Attenuators and the signal path?
Mar 28, 2006 at 6:02 AM Thread Starter Post #1 of 2


500+ Head-Fier
Jul 6, 2004
I am placing this question here because it primarily deals with headphone amplifier designs and/or modifications. If a headphone amplifier is capable of being used as a preamaplifier, but it needs to have its signal attenuated because of a high noise floor, is ther any difference to the signal path if the amp is modified "in the box" (i.e. the needed resistors or other electrical components are added to, or replace existing components, inside the actual amplifier) vs. an external solution like in-line attenuators?

Simply speaking (and I do mean simply because I am not an engineer), Rothwell Audio believes that in-line attenuators are a better solution because they allow a better signal to noise ratio - . I know that my ears will be the final judge, but I am curious to learn more about the often referred to Holy Grail - "the purity of the signal path". It seems that we often look down upon anything we add "outside the box", but many who do so often seem to not know what is "inside the box". Any useful comments or information would be greatly appreciated.
Mar 28, 2006 at 7:33 AM Post #2 of 2


500+ Head-Fier
Jan 22, 2005
If you are interested about hearing about attenuators I would give Ric Schultz of Electronic Visionary Systems a call.

He is super friendly and can talk your ear off. He also knows a thing or two about attenuators

Here is his now expired ultimate attenuators page courtesy of google cache. His web page is The man doesn't fool around when it comes to achieving the ultimate sound. I asked about his system once and he said he doesn't use any attenuation at all but has it hardwired at 80db.

Here is an excerpt from the google cache explaining the different attenuator types which you may find useful.


A shunt attenuator uses a fixed value resistor is series with the signal (nothing else)
and varies the resistance to ground following it. We us a 24 position switch to switch
different resistors to ground. The shunt attenuator has only one element in series with
the signal, so tends to sound the most transparent. The shunt attenuator has a variable
input and output impedance. For instance, our 1K attenuator has at least 1K input
impedance but rises in impedance towards full volume. Some people think that you need
to load the source with a constant input impedance. In our experience of over
14 years with the shunt attenuator we have not found this to be necessary. As long
as the source can drive the input impedance of the attenuator then it works fine.
If you were using a shunt attenuator in a preamp between a phono stage and a line stage,
then there might be some difference of loading on the phono stage, if the phone stage
uses RIAA feedback rather than passive EQ. When using a shunt attenuator the series
element (resistor) can be soldered directly to the circuit with no extra wires in the
signal path.

The Ladder attenuator uses a two pole switch per channel and switches a series resistor
and a resistor to ground at the same time. This type of attenuator costs more to make
than the series type but sounds better. The Ladder type of attenuator has two pieces
of wire, one resistor and two switch contacts in series with the signal. The Ladder type
of attenuator has a fixed input impedance and a varying output impedance.

The series attenuator has a series of resistors that the switch taps off of. So if you
are using a series attenuator half way up, you have half the resistors and older joints
in series with the signal and the other resistors and solder joints to ground. You also
need wires from the switch to the source and load. The series attenuator tends to be the most
veiled but fairly easy to make. It has a constant input impedance and a varying output


Users who are viewing this thread