Best Way to Attenuate
Eeyore wrote:
west wrote:
Can we take an off the shelf CD Player and install , let's say a 10K stepped
attenuator or Alps type pot inside the CD player on the output line? Would
this make a cleaner audio input to the tube amp?
Cleaner than WHAT ?
The 2 alternatives are an outboard passive "preamp" with let's say 100k
attenuator
Bad choice. As previously described, 100k is too high a value if you're
concerned about possible treble loss with longer cables.
Use 10k. No CD player will have any trouble driving it.
or installing it inside the amp. I tried the passive preamp route with mixed
results but got straightened out from info in a previous thread by using a
much shorter
interlink cable.
What exactly are you concerned about ?
Graham
I know a guy who never would use a 100k pot or switched R bank as an
attenuator.
He prefers to have 1k from the input terminal to an amp input, and
shunting pot
from input to 0V, and thus the series R from source to amp is never less
than 1k.
he says all the CD players have Rout = approx 500 ohms, and
effortlessly can produce say 3Vrms into 1k, ( 3mA rms current ) at
0.001% N&D,
so having to produce 0.1Vrms for average levels for a power amp isn't a
problem.
Many preamps have an attenuator BEFORE a gain tube, so a 1Vrms signal
arrives
from a CD player and is amplified to 10Vrms, and then attenuated down
-40dB to 0.1Vrms
for feeding into a power amp.
Fortunately, its easy to maintain good enough linearity with a triode to
do all this,
but its better if the triode has a CCS anode load and some shunt NFB to
reduce A from say
16 max to 4, and all aspects of technical operation are vastly improved.
Crown opt for using a pot in a shunt FB path to attenuate the signal
around an opamp
in a preamp.
Using a 100k pot before a gain tube isn't all that bad. Usually SNR is
still good enough, and N&D remain very low because maybe only 0.1Vrms
needs to be
produced by the tube, with the gain pot at the input set at -40dB where
its
resistance in shunt with the grid input is 1k, and HF losses are
entirely negligible.
The highest Rout for a 100k pot fed with a lowZ source is 25k if the pot
is set at -6dB,
which is rarely ever used; its always around the -20dB position, noon
position,
on a log pot. Rout is around 9k.
If the pot feeds a cathode follower output buffer of a preamp, the HF
losses are utterly negligible.
There are many ways for anyone to try arrangeing the source - attenuator
- pre-amp - power amp
line up, but hacking into the guts of a horrible pcb board in most CD
players
with micro circuitry to instal some attenuator is never going to be
something I would do.
Many CD players already have a facility for level control, and come with
a remote
for total control of the player from the lazy chair.
Patrick Turner.
|