase of each stereo channel of the VGT contains 46 discrete resistors that, selected with a rotary switch, create 46 distinct variations on the preamp's voltage-gain stage, for a range of 46 volume levels. Thus there are fewer parts than usual in the signal path, and the preamplifier as a whole, from signal inputs through output buffer, is always maintaining an optimal S/N ratio."

I am especially intrigued by the optimal(sic) S/N ratio. How does one determine this optimality?

Axiom05's picture

The volume is controlled by changing the gain of the preamp. Changing the volume changes the resistor value in the signal path. This is not a passive preamp, every volume step has a specific amount of gain. Despite being conditioned to think that a passive preamp should sound best, this preamp sounds absolutely amazing: extremely quiet and smooth. The drawback to this approach is that you need a lot of resistors to effect the volume changes, the smaller the steps, the more resistors, the higher the parts cost. This is why the KX-5 Twenty has about 1.5 dB volume change for each "click" of the control vs 1 dB for the KX-R Twenty. The optimal S/N ratio can be maintained by choosing parts with close tolerances. You have more control over part values. Typical volume potentiometers are not accurate over their entire range often resulting in an increase in noise and/or channel imbalance.

mrkaic's picture

Thank you for an exhaustive and informative answer. So, if I understand you correctly, the bank of 46 resistors essentially selects the closed loop gain -- like the resistors in an op amp circuit?

Could this work with a digital potentiometer? Those can give very close tolerances and even more choices that a bank of resistors.

Indydan's picture

Would it not have been better to test the preamp with an Ayre power amplifier? Of course the Shindo preamp will sound better with a Shindo power amp, DOH!

Can generalizations be made when the Ayre was tested with 50 year old speakers?

I wish Mr. Hansen was still around. He would have surely commented on the review.

johnnythunder's picture

As much as I love Art's writing and reviews, I agree with Indydan about this somewhat flawed approach to reviewing these state of the art components in a "vintage" reference system. I think it's an interesting angle to hear how these components blend together but it's sort of a tangental idea not one for a comprehensive review. It should be classified as a separate type of review - this mix of old and new. How do AR or KLH bookshelf speakers stack up against today's ELACs or KEFs or NOLAs or JOSEPHS? that would be interesting to read...but speakers designed to be used with low powered tube amps powered by a new amp is not the same thing...

mrkaic's picture

Johnny, what is your problem?

Mr. Dudley writes SUBJECTIVE reviews, so he can pair any components he wants and evaluate them by LISTENING. [Gee, I wonder where the master title of his editorials comes from? :)] His approach is not scientific.

If you want rigor, read JS'a measurements.

johnnythunder's picture

and especially not attacking Art D. who is a national treasure and who's ears and musical taste are impeccable. He's one of my favorite magazine writers. My issue was the odd mismatch of amp and speakers as the only point of comparison in the review of the amp. I am not in the double blind comparison camp or a skeptic about tweaks, cables or AC power enhancements. I simply felt that the Ayre should ALSO best be paired with a modern state of the art speaker that it was obviously designed for. I wouldn't be stating any of this if the review was of a modern state of the art push pull or single ended tube amp used w his Altecs. So this has nothing to do with science or rigor but of what is in my opinion a slightly incomplete way of assessing this particular piece of equipment. It's just my opinion in this instance.

mrkaic's picture

You see, I'm a passionate objectivist and rely on measurements first and foremost. However, if the genre of a piece is subjectivist, then anything goes. Including pairing amps and speakers that are not from same period of audio development.

So, I guess we see this issue somewhat differently.

CG's picture

Kind of late to reply...

I was motivated to reread the review based on Jim Austin's recent comparison of this preamp to a PS Audio unit.

Personally, I straddle the objectivity camp and the subjectivist camp. Here's why...

In my day job (nothing to do with audio), I am exposed to more electrical engineering measurements than any reasonable person should ever be subject to. What I have found is that it's very, very easy to oversimplify any form of analysis by relying on a rote set of measurements, industry standards or not. Very often, they don't really tell the whole story. That isn't to say that you can't invent tests that tell you more, but that's not the usual modus operandi, at least until something breaks.

So, to me, measurements are often - not always - what my geometry teacher used to call necessary but possibly insufficient conditions." In other words, they may not tell the whole story.

But, if that's not your belief set, I'm not here to argue with you or dismiss your philosophy. Just explaining where I am coming from.

That said, you do raise a really good question. (Whether you knew it or not!)

Why are all the pieces of audio equipment tested in isolation, by themselves, under the most pristine test conditions possible? Why are they not tested when used in conjunction with other gear or an entire system? Doing this all in isolation is hardly a "sufficient condition."

The list of possible system interactions between components is far, far too long to even compile here. That's even when excluding the acoustic performance of the loudspeakers in concert with the room. Just the numbers of possible electronic interactions is staggering.

Yeah, I know it'd be very labor and time intensive. But that might be preferable to a set of tests that tell you just enough to lull you into the idea that you know what an audio component will do. That's why trying a new piece in your own system and deciding whether you like the sound might be a way better approach.

Any thoughts?

villager56's picture

Hi John,
The measurements reported for this unit appear very similar to the measurements that you reported for the K-5xeMP in 2011. Would you kindly add a few words comparing the measurements of these two units and how the new technologies are reflected in the measurements of the latest model? Thanks!