I understand that an impedance mismatch causes part of a signal to be reflected back toward the source, but the physical mechanism behind it isn’t clear to me.
The case I keep coming back to is series termination at the driver. The standard explanation is that matching the source impedance to the transmission line prevents the returning wave from being reflected again, but what is actually happening at that boundary?
When the reflected wave reaches the source, how does the source impedance absorb the wave instead of sending another reflection back toward the load? What is happening to the voltage, current, and energy at that point?
I can follow the equations, but I’m trying to develop an intuitive physical understanding of what happens at the discontinuity. Any explanations or recommendations for good resources would be appreciated.
A transmission line behaves as a continuous LC network that stores and transfers electromagnetic energy as the signal propagates. Because the signal travels at a finite speed, the source has no information about the load until the wave reaches it and any reflection returns.
When the reflected wave arrives at the source, the source impedance acts like any other load on the line. If that impedance matches the characteristic impedance of the line, the returning wave sees the same V/I ratio it carried on the line; its energy transfers into the source resistance and is dissipated as heat. No mismatch at that boundary means no mechanism to generate another reflection.
If the source impedance is too low or too high, the returning wave cannot transfer all of its energy into the source. The remaining energy has to satisfy the new boundary conditions, so another reflected wave is launched back onto the line. The source termination is simply the load seen by the returning wave: the physics at that boundary are no different from what happens at the far end.
Characteristic impedance sets the fixed ratio between voltage and current for a traveling wave. On a 50Ω line, a 1V traveling wave carries a 20mA current; that ratio holds as long as the impedance doesn’t change.
When the wave reaches a section with a different impedance, the incoming wave carries a V/I ratio the new line can’t match. Voltage and current must both satisfy the boundary conditions at the junction, and the incident wave alone can’t do that, so a reflected wave is launched. The incident and reflected waves together satisfy those conditions.
At a matched source impedance, the returning wave already carries the correct V/I ratio for that boundary. Its energy transfers into the source resistance and is dissipated as heat; no mismatch means nothing is reflected back.
Transmission-line equations permit energy to travel in only two forms: forward-propagating waves and backward-propagating waves. In a forward wave, voltage and current always maintain a fixed ratio equal to the characteristic impedance of the line.
At a discontinuity, that ratio no longer satisfies the boundary conditions the new impedance demands. A backward-propagating wave is therefore required, and the superposition of the forward and reflected waves produces the voltage and current that do satisfy those conditions.
When the line is terminated in its characteristic impedance, the arriving wave carries exactly the voltage-to-current ratio that boundary requires. The boundary conditions are met by the incident wave alone, so no reflected wave is generated.
One point that hasn’t been mentioned yet is why the reflected wave effectively “sees” only the source impedance, even though the driver is actively driving the line. For a linear driver, superposition applies: the driver’s output is treated independently of the returning wave, so the ideal voltage source presents zero small-signal impedance. That leaves the series resistor as the impedance seen by the reflected wave.
If the series resistor equals the line’s characteristic impedance, the returning wave already has the correct voltage-to-current ratio for that boundary, so no additional reflected wave is needed. If the impedance is different, the boundary conditions can’t be satisfied by the incident wave alone, and another reflection is generated toward the load.
In that sense, “the source absorbs the wave” simply means the source termination provides the impedance the returning wave expects.
Another way to think about it is in terms of settling behavior. With a matched source impedance, the returning wave is absorbed at the source, so no new reflection is launched and the line settles after the expected round trip. If the source impedance is mismatched, only part of the wave is absorbed and the rest is reflected back onto the line. That repeated exchange of energy is what produces ringing and causes the signal to converge over multiple round trips instead of one.