Well, the tube modifications may not be for everyone.
It may not be your cup of tea.
It may repell you for no reason at all.
Some people will say that they slept through the Ohms-Law lesson.
Others will complain, that the tubes are not true, that they color the
sound with pleasant 2-nd harmonics.
Others simply can't be bothered.
Someone may not have the guts to be so radical.
Last but not least - someone may not afford the lampizator job - that
is understandable too.
The good news is that we can make HUGE PROGRESS without any tubes
Here will follow simple straightforward recipes for four great sonic
improvements using just a pair of snips with occasional help of the
soldering iron and a meter.
So lets do it folks! Lets SNIP IT !!!
I can see 4 important areas of improvement that are totally cost-free
and take just minutes, plus they are all easily reversible.
And the result is 100% guaranteed every time, unlike some hard-ball
mods which for some - may be questionable.
SNIP ONE: CD output filter.
The big secret that I am about to reveal for free (that's why
I am a self proclaimed Prometheus, remember) (by the way, I don't
know if the eagle will ever eat my liver, so I am trying to destroy it
myself with red wine) is that every budget CD player that I have ever
seen - and by that I mean a player lighter than 10 kilos or cheaper
400 Euros - has OUTPUT FILTER CAPS. These are two small brown (usually)
tablettes nearest thing to the output RCA.
This looks like this:
The HF filter caps in AKAI CD-55 player - marked with red dots.
This is Technics SL-700 player with two blue capacitors.
This is otherwise excellent ZERODAC with two 220p capacitors nearest to
Otherwise excellent Nakamichi CDP2-e with snipped caps.
I think you get the idea.
The purpose of shunting the signal to the ground just before output is
that it shall ground (kill) high frequencies which are trebles,
midrange upper harmonics, and some noises too.
The idea behind it was that the designers were forced to play the SPECS
game. Every CD in the 80-ties and 90-ties was being compared to the
others by specs. And because the spec of CD - when it was introduced
back in 1982
- categories were derived from the cassette, the LP, the tuner and the
reel to reel, they were quoting signal to noise ratio (S/N), wow and
flutter, and dynamic range. All completely meaningless when applied to
CD. But to be ABSOLUTELY SURE that the CD does not produce ANY noise,
they sacrified the trebles.
I mean - the CD was supposed to be PERFECT as compared to noisy LP and
tuner and cassette. So IT COULD NOT HISS, NO MATTER WHAT !
Nobody blamed these CD players for having no trebles and no "air", but
it hissed with ear to the tweeter - boooooo! That was a no-no. No good
for CD !!!!
Anyway, we can now call these engineers names like morons, idiots,
audio-terrorists - enough said. What we don't want is to spread
hatered, we want to snip these caps out and enjoy for the first time
the REALLY OPEN SOUND.
One - two - three - SNIP!
two - Amplifier input filter
Number two is a really easy one. Some paranoid engineers from AMPLIFIER
department did not trust their colleagues from CD department and so
they added similar caps at every INPUT of the amp. Just in case.
So just open the amp, look on the small PCB where the RCA's are
mounted, and identify a pair of small caps near them.
SNIP them as well...
Now you will begin to hear niuances and some air too.
The real ear-opener is when you do it on BOTH - the CD and the amp.
And just imagine, you were just about to go to your very friendly and
helpful dealer, who never let you down before, and ask him for
recommendation for an IC cable which would make your system a little
brighter and more airy. You were prepared to spend 300 Euros on a good
The MARANTZ amplifier after the input PCB has been cleaned of the
The removed caps.
Do you see what I mean ????
THREE - CD output stage
This is becoming somewhat more advanced.
We will BYPASS the entire output stage of the CD player (filter capc
included) which is useless and unnecesary, as explained elsewhere in my
The output stage is something which is mandatory and unnecessary at the
same time. Don't ask why. This stage is a waste of time, waste of
of space and waste of music as well.
Lets consider two scenarios:
A) The DAC chip has voltage output : we will find the V out R and V out
L legs of the chip and wire a cable from that point to the newly
installed (in 10 mm pre-drilled holes) RCA sockets.
New RCA's are a
must. I pay for mine 2 Euro per pair. The old openings are just perfect
size - diameter is 10 mm.
The wire must be
connected to the RCA via a foil capacitor like MKP or better paper in
oil, value of
capacitance between 0,47 and 4,7 uF. Voltage is unimportant.
So first we go to www.alldatasheet.com, dig out our dac chip, download
the PDF, read it, find the legs etc.
Then find the respective legs in the real life - on our PCB.
Then solder your wire to first point AFTER this leg to avoid damaging
the chip itself by heat.
resostors smd - with red X soldering signal source points
Then CUT the remaining track by means of scalpel-cut or by lifting a
resistor or something like this.
Wire the wire to RCA, sit back and enjoy.
Note that the caps are necessary to de-couple the DC voltage component
present on these dac legs - usually from the range between 2,4 to 3,5 v
This bypass is giving us the purest possible sound from a given
configuration. We are eliminating usually betweem tens and even
hundreds of electronic components which play the role of muddying the
In SCENARIO B - where the DAC chip is a current output type - we can't
do such simple bypass, but the one we will make is simple enough.
So we identify the NEAREST op-amp to the DAC - usually a pair.
Then we must determin, which type of op-amp they use - dual or single.
(usually dual because it is cheaper).
In single opamps the input is leg 2, inverted input is leg 3, and
output is leg 6.
In dual op-amps - the input one is leg 2, inverted one is leg 3, and
output one is 1.
The second half is input = 5, inverted input = 6, output is 7.
Now backtrace the signal - which leg connects to the DAC output. Say
we have a dual op-amp, we trace that the DAC leg 6 (like in TDA1541) is
wired to opamp leg 2 , so op-amp output is leg 1.
If the DAC was wired to leg 5, output would be 7. If the op-amp was
single one, the DAC would have been wired to leg 2 and output would
have been leg 6.
ABOVE - The new RCA sockets and MKP Arcotronics caps in bypass of a
PICTURE: Single op-amp
Picture: Double op-amp
op- amp applied as stereo
leg 11/14 is Iout
Anyway - we wire our signal from FIRST AVAILABLE op-amp OUTPUT to the
RCA via same capacitor as in scenario A, so the Current Iout
signal is being converted to V out by so called "half op-amp
conversion". So after the first opamp the scenario B becomes A again.
The difference is that we CAN NOT cut off the existing following path
because it contains the local feedback loop.
Picture: signal pick-up point.
Picture: different signal pickup point in TDA1541A player.
now it is perhaps a right time to mention the snags, the things that
may go wrong.
Some DACs more than others have some residual digital square-wave
byproducts of conversion process mixed into the music. This is being
filtered in the analog domain after the signal leaves the DAC by means
of op-amps with heavy capacitive local feedback crearing a firewall for
Some DACS do not need this, some - do need a LOT of filtering.
In particular, Sony DACS CDX / CDA need filtering, as well as JVC,
Pioneer, so I drew the conclusion, that the Japanese school of
conversion is noisy, and USA/European - deals with this artifact inside
Burr-Brown, Philips, Wolfson , Cristal Semiconductors, Analog Devices
and Cirrus Logic are all very
clean and silent DACS.
Shall you hear any problem with the "cricket in a tweeter" and it
bothers you, I suggest to experiment with passive RCR filter. Use the
filter discretely, better to under-filter than over-filter.
I suggest to choose a signal stealing point from the DAC, add R=1K in
series, capacitor 1nF - 10 nF to ground, and again - a 1K resistor.
This should bring the noise do a level acceptable from listening seat.
Of course if you crank the amp all the way and put the ear to the
tweeter - you will still hear the hiss. I say - be NORMAL and go back
FOUR : AMPLIFIER INPUT STAGE ELIMINATION
Number four gets even more tricky but the reward is well worth the
I am talking about integrated amplifier INPUT BYPASS.
If our amp has tone controlls, input selector switches, balance
controll and loudness - we want to bypass all that.
In case of my favourite DUAL CV40 amp (made circa 1969) there is a
superbly clean and simple power stage preceeded by VERY complex input
SEE THE SCHEMATICS HERE
Just lift the left leg of the capacitor C21 and connect the input
signal from RCA straight to the lifted leg (via a potentiometer).
What I did I drilled new speaker terminals at the back as well as new
RCA sockets to replace these DIN ones.
I wired the RCA directly via a cable to the volume pot, but first I had
to isolate the pot "legs" from EVERYTHING ELSE.
A freed up potentiometer is grounded on one side, and the ground also
is wired to RCA ground.
The pot OUTPUT TAP (center) is wired to the power board input point.
Then I realized that the volume pot is a wrong type, I needed something
from 47K-100 K range, so I used BASS pot instead. It has the right
type. I isolated (floated) the legs and wired input to one side,
output to centre and ground to the other side.
The cheap and simple transistor amp becomes a giant killer in terms of
sound quality. It rivals the medium range of tube amps, like push-pull
el-34 kind of type for sure.
Having done all that - the snip number 3 and 4 - we do not need anymore
the type 1 and 2 as described above - we get VERY VERY close to decent
sound at no cost at all.
If we buy ready products, the DUAL amp (or any other amp where at the
back side you see four TO-3 transistors (looking like a Stetson cowboy
hat). It will cost circa 50 Euros, and a decent CD like a philips,
yamaha, grundig, marantz, technics, cambridge , pioneer or
kenwood fore example will cost even less.