Audio Response: HP8920 Service Monitor
Much of the TNC testing was done using a service monitor as the signal source
and the receiver. The monitor has selectable pre- and de- emphasis, so I
thought I'd plot and display its results here for reference.
First, the receive side:
This is the audio response with the deemphasis circuit switched off. For all
practical purposes, it's flat through the frequencies of interest. (There is
a 50Hz high pass filter, and a 15kHz low pass filter, in the circuit, but
they don't have any effect at the frequencies we're interested in.)
This is the response with 750uS deemphasis switched in. It's very close to
the theoretical response; between 1200 and 2200 Hz the rolloff is 5.17dB.
Now, on to the transmit side. For these tests, I used the microphone input
connector as the audio injection point. This provided switchable pre-
emphasis, but made level adjustments a bit complex.
This is the transmit response with the pre-emphasis circuit switched off.
There are a 300Hz high pass, and 3kHz low pass, filters in the microphone
input circuit, and you can see their effect at the ends of the chart. The
high pass cutoff is far below the frequencies used by the packet modem, so
it can be ignored. The low-pass effect starts to show up a bit earlier than
would be ideal, but the rolloff is so slight -- only about 1dB -- that it
shouldn't have any impact.
This is the response with pre-emphasis switched in. The boost is a bit
strong; between 1200 and 2200 Hz the increase is 6.2dB.