Sony ES CD player mods performed succesfully
Marc Yun
In article <4smg02$1...@agate.berkeley.edu>, mba...@midway.uchicago.edu
(Miguel Barrio) wrote:
> OK, the mods are the following:
>
> - Change NE5532 opamps for Burr-Brown's OPA2132.
> - Add a combo of caps (470uF in elect., 10uF in tantalum and .1uF in
> ceramic) to the common supply rails of the opamps. That is, one combo
> for the V+ and one for the V-. The rails are wide and go straight to
> the input pins so they don't have much inductance.
>
> That's all I did, it took about an hour of solder work and stuff.
A couple comments....
- You know, you might want to install a IC socket where the NE5532 was
so you can easily play around with different chips and find one that
suits you best... then you could try the Burr-Brown OPA2604 and Analog
Devices OP275 also, as well as almost any other dual op amp. When
you've decided on one that you like you can then permanently solder it
in.
- Why did you bypass the elect. with a tantalum and ceramic? I know
that's what all the databooks suggest, but I dunno how good it
sounds. Tantalum is generally not used in audio applications, and most
designers avoid ceramic like the plague. Here's what type of bypass I
think sounds really good:
-470uF Nichicon Muse
-10uF Black Gate Standard or BG-N
-0.47uF Black Gate BG-NX
-0.01 Rel-Cap RT polystyrene
By staggering progressively smaller and higher-quality caps, you might
get improved High-frequency performance by avoiding the larger
capacitor's area of resonance, as well as their parasitic inductances,
HF distortion, and so forth. If you get a chance try that combo out on
your power supply lines... you might be suprised at just how good
whatever op-amp you use can really sound. I think this might give you
a very noticable improvement in treble smoothness and extension,
transparency, detail, bass power, and overall musicality. In fact,
someone tried out a bunch of capacitors in his Assemblage DAC-1, and
he said changing the type of a pair of 10uF caps in the supply could
really affect the sound.
Anyway, it's good to hear things worked out nicely for you... keep up
the good work, and remember, before you buy, try DIY!!!
Regards,
Marc Yun
my...@sas.upenn.edu
Pete Goudreau
A couple of thoughts here,
> In article <4smg02$1...@agate.berkeley.edu>, mba...@midway.uchicago.edu
> (Miguel Barrio) wrote:
>
> <snip>
>
> - Why did you bypass the elect. with a tantalum and ceramic? I know
> that's what all the databooks suggest, but I dunno how good it
> sounds. Tantalum is generally not used in audio applications, and most
> designers avoid ceramic like the plague. Here's what type of bypass I
These type caps are generally avoided as coupling caps, not as bypass
caps. The aversion to them as bypass elements results from
misunderstanding their exact application in most situations. In fact,
tantalum caps make excellent high frequency bypass elements as the
rather large ESR overdamps the tank formed by the ESL to the point
that there is a real pole and a real zero that are widely
separated...this inhibit ringing on the power rail during a high speed
current transient...definitely a good thing...
> think sounds really good:
>
> -470uF Nichicon Muse
> -10uF Black Gate Standard or BG-N
> -0.47uF Black Gate BG-NX
> -0.01 Rel-Cap RT polystyrene
>
> By staggering progressively smaller and higher-quality caps, you might
> get improved High-frequency performance by avoiding the larger
> capacitor's area of resonance, as well as their parasitic inductances,
> HF distortion, and so forth. If you get a chance try that combo out on
Actually, what you will get is at least 3 very high Q transmission
zeroes formed by the complex pole pairs from the interaction of the
ESL of the larger cap with the capacitance of the smaller cap for each
adjacent pair in the frequency response. Without the |Z|vs.f curve of
each part, it is essentially impossible to know what the driving point
impedance will be at the bypassed node's location in the circuit. It
would be much better to get this information and work out a network
that provides a constant resistive driving point Z. This will
eliminate any possibility of forming resonances in the power feed
network that will dramatically effect crosstalk, distortion, and high
frequency response of the circuit. Remember that the layout
parasitics will control the highest frequency response so use surface
mount parts and locate them immediately adjacent to the pin to be
bypassed.
> your power supply lines... you might be suprised at just how good
> whatever op-amp you use can really sound. I think this might give you
> a very noticable improvement in treble smoothness and extension,
> transparency, detail, bass power, and overall musicality. In fact,
> someone tried out a bunch of capacitors in his Assemblage DAC-1, and
> he said changing the type of a pair of 10uF caps in the supply could
> really affect the sound.
Uhhh...that was me...
And note that without the correct bypass structure, it sounded quite
bad due to unintentional resonances on the power rails which were
eliminated using the appropriate parts. Those parts were actually low
cost 'lytics (Panasonic HFQ) and specialized surface mount ceramics
(AVX 0612 series). Without the 500pH of the ceramic caps, it is
essentially impossible to provide a resistive driving point Z at any
frequency above a few tens of megahertz, it simply cannot be done with
leaded caps of any type.
If the active circuitry being bypassed has a much lower gain-bandwidth
product, then it is likely not to matter, but if you are using current
mode op-amps and/or wideband buffers, then it is critical that this
matter be addressed properly or the response of the circuit will
depend rather significantly upon the bypass network and this is
generally a bad thing...
<snip>
Cheers,
Pete
Marc Yun
In article <4t31ic$b...@agate.berkeley.edu>, Pete Goudreau
<pgou...@lsi.dsccc.com> wrote:
> > your power supply lines... you might be suprised at just how good
> > whatever op-amp you use can really sound. I think this might give you
> > a very noticable improvement in treble smoothness and extension,
> > transparency, detail, bass power, and overall musicality. In fact,
> > someone tried out a bunch of capacitors in his Assemblage DAC-1, and
> > he said changing the type of a pair of 10uF caps in the supply could
> > really affect the sound.
> Uhhh...that was me...
Sorry Peter, I knew someone here told me about that but I had deleted
my email and I couldn't for the life of me remember who it was...
sorry... :)
Anyway, I had no idea you were so knowledgeable in this area... if I
had known I would have asked you a lot more questions! I'm still
trying to think of a good bypass combination for both digital and
analog supply lines, and the Muse/Black Gate/Rel-cap combo was
something I arrived at rather arbitrarily at the top of my head,
though I have used similar structures in my equipment. Could you
suggest an effective combo for the rest of us mere mortal DIYers?
Thanks again...
Regards,
Marc Yun
my...@sas.upenn.edu
Marc Yun
> In article <4t31ic$b...@agate.berkeley.edu>, Pete Goudreau
> <pgou...@lsi.dsccc.com> wrote:
>
> > > your power supply lines... you might be suprised at just how good
> > > whatever op-amp you use can really sound. I think this might give you
> > > a very noticable improvement in treble smoothness and extension,
> > > transparency, detail, bass power, and overall musicality. In fact,
> > > someone tried out a bunch of capacitors in his Assemblage DAC-1, and
> > > he said changing the type of a pair of 10uF caps in the supply could
> > > really affect the sound.
>
> > Uhhh...that was me...
>
> Sorry Peter, I knew someone here told me about that but I had deleted
> my email and I couldn't for the life of me remember who it was...
> sorry... :)
Peter, I checked my email server again and it was actually NOT you who
originally informed me about the cap mod in an Assemblage DAC-1 but a
fellow named Mike Walkford. Mike, if you're out there, my sincerest
apologies...
Anyway, getting back to the subject of caps, I've been wondering, at
what point do large (over about 470uF) electrolytics start exibiting
lowered resonance points that might inhibit HF performance, and how
effective really are film cap bypasses? Film cap bypasses have become
the mainstay of the audiophile tweaker, but some "golden-eared"
designers such as those at Audio Note actually believe that bypassing
electrolytics with film caps can actually introduce colorations, even
in the power supply lines. In any case, with the availability of Black
Gate wide-bandwidth electrolytics I'm starting to doubt the virtue of
multiple "audiophile" film cap bypasses. Any thoughts?
Regards,
Marc Yun
my...@sas.upenn.edu
Pete Goudreau
On 24 Jul 1996, my...@sas.upenn.edu (Marc Yun) wrote,
> In article <4t31ic$b...@agate.berkeley.edu>, Pete Goudreau
> <pgou...@lsi.dsccc.com> wrote:
>
> > > your power supply lines... you might be suprised at just how good
> > > whatever op-amp you use can really sound. I think this might give you
> > > a very noticable improvement in treble smoothness and extension,
> > > transparency, detail, bass power, and overall musicality. In fact,
> > > someone tried out a bunch of capacitors in his Assemblage DAC-1, and
> > > he said changing the type of a pair of 10uF caps in the supply could
> > > really affect the sound.
>
> > Uhhh...that was me...
>
> Sorry Peter, I knew someone here told me about that but I had deleted
> my email and I couldn't for the life of me remember who it was...
> sorry... :)
>
> Anyway, I had no idea you were so knowledgeable in this area... if I
> had known I would have asked you a lot more questions! I'm still
> trying to think of a good bypass combination for both digital and
> analog supply lines, and the Muse/Black Gate/Rel-cap combo was
> something I arrived at rather arbitrarily at the top of my head,
> though I have used similar structures in my equipment. Could you
> suggest an effective combo for the rest of us mere mortal DIYers?
Absolutely, glad to...
As a matter of fact, I just completely changing out the bypass
structure of the CS8412 and SM5842AP (having recently replaced the
DF1700) to match that of the bypass on the analog side, which worked
so well. The results were amazing and not expected. I had initially
bypassed these parts with a 3.3uF tantalum in parallel with an AVX
1210 220nF X7R chip cap and replaced the 10uF 'lytic at the output of
the 7805 with a 10uF tantalum, apparently this combination was better
than stock but not as good as the latest improvement.
The best combination I have found to date that works by calculation,
simulation, measurement, and listening is as follows:
A Panasonic HFQ 'lytic, 120uF/25V paralleled with an AVX 0612
270nF/25V Z5U chip cap paralleled with an AVX 0612 10nF/50V X7R chip
cap. Note that these chip caps are 0612s, not 1206s. This yields a
nearly perfectly resitive driving point Z out to about 90MHz or so.
The residual inductance is on the order of 180pH which is swamped by
the internal parasitics of the bypassed device's leadframe and bond
wires. The 10nF should be mounted to the board first with the 270nF
on top so that inductance at high frequencies is minimized. This is a
very tedious task (having installed 28 sets of these on the DAC-1 I
can swear by that) but it can be done and provides a serious
improvement in performance by itself.
The rest of the bypass structure is defined by the regulator type
feeding the rail. If it is a 7805 or 7905 reg, then bypass it at its
output with a Panasonic HFQ 'lytic, 12000uF/6.3V...if it is a 7815 or
7915, use an HFQ 6800uF/25V 'lytic. These will closely cancel the
apparent inductive output impedance of the regs and provide a nearly
resitive driving point Z at the source of power feed network. Do not
use FA type capacitors as their ESR is too low to provide the correct
damping. Keep all the leads short and it should work just fine in any
other application. As a final measure, use Panasonic FA series caps
as the main filter caps as that is where their very low ESR matters
the most by minimizing ripple voltage due to the ripple current from
the rectified secondary voltage of the transformer...this matters
whether the filter topology is a choke input or a capacitor input
type.
Hope this helps,
Pete
P.S.: And if you ever see this type of bypass in a future
product...remember you saw it here it first...;-)
Jonathan Ives
Following this thread does the Sony have 100uF DC blocking caps on the
OpAmp outputs ? some older/cheaper cd players do as a simple way of
stopping any nasty DC thump at turn on from making it's way into your
pre-amp.
My old Marantz had 2 * 220uF Elna Cerafines back-to-back making a
bipolar in this role. I *believe* later Marantz players use Blackgate
bipolars. Does anyone have any thoughts or reccomendations of the best
type in this role ? I guess no cap is the best cap but I'm in
component swapping mode. Does the quality of the cap just effect the
hi frequency response. My player has an RC filer following on from
this which gives a gentle slope off above 16Khz. I like this as it
makes my CD's listeneable but wonder if replacing the DC blocking cap
is worth while given that further downstream is a filter anyhow ?
Any thoughts greatfully received.
Marc Yun
In article <4tfmca$e...@biosun.harvard.edu>, Jonathan Ives
<ji...@uk.oracle.com> wrote:
Hmm...
I have heard that the absolutely best caps for power supply bypassing
and coupling applications are the Black Gate BGN and BGNX non-polar
electrolytics... especially if you paralell or series them in a
"L-cancelling Pair" configuration...
Haven't tried it myself, but the performance claims are impressive,
and a lot of people who have gone to the expense of using them seem
very satisfied...
Regards,
Marc Yun
my...@sas.upenn.edu
Pete Goudreau
On 25 Jul 1996, my...@sas.upenn.edu (Marc Yun) wrote,
> <snip>
>
> Peter, I checked my email server again and it was actually NOT you who
> originally informed me about the cap mod in an Assemblage DAC-1 but a
> fellow named Mike Walkford. Mike, if you're out there, my sincerest
> apologies...
Will have to find these posts on AltaVista...
> Anyway, getting back to the subject of caps, I've been wondering, at
> what point do large (over about 470uF) electrolytics start exibiting
> lowered resonance points that might inhibit HF performance, and how
> effective really are film cap bypasses? Film cap bypasses have become
> the mainstay of the audiophile tweaker, but some "golden-eared"
> designers such as those at Audio Note actually believe that bypassing
> electrolytics with film caps can actually introduce colorations, even
> in the power supply lines. In any case, with the availability of Black
> Gate wide-bandwidth electrolytics I'm starting to doubt the virtue of
> multiple "audiophile" film cap bypasses. Any thoughts?
<.sig snipped>
'Lytics do not exhibit resonance points, they have a real zero at ESR*C
and another at ESL/ESR. There are some complex response
charactersistics at very high frequencies but they are typically
negligible compared to circuit parasitics.
Film caps almost always have a very high Q resonance as their ESR is
very low due to the method of lead termination. The construction of the
cap defines the ESL and can generate a rather broad range of SRFs.
There is nothing intrinsically better about a film bypass except that
the capacitance is more independent of applied voltage than most
ceramics, keeping the CV product constant in comparison.
Yes, film caps across 'lytics is not a very good idea...this is very
well known in analog design circles. Generally, if 'lytics are to be
used as bypass elements then they are almost always chosen to have
either highish ESR or an external resistor is placed in series to
provide the requisite damping. Without the damping, severe power line
resonances can occur that will actually drive wideband amps into
oscillation or at least wreak havok with their closed loop response.
This appears to be pretty much unknown amongst most audio equipment
designers, for some reason...
Don't know much about "Black Gate" caps except that it is most
unlikely that they even manufacture them themselves as nearly everly
last bit of 'lytic foil is made by NCC in Japan. A company would have
to be mighty big to get them to make even a small amount of custom
foil. I suspect that their caps are in reality either cherry picks or
relabels of existing caps that exhibit ultra low ESL. Remember that
the ESL defines the "bandwidth" of the cap. One can get the exact
same effect using 'lytics aimed at the switchmode power supply market
for a lot less money...
In order to get a resistive driving point Z at the power pin of an
active device, a designer has no choice but to parallel bypass caps as
there exists absolutely no capacitors that can deliver a sufficient CV
product with a low enough ESL...
Hope this helps,
Pete
Marc Yun
In article <4tjp06$p...@biosun.harvard.edu>, Pete Goudreau
<pgou...@lsi.dsccc.com> wrote:
> Don't know much about "Black Gate" caps except that it is most
> unlikely that they even manufacture them themselves as nearly everly
> last bit of 'lytic foil is made by NCC in Japan. A company would have
> to be mighty big to get them to make even a small amount of custom
> foil. I suspect that their caps are in reality either cherry picks or
> relabels of existing caps that exhibit ultra low ESL.
[quote deletion - wsr]
The Black Gates are manufactured by Jelmax/Rubycon I believe, and they
have a very unique construction so I doubt they are made by another
OEM. They use very large quantities of conductive graphite particles
in the seperator paper which they claim increases electron transfer as
opposed to the slower ion transfer on other wet electrolytics. They
make a unique non-polar electrolytic that is especially
intriguing. They claim much lower levels of noise and distortion than
any other wet or dry electrolytic cap, tantalum included. Try to get
the technical papers if you can, they're quite interesting...
Regards,
Marc Yun
my...@sas.upenn.edu