What It All Means: line-level preamplifier measurements

Though we sometimes take for granted that the basic "language" of our measurements is clear to all of our readers, letters to the editor tell us that this is not the case. Periodically, then, we will attempt to explain exactly what our measurements are and what they purport to show. Though those with technical training may find our explanations a bit simplistic, they're aimed at the reader who lacks such experience.

Perhaps the simplest type of audio product is a basic, line-level preamplifier. Such devices accept line-level signals: ie, a signal with little current flow, typically in the high millivolt to low volt range. They deliver a usually amplified line-level signal, generally to a power amplifier. Such devices are shielded from the interface problems inherent in transducers at the driving (phono cartridges) and receiving (loudspeakers) ends.

JA explains in the sidebar what is meant by the output, or source, impedance of a device and how we go about measuring it. The input impedance is measured with similar indirection. Using a fixed input voltage (usually 100mV) at a frequency of 1kHz, the output or source impedance of our Audio Precision Dual Domain System One test device is set to 25 ohms (50 ohms for a balanced output/input lash-up), and the output voltage of the device under test (DUT) is measured. Then the source impedance of the test set is changed to 600 ohms and another output voltage measurement is made.

A simple computer program written by JA takes these measurements and calculates the DUT's input impedance by comparing the relative voltage "drop" between the two source impedance settings. The output impedance of a source and the input impedance of a DUT act as a voltage divider. Assuming the voltage of the source remains unchanged, the higher the input impedance of the DUT relative to the output impedance of the source, the higher the voltage appearing across the input of the DUT---up to the point where the DUT input impedance is so large that it effectively swamps the output impedance of the source, the latter becoming a negligible part of the total. At this point, effectively, there is no voltage divider effect, and the full output of the source appears across the input of the DUT (footnote 1). The output voltage of the DUT---which is where we take our reading---is directly proportional to the voltage seen at the input of the DUT (changed only by the fixed gain of the DUT). By comparing the two output voltages taken at two known source output impedances, Ohm's Law gives the DUT's input impedance.

The higher the input impedance of the DUT, the smaller the change in the voltage readings at the two source impedances, the 25 or 600 ohm source output impedances becoming ever more negligible in providing any voltage division with the input impedance. Thus, the less precise the reading becomes. For example, in the case of the Jadis JPL, for the left channel from the CD input we measured 652.8mV output for a 600 ohm source impedance and 653.3mV output for a 25 ohm source impedance. This computes to an input impedance of 750,695 ohms. A simple 1mV change in the 600 ohm reading, to 651.8mV, would result in an impedance calculation of 249,832 ohms. Fortunately, the stability and accuracy of the Audio Precision's readout is much better than 1mV, but even a 0.1mV error will change the result by 100k ohms. Fortunately the readings are much more accurate at the lower, more common, input impedances.

Footnote 1: This point is never truly reached, of course. Sort of like the old paradox of covering half the distance to your destination with each step, thereby never reaching it. You get pretty close, though.