Re: Questions on impedance matching

Ted Rook



Compared to the precision and rigor that are found in RF work computing and video the
world of audio can sometimes seem sloppy by comparison. Our hearing is a something of an
enigma because in some ways it is sophisticated, sound source location, and frequency
discrimination for example, and in some ways it is primitive, the thresholds for just detectable
changes can be quite large under some conditions and quite small in others. For an audio
device to reproduce sound free from obvious degradation does not require sophisticated or
expensive hardware, this was achieved in the 1950s with tube amplifiers and moving coil

There is a website you may find helpful, it is what remains from the work of the late Dr Jim
Leach at Georgia Tech, there is probably a link there to his course text book


Ted Rook

On 12 Feb 2018 at 20:05, Ted Rook wrote:

Yes I agree, with 600 ohms attached the terminal voltage is the nominal value, with open
circuit it is double.

In audio we NEVER match source and load impedances, not with small signals nor with
power amplifiers and loudspeakers. The universal system comprises low source impedances
and high load impedances, ratios of 1:10 to 1:1000 are common.

Things are happening in audio at such a slow speed that reflected energy is irrelevant.

Reflections don't cause "distortion" in audio. Distortion in audio occurs due to non-linearity in
the amplification and from abuse of amplification by the connection of load impedances
below the rated value.


ratio of two voltages expressed in dB is 20 times the log base10 of the ratio

in round numbers:

double is +6dB
half is -6dB
3x is +10dB
10x is +20dB
100x is +40dB

ratio of two powers expressed in dB is 10 times the log base 10 of the ratio

in round numbers:

double is +3dB
half is -3dB
10x is +10dB
100x is +20dB

Hope this helps


On 12 Feb 2018 at 6:34, David Berlind wrote:

thank you @tedR: when you say "if it (the 600 ohms) is absent," you actually mean if there's
no load at all. Because, according to the math, other loads (above or below 600 ohms)
should not necessarily yield double the voltage. It will be somewhere in between. I'm trying to
understand the use cases for where maximum power transfer (watts) trumps maximum
voltage (amplitude) and vice versa. My assumption is that as you move off the maximum
power transfer point (where impedances match), you get signal distortion due to refection
and at that point, maximizing voltage means you're just maximizing a distorted signal which is
not helpful in radio or cable communications, but might be desirable in audio. The
relationship to dB is a new dimension for me. What's the math behind any dB calculations
and what is optimal? What dB am I shooting for?

Join to automatically receive all group messages.