It has been claimed that using an attenuator in parallel with a speaker instead of inline as they are normally used (for a -3dB cut only) will not affect the frequency response, tone, or feel of the amp. This is not the case, because putting a fixed resistor in parallel with a reactive load will indeed change the frequency-dependent "damping" factor - the amp will no longer "see" a fully reactive load, but rather a reactive load in parallel with a resistive load., which isn't the same thing at all.
This parallel resistance will reduce the resonant peak on the lows and reduce the impedance rise at higher frequencies that you get with a speaker, so you will lose some of the natural bass and treble boost you get when driving a speaker, which will result in a more midrangey sound, and the "feel" of the amp will change somewhat. This will be more apparent in amps that do not use global negative feedback.
Following is a plot showing the effect of adding a resistive load in parallel with the normal reactive load of a speaker, in an amplifier with global negative feedback (in this case, a Marshall 50W output stage). The green trace shows the response of the amp when driving an 8 ohm reactive load, with the amp set for an 8 ohm output impedance for correct matching. The red trace shows the same amp driving a 16 ohm reactive load in parallel with a 16 ohm resistive load, with the amp set for an 8 ohm output impedance for correct matching. As you can see, the parallel resistive load increases the amplifier damping factor, which reduces the way in which the amplifier "reacts" to the speaker. Note that the midband response doesn't change appreciably. The overall result is less apparent bass and treble as mentioned previously.
Following is a plot showing the same loads, this time with an amplifier using no global negative feedback. As you can see, the effect of a parallel resistive load is even more apparent, because global negative feedback tends to reduce the effective output impedance of the amplifier, increasing its damping factor, and decreasing its response to reactive loads. Removing the feedback makes the amp respond much more to load impedance variations.
While using an attenuator in parallel with a speaker load may in some cases sound better than using the same attenuator in series, it is not without its own flaws. There will be a change in the amplifier's tone and feel.
Copyright © 2003, Randall Aiken. May not be reproduced in any form without written approval from Aiken Amplification.