Results of op-amp swap for Sony CDP-X339ES (CD player)
Julian Vrieslander
the Sony CDP-X339ES CD player and the op-amp upgrade that I had planned
for that unit. Since this type of tweaking seems to be quite popular, I
thought I'd post a note on the results I obtained.
The 339ES uses Sony's CXD2562Q DAC. This is a one-bit pulse length
modulation convertor, clocked at 45 MHz, with balanced (differential)
outputs for each channel. Although Sony does market other convertors,
this one is apparently a proprietary part - they would not provide me
with any data other than the pin identifications shown in the 339ES
service manual. The outputs of the DAC feed a simple analog circuit
using two dual op-amps per channel (Mitsubishi M5238 and JRC 5532). The
op-amps provide differential-to-single-ended signal conversion, low-pass
filtering, and buffering for the outputs of the player.
My impressions of the stock player's performance were generally
favorable: a smooth detailed sound, decent imaging, and good bass. But
previous experiments led me to suspect that the 339ES could be improved
by swapping out the 5532 chips. I had no previous experience with the
M5238 op-amps.
The op-amp chips were replaced with four DIP sockets (machined-pin
gold-plated sockets from Texas Instruments). I picked out some
candidate replacement chips from my current home inventory: the Analog
Devices AD827JN and the Burr Brown OPA2604AP. Pairwise listening tests
were done with these and the original chips. Interchanging the op-amps
between trials required about 2 minutes. The tests were not blind.
I am not representing this procedure, or my conclusions, as definitive.
Although I usually like to back up such tests with distortion and noise
measurements, I currently do not have access to the necessary equipment.
The best subjective results were obtained with the AD827 installed in
all four locations. They seemed to do the best job of recreating subtle
features in the characteristic sounds of instruments. It became easier
to hear the transient sounds of fingers hitting and releasing acoustic
guitar strings, the changes in the overtones of a cymbal as its sound
decays, the tiny breathy noises from a sax, etc. On some recordings,
stereo localization seemed more precise. The AD827 has poorer noise
specs than the other amps mentioned in this note, but I could hear no
noise at my normal listening levels. The OPA2604 amps, which I have
used with good results in other mods, did not sound as good as the AD
parts in this application.
The PCB's in the 339ES are quite spacious, and it would be easy to
install other mods: capacitor swaps, faster voltage regulators, etc. But
I'm pretty happy with the sound of the Sony in its current state, so for
now I'll just kick back and listen.
--
Julian Vrieslander
Neurobiology & Behavior, Mudd Hall, Cornell University, Ithaca NY 14853
INTERNET: ea...@theory.tc.cornell.edu BITNET: eacj@crnlthry
UUCP: ..cornell!batcomputer!eacj
Daniel Moos
>
>The best subjective results were obtained with the AD827 installed in
>all four locations. They seemed to do the best job of recreating subtle
>features in the characteristic sounds of instruments. It became easier
>to hear the transient sounds of fingers hitting and releasing acoustic
>guitar strings, the changes in the overtones of a cymbal as its sound
>decays, the tiny breathy noises from a sax, etc. On some recordings,
>stereo localization seemed more precise. The AD827 has poorer noise
>specs than the other amps mentioned in this note, but I could hear no
>noise at my normal listening levels.
>
I know I'll probably get a bit of heat for this, but...
Is it possible that a bit of "noise" does for the analog side of
playback the same thing
that dithering does for digital? I.e., brings up the intelligibility of
sub-resolution information(?) So that a slightly noisy analog source will
allow the ear to recover a bit more ambient information???
And, of course, by that token explain a bit about some people's preference
of LP's over CD's?
Dan Moos
PS - did I add enough question marks to avoid being flamed for idiocy? ;-)