Pass Laboratories XP-32 line preamplifier

When I studied physics at university too many years ago to admit, I learned about Occam's razor. Many, many more years ago than that, Franciscan friar William of Occam stated that a hypothesis should provide the simplest possible explanation for a phenomenon.

I apply Occam's criterion to my audio system: It should be no more complicated than it need be. The D/A processors I have written about over the past few years—dCS Rossini and Vivaldi, MBL Noble Line N31, PS Audio DirectStream, Weiss DAC502—all have high-precision digital volume controls. I therefore haven't had a conventional preamplifier in my system for a long time, instead connecting the processors' balanced outputs directly to the power amplifiers. Even for LP playback—don't tell Michael Fremer!—I connect the balanced, RIAA-equalized output of my Channel D Seta L phono preamplifier to an Ayre QA-9 A/D converter and send the 24/192 PCM data to the digital processor!

However, when I reauditioned the MBL N31 for my follow-up review in the December 2020 issue, I had a disturbing experience when I set the N31's volume control to its maximum and inserted MBL's N11 line preamplifier between the DAC and the power amplifiers to control volume. As I wrote in my follow-up, "I was not expecting an increased sense of palpability to the acoustic objects within that soundstage, particularly vocals, but that is what I heard. ... [B]ut how can inserting an active device with vanishingly low levels of noise and distortion result in improved sound quality?"

I must have been having a senior moment when I penned those words, because I had forgotten that in my December 2014 review of Ayre Acoustics' KX-R Twenty line preamplifier, I had described the same experience with the original KX-R: "To my astonishment, the sound of my system with a [Logitech] Transporter D/A processor feeding the preamplifier was better than when the DAC fed the power amplifier directly."

As I had previously found that the KX-R's sound quality was equaled by Pass Laboratories' XP-30 three-chassis preamplifier, when I learned that the XP-30 had been replaced by the XP-32 ($17,500), I asked for a review sample.

221pass.bac

The '32
The new preamplifier, like the XP-30, was designed by Wayne Colburn, who has been responsible for Pass Labs preamplifiers since the mid-1990s. It looks identical to the XP-30, with each channel's audio circuitry in a separate chassis and the power supply and control circuitry in a third chassis. This last chassis has a rectangular, alphanumeric, blue-fluorescent display in its center, a large volume-control knob to its right, and five control buttons to its left. These controls are duplicated on the aluminum remote control, which also has buttons for channel balance and "Pass Thru." The latter sets the gain to 0dB ("179" on the display) and locks the other controls so that the XP-32 can be used in a home-theater system, where volume is adjusted elsewhere.

Keeping the "dirty" control circuits and power supply separated from the actual preamplifier channels minimizes the possibility of RF and other spuriae from contaminating the analog signals. (Pass Labs' XP-22 preamplifier, which Jim Austin favorably reviewed in June 2019, is similar other than having both channels' audio circuitry in one chassis.) Umbilical cables connect each audio chassis to the control chassis, terminated in locking, military-spec, multipin circular connectors rather than the XP-30's rectangular DIN-25 connectors. These must be connected prior to powering up the XP-32. (The power supply has to "see" the audio channel before it will send power to the audio chassis.)

221pass.bac2

Each audio chassis has five numbered inputs, a dedicated home-theater input, and a tape-loop input, all on both XLR and RCA jacks. (When the XLR is used, a jumper connecting the jack's pin 3 to pin 1 must be removed.) The ground of each RCA input is in parallel with pin 1 of the XLR, and the RCA input feeds a summing junction that, according to the manual, "[maximizes] the patented supersymmetry (X circuit) and preserves the balanced character of the XP-32 from input to output when fed from a single-ended source." There are three outputs, again on both XLRs and RCAs: one main output that follows the volume control setting; a second output whose level can also be adjusted with the channel's front-panel gain control to allow use with power amplifiers of differing sensitivities in multichannel applications; and a unity-gain tape-loop output.

XP-32 vs XP-30
Given that the XP-32 appears identical to its predecessor, other than having a volume control that operates in 0.5dB steps rather the earlier one's 1dB steps, I asked Wayne Colburn what the internal changes were.

221pass.life

"The volume control is single-stage as used in the Xs preamplifier and has more range, with 0.5dB steps; it is quieter and more accurate. While the XP-30 had a 1µF polypropylene input cap, the [XP-32's] inputs are DC-coupled and a servo is used after trimming. The XP-32 has a new input circuit that is based on the Xs: it is lower in noise and distortion with greater drive capability. Noise is the most prominent part of THD+N at low levels. By lowering noise we get better resolution and dynamics....The basic circuit topology is the same, but the plug-in gain modules are unique to the '32 and use some new cascode transistors and higher bias. The transistors used are [matched pairs of] Toshiba FETs on the input and MOSFETs on the output. We have a nice stash of these purchased before they were discontinued. For the basic signal path, these are tough to beat and I consider them the standard, as do other audio engineers.

"The output stage runs a higher class-A bias with auto bias and lower output impedance. After all, why would anybody make a pre that wasn't class-A? The XP-32 can drive long cable runs and low input impedances easily. A servo controls DC offset after it is set by hand and balance-adjusted. The output is still capacitor-coupled even with the servo but can be bypassed with internal jumpers. It is really an interface to other products since I have no guarantee of what it will connect to. While caps still have a sonic signature, they have gotten really good."

COMPANY INFO
Pass Laboratories Inc.
13395 New Airport Rd., Suite G
Auburn, CA 95602
(530) 878-5350
ARTICLE CONTENTS

COMMENTS
Kal Rubinson's picture

A frightening thought:

Can Pass whip up a 6 or 8 channel version of this?

John Atkinson's picture
Kal Rubinson wrote:
Can Pass whip up a 6 or 8 channel version of this?

I believe you can add additional audio-channel chassis, each controlled by the same control unit.

John Atkinson
Technical Editor, Stereophile

Kal Rubinson's picture

John Atkinson wrote: I believe you can add additional audio-channel chassis, each controlled by the same control unit.

I was really thinking of the price tag and the rack space.

Anton's picture

Didn't a prominent Stereophile editor once say that it was the preamp that was the most important part of a system?

;-D

It's out of my budget, but I love Nelson Pass.

liquidsun's picture

Hm, I see Sonus Faber Lumina III used as speakers during the review. Does that mean we can expect the Lumina's to be reviewed sometime in near future? :)

John Atkinson's picture
liquidsun wrote:
Hm, I see Sonus Faber Lumina III used as speakers during the review. Does that mean we can expect the Lumina's to be reviewed sometime in near future? :)

My review of the Lumina III will be published in the April issue of Stereophile.

John Atkinson
Technical Editor, Stereophile

liquidsun's picture

It's a pity that stereophile didn't review the SF sonetto line so we could make a direct comparison between both lines. However it's still good news to hear :)

tonykaz's picture

Didn't you say that you returned it ?

Did you have to purchase it or is it on long term loan?

Tony in Venice Florida

John Atkinson's picture
tonykaz wrote:
Didn't you say that you returned it? Did you have to purchase it or is it on long term loan?

After I measured the MBL N11 for Jason's review, I arranged to hold on to the preamplifier first to use in my followup review of the MBL N31 CD player/DAC in the December 2020 issue, then to use as a reference for this Pass Labs review. It has since been shipped to another Stereophile reviewer to use as a reference for a review of another product.

John Atkinson
Technical Editor, Stereophile

MarkusL's picture

I once connected my Wadia 781 directly to my power amplifiers (Pass X 600). The sound is much better with my preamplifier (Audio Research Ref 5) in between!

Kal Rubinson's picture

I know what you mean:

Quote:

I once connected my Wadia 781 directly to my power amplifiers (Pass X 600). The sound is much better with my preamplifier (Audio Research Ref 5) in between!

I gave up on finding a suitable preamp since I need 8 channels. I run my DACs directly to my power amps and it is, indeed, much better.

LTig's picture

I have some trouble looking at the signal to noise specs. The claim is 150 dB at max output voltage, and 500 nV residual noise, which is more or less what you expect with an output voltage of 23V and 150 dB S/N.

However JA measured 85 dB at 1 V. At 23 V we should then expect 27 dB more which is 112 dB, a whopping 38 dB less (a factor of 80). This would result in a residual noise of 56 mV, more than a factor of 100 higher, Even with bandwidth restriction and A-weighting the measured S/N reaches "only" 124 dB, 26 dB less than claimed.

WayneC's picture

The SN ratio was calculated from max output of 32 Volts with 1uV as the noise. The noise floor is actually lower than 1uV. It all depends how you measure it and the reference. Of course we spec it to look good, the suns output would make a good max reference. Measured on the AP with a 30 KhZ filter it is about 14 uV broadband. John's AP FFT show -120 to -140 dB range. 56 mV would be rather crappy, its a very quiet preamp.

LTig's picture

Aah, there is a typo on the specification page: 23 V max (correct is 32 V). So my calculations are 3 dB off.

14 uV noise broadband (30 kHz) is almost 30 times more than the specified 0.5 uV. May I ask under which conditions you did measure those 0.5 uV?

Regarding JA's FFT plot: one have to keep in mind that the FFT noise floor does not show the S/N value. You have to subtract the FFT gain (3dB per doubling of bin size) to get the real value. I don't know the FFT size of JA's FFT plot. With 16k FFT size the FFT gain is 39 dB. Subtracting 39 dB from 120 - 140 dB we get 81 - 101 dB which correlates quite well with JA's measured S/N.

Michael Fremer's picture

No one else has to tell me that you digitize your analog. You just did! One thing's for certain: we ALL have to analog our digital in order to hear it!

John Atkinson's picture
Michael Fremer wrote:
No one else has to tell me that you digitize your analog. You just did!

Me and my big mouth :-)

John Atkinson
Technical Editor, Stereophile

krahbeknudsen's picture

I hope you don't mind if I point out that your turntable setup is perhaps getting a bit long in the tooth? I love the Linn and all that but at lot has happened since 1991 even in Linn world. That old MDF tone arm board is not exactly known for it's high resolving power. Could it be that your digital setup is outresolving your turntable in the time domain and perhaps that would not be the case with a modern turntable or even an up to date Linn?

John Atkinson's picture
krahbeknudsen wrote:
I hope you don't mind if I point out that your turntable setup is perhaps getting a bit long in the tooth? I love the Linn and all that but at lot has happened since 1991 even in Linn world...

Indeed it has, but I am so familiar, so comfortable, with the sound of my LP12-based LP system that I fear change.

John Atkinson
Technical Editor, Stereophile

a.wayne's picture

Why on earth would you not want to run the pre Amp Direct , favoring the sound of caps maybe ..??
Btw JA ,

Are those stereophile CD’s mentioned still available ..??

Regards

John Atkinson's picture
a.wayne wrote:
Are those stereophile CD’s mentioned still available ..??

Yes, see http://shop.stereophile.com/music-cds/

John Atkinson
Technical Editor, Stereophile

Charles E Flynn's picture

Looking at my test CDs, I found, with no jewel box, cover art, or record of purchase, Stereophile CD STPH-002-2. It was apparently the first test CD Stereophile made, released in 1990.

jahnghalt's picture

I suggest a follow up to address those coupling caps.

"The output is still capacitor-coupled even with the servo but can be bypassed with internal jumpers."

On the back panel - "no user serviceable parts"

Question for Wayne - why not have jumpers accessible to the exterior? Note all those XLR jack jumpers/

(good photos BTW)

X