You are on page 1of 8


Parasound P5 and A23


lever! That’s what I cried out once distortion in that amplifier, plus effectively higher frequencies to your power amplifier
I’d finally got my head around the give it a far higher dynamic power rating. and (therefore) main speakers. And yes, I
Swiss Army knife of a compo- You gain all these advantages by using the know there are some subwoofers on sale that
nent that Parasound has created electronic crossover to send all the low allow you to do this, but most don’t, and of
with its P5 pre-amplifier. Then I thought to frequencies to a subwoofer (or in the case of the few subwoofers that do, not many do
myself: ‘Really clever!’ Because although the the Parasound P5, up to two subwoofers) for it well. Far better to perform the crossover
Parasound P5 is a fully-featured preamplifier, amplification and reproduction, and only the function in the preamplifier section, as is
and a digital-to-analogue converter, and a done here.
headphone amplifier… and a lot more be- So on the back of the P5 you’ll find two
sides… it’s also an electronic crossover, and so crossover outputs: a low-pass output with a
far as I know, no-one has previously thought rotary control and settings at 20Hz, 40Hz,
to integrate an electronic crossover into a pre- 60Hz, 80Hz, 100Hz, 120Hz and 140Hz, and
amplifier. And I, personally, think that doing a high-pass output also with settings from
so is an absolutely wonderful idea. 20Hz to 140Hz and calibrations at 20Hz
Newport Test Labs

intervals. Associated with both crossovers are

The Equipment slider switches that allow the crossovers to be
Why do I think incorporating an electronic bypassed, should you so desire.
crossover into a pre-amp is so wonderful? Despite my enthusiasm for the idea of in-
Because it means you can use it to prevent corporating an electronic crossover in the P5,
Doppler distortion in your main loudspeakers it appears this wasn’t really Parasound’s pur-
and effectively increase the power-handling Power Output: Single channel driven into pose in so doing. According to Parasound’s
8-ohm, 4-ohm and 2-ohm non-inductive
ability of those speakers. And insofar as your long-serving CEO, Richard Schram, the actual
loads at 20Hz, 1kHz and 20kHz. [Parasound
power amplifier is concerned, it can reduce A23 Power Amplifier] reason was: ‘to offer a true audiophile two-chan-

24 Australian
Parasound P5 Preamplifier & A23 Power Amplifier ON TEST

headphone output (also via 3.5mm phone rear illumination when the muting is not ac-
socket…not my favourite physical connector tive, and red-coloured rear illumination when
for a headphone socket on a full-sized hi-fi it is), the tone on/off button isn’t nearly
component). so intuitive, because it’s illuminated from
So far as the DAC inside the P5 is con- behind with blue light when the circuit is ac-
Newport Test Labs cerned, Parasound is using a Burr-Brown tive, but there’s no light at all when it’s off.
24-bit, 192kHz-capable PCM1798 DAC that’s The shape of the power on/off and mute
fed via optical (Toslink), coaxial (RCA), or buttons is strange, presumably so they
USB (use the USB option though, and you’re can be rear-illuminated more effectively.
limited to 24-bit/96kHz performance). This They’re lozenge-shaped and protrude from
is the same DAC PS Audio uses in its stand- the panel on a stalk that makes them look
alone NuWave DAC, which costs more than almost mushroom-like… if you can imagine
Power Output: Both channels driven into Parasound is asking for its P5. a lozenge-shaped mushroom. They do work
8-ohm, 4-ohm and 2-ohm non-inductive As you can see from the front panel, the beautifully though… a very ‘positive’ feel
loads at 20Hz, 1kHz and 20kHz. [Parasound
P5 has good old ‘old-fashioned’ (but very use- with great tactile feedback.
A23 Power Amplifier.]
ful!) bass and treble controls, but if you’re not The remote control isn’t particularly flash,
nel pre-amp that could be seamlessly integrated a fan of tone controls, you can press a button and you don’t get any additional features
into a multi-channel system’ because ‘few people and take them out of circuit. A balance con- when using the remote, so it’s really only for
have the space for completely separate two- trol (also useful, and inexplicably AWOL from the convenience of operating the unit from
channel and surround sound systems.’ There’s a great many pre-amps and integrated ampli-
no doubt that Schram is right and that, as fiers) is also a fixture on the P5, so it also gets
he says, ‘the P5’s Bypass and 2.1 modes allow a big tick from me. There’s a separate volume
Parasound P5 & A23
the user to preserve the true audiophile integrity control for the subwoofer outputs, plus a Pre & Power Amplifiers
of their stereo set-up yet also offers the ability global muting button. This muting button
to engage their surround sound system at the is not overly ‘smart’ in that it will turn off Brand: Parasound
touch of a button’ but I would prefer to think automatically (as it should) when the up or Model: P5/A23
of the P5 as an exclusively stereo high-end down buttons on the remote are pressed, Category: Pre/Power Amplifiers
pre-amplifier. but not if the front panel’s volume control RRP: $2,195 (P5)/$1,995 (A23)
As a dedicated pre-amplifier, the P5 Warranty: Three Years
certainly offers tons of features and opera- Distributor: Network Audio Visual Pty Ltd
tional flexibility. In addition to the digital Address: Unit 6B, 3–9 Kenneth Road
inputs (about which more later), the P5 has Manly NSW 2093
five line-level analogue inputs, four of which (02) 9949 9349
are unbalanced (RCA terminals) and one of info@
which can either be balanced (XLR terminals)
or unbalanced (RCA)—but not both at the
same time… you can only connect one or the • Built-in electronic
other! It also has a phono input with switch- crossover
ing for both moving-coil (MC) and moving- p ort T • Hugely powerful
est L
magnet (MM) cartridges, with selectable load abs • Incredibly low price
(100Ω or 47kΩ) for the moving-coil option.
There’s also a home theatre bypass option, as • 3.5mm headphone
I mentioned earlier, which outputs the left-
Power Output: Single and both socket
channel, right-channel and subwoofer signals channels driven into 8-ohm, 4-ohm • Utilitarian build
without any processing or volume adjust- and 2-ohm non-inductive loads at
ment. There’s also a pair of fixed-level Record 20Hz, 1kHz and 20kHz. [P5/A23]
Outputs, so you can send analogue audio to a
recorder of some kind if you wish.
On the ‘outputs’ front, you have the
is turned, which means it could be possible
for your speakers to be blasted with sound at
already-mentioned outputs from the two high volume levels if the front panel volume Readers interested in a full technical
crossovers (which if bypassed operate as control is turned up while the amplifier is appraisal of the performance of the
conventional full-range unbalanced analogue muted, and the muting then removed. Also, Parasound P5 Pre-amplifier and A23
outputs, via RCA terminals) but you also have the P5 allows you to switch inputs while its Power Amplifier should continue on and
a pair of full-range balanced outputs (XLR) output is muted, so if you switch to a source read the LABORATORY REPORT published
and a balanced subwoofer output (XLR). The that has a substantially higher output than on page 30. Readers should note that
output from these can also be optionally the source you switched from, you could the results mentioned in the report,
routed through the crossover or switched again be ‘blasted’ by sound at high volume tabulated in performance charts and/
to operate as full-range outputs. And if all when you un-mute. or displayed using graphs and/
this wasn’t enough, there’s also an analogue Interesting, although the muting or photographs should be con-
input on the front panel (a 3.5mm stereo button has associated colours that strued as applying only to the
phone socket) for a portable device, plus a make its status clear (blue-coloured specific sample tested.

Lab Report on page 30 25
ON TEST Parasound P5 Preamplifier & A23 Power Amplifier

the comfort of your listening chair. It’s sup- you look at the rear of the A23, you can see it terminals, but if there’s a MkII version of the
plied with a set of ‘Titen’ (sic) AA batteries. has far more controls and features than you’d A23, I would really like it if Parasound could
Don’t even think of installing these in the expect. It has both balanced (via XLR) and allow Curl a bit of extra budget to improve
remote… chuck ‘em and install a decent set unbalanced (via RCA) inputs, plus a loop- the terminal quality. However, since the A23
of Eveready alkalines that are guaranteed not through so you can route an input signal has been in continuous production since the
to leak! to an additional amplifier if you like. There turn of the century, I can’t see there being a
The exterior of the P5 is thick, solid are also two volume controls—one for each MkII anytime soon!
metal, and the assembly and build quality is channel—so not only can you adjust volume, Internally, the most visually impressive
excellent, but it’s not exactly pretty. It looks you can also adjust channel balance. These component is the massive custom-made 1kVA
like it’s been built to slide into a profes- controls have a smooth action, rather than a encapsulated toroidal power transformer
sional equipment rack, or be tucked away ‘click-stop’ action, so if you need precise level with its independent secondary windings
in a cupboard rather than as a display piece. adjustments, you’ll likely have to use test for each channel. Storage and smoothing
(And, since you can indeed order the P5 with equipment to get the levels perfect, unless is by way of four 10,000µF/80V electrolytic
rack-mount handles, I’m probably not too far you set both controls to maximum (labelled capacitors, for a total of 40,000µF. (Parasound
wide of the mark with this description.) as ‘Reference/THX’). specifies a total of 48,000µF, so presumably
Three toggle switches are provided for there’s been some slight modifications to
Parasound Ground Lift (which can be used to prevent the circuit over the years the A23 has been
A23 Power Amplifier ground loops that might otherwise result in Parasound’s line-up… the PCB on our
The circuitry for the A25 was designed by in audible mains hum), balanced/unbal- review sample, for example was last revised
John Curl, whose name is legend in au- anced input selection and selecting between in 2002.) The amplifier topology is that of
diophile circles. Although Curl was already ordinary stereo operation and bridged mono a complementary MOSFET driver stage and
well-known in professional audio circles for operation. As the name suggests, this means JFET input stage, with the final output being
his work with Ampex Corporation, and with you can use the A23 as a standard two-chan- via twelve beta-matched 15-amp, 60MHz bi-
musicians for his work with The Greatful nel stereo amplifier, in which mode it’s rated polar output transistor pairs—Sanken C3519/
Dead, he didn’t really come to the attention to deliver 125-watts into 8Ω loads, or you can A1386 types, to be precise. There are two big
of the audiophile fraternity until he worked switch it for single-channel monobloc opera- internal aluminium heatsinks running down
on the JC1 and JC2 amplifiers for Mark tion, in which case the A23 is rated to deliver either side of the PCB, so there’s no need
Levinson (and if you didn’t know that, you 400-watts into the single 8Ω load. for any noisy fan-cooling. The amplifier is
now know what the initials of those Mark At the left of the rear panel is a suite of well-protected from misadventure thanks to
Levinson products stood for). Curl eventu- options for turning the A23 on (and off). the use of d.c. servos, relays and fuses, along
ally left Levinson to found his own high-end You can elect to do this manually, or you can with thermal protection. Rather unusually,
amplifier company (Vendetta Research) but have the Parasound auto-sense the presence all the individual components on the PCB are
has since 1989 been consulting for Parasound of an audio signal and switch on automatical- ‘old-school’ through-hole types… there’s not
in San Francisco, primarily developing the ly (in which case it will switch off automati- a surface-mount component in sight!
Halo series amplifiers for the company, but cally once it no longer detects an audio sig-
also working on other Parasound projects. At nal). You can also have it triggered remotely In Use and
Parasound, Curl is insistent that his designs via the usual 12V d.c. system used by home Listening Sessions
use the finest parts possible, avoids the use theatre (and home install) components. To It wasn’t until I had mounted the Parasound
of capacitors and inductors in the circuit ensure there isn’t any ‘nuisance’ switching, P5/A23 combo on my equipment shelving
path, and continues to use balanced circuits, the A23 waits for around five minutes after that I really took any notice of the red power
despite the additional costs. Schram once said it no longer detects a signal before switching lights that indicate each component’s power
of Curl: ‘He wouldn’t last a week in a mass-mar- itself off. Switch-on is instantaneous. status. Each centrally-located light is a rectan-
ket factory … the accountants would probably re- The speaker terminals are certainly gular section of black plastic that’s embossed
ject every part he picked because it cost too much. serviceable, and they’re gold-plated multi- with a metallic ‘P’. These glow a fairly bright
At the same time, he knows how to make a very, way types, so they’re not exactly ‘budget’ red when the Parasounds are operational,
very good product at what
we consider to be a reason-
able price.’ Despite being
lauded for his amplifier
designs Curl, at age 72, is
still chasing perfection.
‘We have never yet been
able to make a completely
perfect amplifier at any
price… the challenge is to
make very good amplifiers
in a way that is cost-effec-
tive, he says.
As you can see for
yourself, the front panel
of the Parasound A23
is almost devoid of
controls, pretty much
as you’d expect for a
power amplifier, but if

26 Australian
ON TEST Parasound P5 Preamplifier & A23 Power Amplifier

and a soft muted red when tor… I lose a lot of CDs that
they’re in Standby. I found way!) From the very first
that on close inspection notes I remembered why
the design wasn’t really to this CD became ubiquitous
my taste, being a little too as a demo disc at hi-fi shows
‘American’ in style… but right around the world.
that’s very much a personal It’s an amazingly-recorded
call. CD, with David Piltch’s
Operationally, every- double-bass ready to reveal
thing worked perfectly, the low-end limitations of
though the rotary controls any loudspeaker, and Aaron
used for bass, treble and Davis’s piano set to high-
balance felt a little stiff light any tonal irregularities,
under my fingertips. The while Holly Cole’s stunning
input source encoder, on voice either soars over the
the other hand, is a delight instruments or whispers
under the fingertips, rotat- and croons in the cracks
ing smoothly and with the between the notes. And I
sound automatically and was reminded how nice it
momentarily muted during is to hear real strings rather
switching. The relays that than synthesised ones, and
do this are a tad noisy, but even a real lap steel guitar
this is of no import. Blue on Don’t Let the Teardrops
LEDs light to show the se- Rust your Shining Heart…
lected input, unless you’ve and what a great version
selected ‘Bypass’ which is instead indicated found that I’d not previously heard, but also of Ben Watt’s (from Everything but the Girl)
by an orange LED. because her interpretations made me go back it is. Actually, every single track on this CD
The motorised volume control works and re-evaluate other, different versions… is a great version, whether it’s Cole Porter’s
reliably and well no matter whether you use also a good thing. When listening, I had am- Get out of Town, Willard Robinson’s Don’t
the remote or turn it manually. If you turn ple opportunity to appreciate what the Para- Smoke in Bed or Casey Scott’s Cry (If You Want
it manually there’s no backlash, so you can sound P5/A23 duo was bringing to the party. To). Through it all, the Parasound duo was
set volume levels easily and repeatedly, and The first and most obvious contribution was audibly invisible… all I was hearing was the
if you use the remote, you can see where the the dynamism that’s on tap, with the A23 music, being delivered better-sounding than I
volume knob is set to thanks to a white line able to effortlessly deliver Dinnerstein’s most ever remember it. For me, there is no greater
scribed on it. thunderous crescendos… and they are thun- proof of a superb amplification chain.
One of the first tasks I set myself with derous, because she has powerful fingers and During my listening sessions I used three
the Parasounds was to use them to seriously an even-more powerful technique, allowing different speaker set-ups: a floorstanding pair
audition a young(ish) pianist of whom I her to extract major SPLs from any piano she of speakers, the same pair but with a sub-
previously was not aware, and had been plays—in this case a restored 1903 Steinway woofer, and a smaller pair of bookshelf speak-
recommended to me by a friend who said Model D. Her ‘not quite staccato’ attack on ers also both with and without a subwoofer.
that I would ‘enjoy her unique Bach inter- the keys in Variation 1 is miraculously good, The Parasound combo worked brilliantly
pretations’. Well I don’t know about ‘enjoy’ as is her tonal shading. The ritardando at the across all four speaker combinations, but
bit but I was certainly in no doubt about conclusion is perfect. In fact, Dinnerstein’s simply excelled when the subwoofer was in
the ‘unique’ because I don’t think I’ve ever Goldberg (Telarc CD80692) is now my third the mix, particularly with the floorstanding
heard any pianist use the same phrasing or favourite version of this work, my favour- speakers, because I could set the crossover at
dynamics. But that was what most people ites being Gould’s 1955 and 1981 versions. such a low frequency that the subwoofer inte-
thought when they first heard a young Glenn What Gould absolutely nails when playing grated perfectly, so I was getting full range
Gould’s take on the Goldberg Variations. And Bach was to my mind identified perfectly by from the floorstanders and only the sub-bass
it was Gould’s performance that triggered a Norwegian pianist Leif Ove Andsnes, who in from the subwoofer. So if you haven’t previ-
worldwide re-evaluation of Bach’s work. His an NPR interview said: ‘The miracle of Gould ously considered a 2.1 system for stereo,
recording, in 1955, was amongst the very few is for me his touch, and the floating quality of now is the time to start thinking about it…
recorded versions available at the time, and his polyphonic playing, where the vertical and because it really works a treat!
the first of the few to gain any prominence. horizontal elements in the music are in perfect
But as a direct result of Gould’s performance, balance. And the millions of nuances within each Conclusion
I can tell you that as of 2015 there are now voice in a four-part fugue, for instance—it’s like Parasound is one of the very few companies
213 versions available, across multifarious four brilliant minds working at the same time. in the world that has a reputation for state-
instruments… including—I’m told but don’t That part of Gould was pure genius, and makes of-the-art sound quality and engineering as
quite believe—a version for jaw harp. both his early and his late Goldberg Variations well as a reputation for delivering superb
So think what you will of Simone Dinner- desert island recordings.’ value for money, and in looking at this duo,
stein’s interpretations, if her performances in- After many enjoyable hours with Bach I it’s easy to see why: buy this pair and you’ll
spire others to investigate Bach’s work, that’s resurrected an old warhorse: ‘Don’t Smoke in be getting an awful lot of power and incred-
a great outcome. However, lest you think I’m Bed’ by the Holly Cole Trio… or for me a new ible performance for what, quite frankly, are
being too hard on her, I’ve been leading you warhorse because somehow I lost my original ridiculously low prices. And that’s true no
on for literary effect, because I absolutely love copy a few years ago. (I probably left in a matter whether you buy the P5, the A23, or
her playing, not only for the inner voices she loaner CD player I returned to the distribu- both. greg borrowman


28 Australian
LAB REPORT Parasound P5 Preamplifier & A23 Power Amplifier
0.00 Newport Test Labs 0.00 Newport Test Labs

Laboratory Test Results -20.00 -20.00

First up, I want to make it clear that very,

very few of these results can be compared -40.00 -40.00

against Parasound’s specifications because -60.00 -60.00

whereas Parasound specs each amp separate-

-80.00 -80.00
ly, Newport Test Labs tested the P5 and A23 as
a pair so, for example, the distortion results -100.00 -100.00

are the sum of the P5’s THD and that of the

-120.00 -120.00
A23. So each individual component would
have lower levels of distortion than is shown -140.00 -140.00

0.00 Hz 4000.00 8000.00 12000.00 16000.00 20000.00

here. [Editor’s Note: I requested the testing be Graph 1: Total harmonic distortion (THD) at 1kHz at an output of 1-watt into an 8-ohm
0.00 Hz 4000.00 8000.00 12000.00 16000.00 20000.00
Graph 2: Total harmonic distortion (THD) at 1kHz at an output of 1-watt into a 4-ohm non-
inductive load, referenced to 0dB. [Parasound P5 Pre & A23 Power Amplifiers]
done this way because in practise, this is the way non-inductive load, referenced to 0dB. [Parasound P5 Pre & A23 Power Amplifiers]

most people will use their Parasound combo in dBFS dBFS

0.00 0.00
real life.] Newport Test Labs Newport Test Labs

Looking at the frequency response first, -20.00 -20.00

you can see that the overall bandwidth of

-40.00 -40.00
this pre/power combo is very wide, being just
3dB down at 4Hz and 116kHz. Even if you -60.00 -60.00

constrain the dB limits, the overall response

-80.00 -80.00
extends from 9Hz to 57kHz ±0.5dB. Across
the audio band (20Hz to 20kHz) the response -100.00 -100.00

is even flatter again, as you can see from -120.00 -120.00

Graph 8, where the black trace (the response

into a standard non-inductive 8Ω laboratory- -140.00 -140.00
0.00 Hz 4000.00 8000.00 12000.00 16000.00 20000.00 0.00 Hz 4000.00 8000.00 12000.00 16000.00 20000.00
grade load) is just 0.3dB down at 20Hz and Graph 3: Total harmonic distortion (THD) at 1kHz at rated output (125-watts) into an 8-ohm Graph 4: Total harmonic distortion (THD) at 1kHz at rated output (200-watts) into a 4-ohm
non-inductive load, referenced to 0dB. [Parasound P5 Pre & A23 Power Amplifiers] non-inductive load, referenced to 0dB. [Parasound P5 Pre & A23 Power Amplifiers]
0.1dB down at 20kHz so, normalised, the re-

dBFS dBr

Very few amplifiers

0.00 Newport Test Labs 3.00 Newport Test Labs


are capable of this



high standard of -60.00



performance -80.00



sponse is 20Hz to 20kHz ±0.15dB. The other -120.00 -2.00

(red) trace shows the amplifiers’ frequency -2.50

response when driving a load that simulates -140.00

0.00 Hz 6000.00 12000.00 18000.00 24000.00 30000.00

that of a typical two-way loudspeaker sys- Graph 5: Intermodulation distortion (CCIF-IMD) using test signals at 19kHz and 20kHz, at an
10.00 Hz 100.00 1000.00 10000.00
Graph 6: Frequency resp an output of 1-watt output into an 8-ohm non-inductive load, tone
output of 1-watt into an 8-ohm non-inductive load, ref to 0dB. [Parasound P5 & A23 Amps]
tem… in other words, a ‘real-life’ load. controls off (black trace) vs. tone controls on but detent position (red trace). [P5 & A23]

You can see that the performance of the dBr dBr

Parasound is virtually identical to that into 20.00 Newport Test Labs 1.00 Newport Test Labs

16.67 0.83
an 8Ω load, meaning the A23 (in this case)
13.33 0.67
will sound the same no matter what speakers 10.00 0.50

it’s being used to drive. Very few amplifiers 6.67 0.33

are capable of this high standard of perfor- 3.33 0.17

mance. 0.00 0.00

-3.33 -0.17
However the linearity of the frequency re-
-6.67 -0.33
sponse does come with one proviso, which is -10.00 -0.50

that the tone controls circuit needs to be de- -13.33 -0.67

feated to achieve this response. The reason is -16.67 -0.83

shown in Graph 6, where you can see the top -20.00 -1.00

10.00 Hz 100.00 1000.00 10000.00 10.00 Hz 100.00 1000.00 10000.00

(black) trace, with the tone controls switched Graph 7: Tone control action referenced to 0dB at 1kHz. [Parasound P5 & A23 Pre/Power] Graph 8: Frequency response of line input at an output of 1-watt into an 8-ohm non-inductive
load (black trace) and into a combination resistive/inductive/capacitive load representative of
out of circuit, is very flat (note the different a typical two-way loudspeaker system (red trace). [Parasound P5 Pre & A23 Power Amplifiers]

vertical scale compared to Graph 8) whereas

the lower (red) trace shows the response the two traces, output volume will increase If you are using the tone controls, I’d sug-
with the tone control circuit active, but with when the tone controls are switched out of gest erring on the side of caution, because the
the bass and treble controls at their nomi- circuit, which will tend to make this setting bass tone control offers far more boost (and
nal ‘0’ position. As you can see, although sound ‘better’ than the other, simply because cut) than most tone control circuits—around
the frequency response is still relatively flat it’s louder, and it’s a peculiarity of human 17dB, compared to the more usual 10–12dB.
(1.2dB down at 20Hz and 0.85dB down at hearing that we tend to judge louder sounds The treble circuit also offers an unusually
20kHz), it’s not in the same league as the as being ‘better’… even if they’re otherwise high level of boost, and one that doesn’t ap-
response above it. Note that because of the identical to the quieter sounds against which pear to be particularly well shelved. However,
overall differences in absolute levels between they’re being compared. so long as you use modest levels of boost

30 Australian
LAB REPORT Parasound P5 Preamplifier & A23 Power Amplifier

(that is, don’t turn either control past its ‘3 few harmonic distortion components are pre-
o’clock’ position) I would not anticipate any sent in the output, and the few that are pre-
issues. sent are low-order (and therefore sound pleas-
Channel separation was good, but not ant to the human ear) and very low in level,
particularly so, particularly at high frequen- so their contribution to the overall sound
cies, where it was measured at 49dB. This would be barely perceptible (if at all). The 8Ω
result might have been improved by using result is clearly the better of the two, with
balanced wiring connections between the P5 a second harmonic at –96dB (0.00158%), a
and the A23: Newport Test Labs for some rea- third harmonic at –100dB (0.001% THD), a
son best known to itself used the unbalanced fourth harmonic at –110dB (0.00031%), a
connections. Cable length and routing would fifth harmonic at –115dB (0.0001778%) and
also affect this result. That said, the result is a sixth at –118dB (0.0001259%). The noise
more than will be required for this pair to floor is significantly more than 120dB down
deliver audibly excellent stereo imaging and over most of the audio band, increasing only
channel separation. Level matching between at very low frequencies, where you can see
channels was slightly less accurate than I am it’s at around –100dB. This put the overall
used to seeing, but the 0.5dB mismatch is THD+N figure at a very good 0.01%, as you
tiny and will be completely inaudible, not can see in the tabulated results.
least because it will certainly be less than will Distortion levels increased with increas-
inevitably be introduced by your loudspeak- ing output but even at full rated output—
ers. Plus, of course, you could match the 125-watts into 8Ω (Graph 3) and 200-watts
channels perfectly using either the balance into 4Ω, (Graph 4)—all harmonic distortion
control on the P5 or the volume controls on components were more than 90dB down
the rear of the A23. Interchannel phase errors (0.0031623%) and most were more than
were low, particularly at low frequencies. 100dB down (0.001%). Although higher-or-
Distortion levels were very low across the der distortion components are now present in
board, and particularly low at lower output the output, it is the lower-order components
levels, which will be where this powerful am- that dominate. Note that the noise floor has
plifier will be operating most of the time in dropped down closer to –140dB across most
normal domestic applications. Graphs 1 and of the audio band, though the very low-fre-
2 show THD into 8Ωand 4Ω loads at an out- quency noise (mains and mains-related) still
put level of 1-watt and you can see that very sits at around –100dB.

Parasound A23 Power Amplifier — Power Output Test Results

Channel Load (Ω) 20Hz 20Hz 1kHz 1kHz 20kHz 20kHz
(watts) (dBW) (watts) (dBW) (watts) (dBW)

1 8Ω 159 22.0 159 22.0 159 22.0

2 8Ω 147 21.7 147 21.7 147 21.7
1 4Ω 254 24.0 254 24.0 254 24.0
2 4Ω 227 23.6 230 23.6 226 23.5
1 2Ω 200* 23.0* 378 25.8 378 25.8
2 2Ω 200* 23.0* 325 25.1 325 25.1
Note: Figures in the dBW column represent output level in decibels referred to one watt output.
* Protection triggered. Auto-resetting once power level reduced.

Parasound P5 Preamp & A23 Power Amplifier — Test Results

Test Measured Result Units/Comment
Frequency Response @ 1 watt o/p 9Hz – 57kHz –1dB
Frequency Response @ 1 watt o/p 4Hz – 116kHz –3dB
Channel Separation (dB) 67dB / 66dB / 49dB (20Hz / 1kHz / 20kHz)
Channel Balance 0.5 dB @ 1kHz
Interchannel Phase 0.04 / 0.11 / 2.02 degrees ( 20Hz / 1kHz / 20kHz)

THD+N 0.01% / 0.004% @ 1-watt / @ rated output

Signal-to-Noise (unwghted/wghted) 78dB / 85dB dB referred to 1-watt output
Signal-to-Noise (unwghted/wghted) 97dB / 102dB dB referred to rated output
Input Sensitivity (Bal or Unbal Input) 31mV / 345mV (1-watt / rated output)
Output Impedance 0.01Ω @1kHz

Damping Factor 800 @1kHz

Power Consumption 0.77 / 56.25 watts (Standby / On)
Power Consumption 86.21 / 478 watts at 1-watt / at rated output
Mains Voltage Variation during Test 238 – 245 Minimum – Maximum

32 Australian
LAB REPORT Parasound P5 Preamplifier & A23 Power Amplifier

Thanks to the reduced noise floor, output to 200-watts, at which point the
overall THD+N was measured at 0.004%. amplifier would otherwise have been
Intermodulation distortion was very delivering continuous current of 20 amps.
low, as you can see from Graph 5, which Unlike many amplifiers, whose protec-
shows CCIF-IMD using test signals at tion circuits require you to turn them off,
19kHz and 20kHz. The resulting IMD then back on again to reset the protection
components at 18kHz and 21kHz are circuit before you can continue to listen
around 80dB down (0.01%) and those to music, the A23’s circuit resets itself au-
at 17kHz and 22kHz are nearly 100dB tomatically (and almost instantaneously)
down (0.001%). There is a small signal so in the extremely unlikely event that it
regenerated at 1kHz, visible at the far left trips during normal programme material,
of the graph, but it’s around 92dB down it won’t cause any inconvenience.
(0.0025119%). Square wave performance was out-
Speaking of noise, it was very low, as standing, as you’d expect given the Para-
you can see from the tabulated figures. sound’s performance in the frequency
Referenced to an output of 1-watt, response and damping factor tests. None-
Newport Test Labs measured S/N ratio at theless, the 100Hz oscillogram exhibits
78dB unweighted and 85dB weighted. some tilt, as a result of the overall re-
Referenced to rated output, these figures sponse being 3dB down at 4Hz. The 1kHz
improved to 97dB and 102dB respective- square wave verges on perfection, while
ly, outstanding results both. the 10kHz square wave shows only the
Absolutely outstanding was the A23’s slight rounding that would be expected
damping factor, which Newport Test Labs given the measured 116kHz bandwidth.
measured at 800, off the back of an out- Performance into a capacitive load was
put impedance of just 0.01Ω. This means outstanding, with only a tiny amount
the A23 will be able to control even the of overshoot and only a few cycles of
most undisciplined loudspeaker cones, quickly-damped ring. As a result—and in
thanks to a damping factor that, accord- conjunction with its proven performance
ing to speaker guru Floyd E Toole, is ten into 2Ω loads—the Parasound A23 will be
times greater than will ever be actually completely stable into even the most dif-
required. ficult and demanding loudspeaker loads.

The Parasound P5 and A23 perform

synergistically as an outstanding pre/power
duo, with particular kudos due the A23
The same could also be said of Input sensitivity testing returned the
power output, because the Parasound voltages I’d expect, but I was surprised to
A23 proved to be able to deliver far find that the sensitivity of the unbal-
more power than most audiophiles will anced inputs—31mV required for an
ever require—no matter how inefficient output of 1-watt, and 345mV in order for
their loudspeakers—with Newport Test the amplifier to deliver its rated out-
Labs recording ‘both channels driven’ put—was almost identical to that of the
power outputs of 147-watts into 8Ω loads, balanced input.
230-watts into 4Ω loads and 325-watts Mains power consumption was
into 2Ω loads. The fact that the Para- fairly high no matter what power levels
sound will deliver even higher output lev- the amplifier is delivering, and peak at
els when only a single channel is driven 478-watts, but I’d plan on the pair draw-
mean that it’s capable of even higher lev- ing an average of around 100-watts from
els of ‘dynamic’ power, so you’ll get even your mains in day-to-day use, so less than
more power again on musical transients. a couple of light bulbs.
Importantly, unlike many competitive In standby, the pair drew only
amplifiers, the Parasound was able to 0.77-watts, so if measured individu-
deliver these very high power output ally each one would easily meet the
levels at the frequency extremes (20Hz Australian standard for stand-by power
and 20kHz), as you can see from the re- consumption.
sults tabulated in the power output table Overall, the Parasound P5 and A23
(and shown graphically in the bar charts perform synergistically as an outstanding
that accompany this equipment review). pre/power duo, with particular kudos due
The only point at which the Parasound the performance of the A23.
revealed a limitation was when driving But then given the credentials of
2Ω loads at 20Hz, where the amplifier’s its designer, I would not have expected
protection circuitry cut-in to ‘limit’ power anything less! Steve Holding

34 Australian

Hegel H160
Integrated Amplifier
Marantz CD6005
Yes, you can read some of our reviews in full on
CD Player
Thor Series V You can now swipe your way through
DSP Subwoofer

Yamaha NS-F901
every page of Australian Hi-Fi on your iPad or Android

March/April 2015 | www.av
Auralic Vega tablet, (or computer) using our snazzy Digital Issues.
Digital Audio

AVM CS2.2 But there’s nothing like the proper print edition,

Nat ty
Streaming AV
landing in your letterbox sealed in shiny cellophane,

Nor wegian
ready for you to read at your leisure, at full size, in
test the order we intended, in the bath, if you like.
amplifier to the
We put Hegel’s H160
Get every issue of Australian Hi-Fi the way it should be.


16-bit vsdo24 7 ISSUES (1 YEAR) $59 SAVE 15%
How many bits
need for good sound?
you really

maha NS-F901
A$9.95 NZ$10.99

New Subscription or YaSubscriptionfior

Extend Flawless fit and nish plus Gift Subscription
extraordinary sonic perf
Mr Mrs Miss Ms
4/9/2015 10:25:12 AM

Your Name
HF MarApr15.indd 1
TOLLFREE 1300 361 146 PO BOX 3355
Your Address
Daytime Phone ( )
I enclose a cheque/money order for $
Please provide phone or email in case of delivery issues
payable to Next Media Pty Ltd OR Charge my credit card:
Mastercard Visa American Express
Recipient’s Name
Recipient’s Address
Postcode Name on Card

Daytime Phone ( ) Expiry Date /

E-mail Cardholder’s Signature

Terms and conditions: Price offer available to Australian and NZ residents. Expires 5/7/15. Includes GST. Savings based on total cover price. Overseas Airmail 14 issues

A$195 or 7 issues A$99. Australian Hi-Fi subscription comprises of 7 issues PA – 6 regular issues and 1 special issue. This form may be used as a Tax Invoice. nextmedia
Pty Ltd ABN 84 128 805 970. Please tick if you do not wish to receive special offers or information from nextmedia or its partners via email mail.
Please refer to for the full Privacy Notice. 35