Re: Questions on impedance matching
David, I believe the following to be accurate but I am no physicist :-)toggle quoted messageShow quoted text
The reflection arrives back at the source with a delay that is proportional to twice the distance
between source and load, the reflection is travelling at approximately half the speed of light.
For the sake of argument assume the signal is a 20kHz sinewave, the period is 50usec, the
half-period is 25usec. Here is a question, given that the reflection travels at half the speed of
light how long would the line have to be for the reflection to arrive back at the source with a
delay of one half period, 25usec?
This question can be rephrased: The line is 5 feet long, the reflection travels at half the speed
of light, the signal is 20kHz sinewave, the reflection arrives having travelled 10ft with a small
delay, what fraction of one period (50usec) is this delay?
The amplifiers in question are negative feedback amplifiers. The physical length of the
feedback path around which the feedback signal must travel from the amplifier output in
order to reach the input network and do its work is a small distance, inches. The feedback
signal travels this path at approximately half the speed of light. The reflection signal arrives
back at the source after a delay analysed in the paragraph above. It is attenuated and the
residual is summed at the output and combines with the feedback signal that travels back to
the input network of the amplifier. Because the distance around the negative feedback loop is
small the feedback signal arrives back at the input very quickly, and long before the reflection
arrives from the load which is 5ft away. The situation might be summarized this way: the
negative feedback loop is very fast compared to the reflections coming back from the load.
Errors induced by reflections are cancelled out immediately by the feedback loop.
On 13 Feb 2018 at 8:55, David Berlind wrote:
Thanks so much for your various responses.The link to Dr. Leach's work is FANTASTIC.
Probably enough bedtime reading for a year or more (at the pace I consume such texts). I
may have to build that Fisher KX-200 for fun. The voltage-db and power-db conversions were
very helpful. In your opinion, is there a typical target decibel improvement? Regarding
reflections, I'm glad I have company in @jafinch78 on the confusion front (I'm not alone!).
You said "In this condition reflected energy is attenuated by the source impedance behaving
as a low impedance sink." I thought I saw or read somewhere that the refections interfere
with the output signal in transit. For example if the reflection is 180 degrees out of phase, it
practically cancels the signal. Do I have that incorrect? I'm trying to understand why distance
or frequency renders this effect moot in audio. Seems like it's just physics to me (yes, we're
at the limits of my understanding!).