These tests were doe using a small AM transmitter. The signal source was a CD player playing:
Track 79 of the Denon Test CD, White Noise - Level 0dB
Track 57 of the Denon Test CD, 1001 Hz Sine Wave -20dB
Initially the levels were adjusted so that the white noise was generating almost 100% modulation.
The Sine wave then was probably about 10% Modulation.
This is the output from an RF diode detector pickup as seen on an audio spectrum analyser on my PC.
The roll off on the white noise line past 9khz is due to the anti aliasing filter in the sound card, not bandwidth
limitations in the transmitter.(Nice wide band transmitter eh!). The 2nd and third harmonic distortion products from the detector changed noticably with
the level into the detector. I was able to tune the detector for least 2nd harmonic. Obviously even this high level doide
detector has quite a bit of non linearity.
This next chart is the same detector with different percentages of modulation
I used slightly differnt frequencies for the different modulation percentages to make the levels easier to see.
At 50% mod the third harmonic is 50dB down from the fundamental. At 16% mod the 2nd harmonic is 50dB down from that
fundamental At 5% modulation the 2nd harmonic is 55+ dB down from the fundamental.
Anyway this gives me an idea of the distortion etc. for my source signal. The question is, how will a reciver handle this?
This reciever dates from sometime between 1983-1992. I use it mainly for AM receive.
Here is the first set of distortion and frequency response I did
So whats going on here? The shape of the Wide AM passband is what I expected, but there is only about 20dB of attenuation
at 8Khz out. Thats pretty bad.
Measuring the distortion reveled what is happening.
At 50% modulation the 2nd harmonic is less than 30dB down. At 5% the 2nd harmonit is just about 40dB down (1%
distorion(THD)) Most, if not all, of the distortion seems to be happening in
the detector or audio stages.
When I applied the 100% mod white noise, the audio stages are generating distortion products thata are masking the true
frequency response. I then did a run with very low level modulation wite noise
The frequency response from the red line is now more like what I expected, dropping to 40dB down by 6.5Khz.
This at least shows that the radio has response bout 6dB down at 4Khz, hardly wide band but better than many other
radios.
This next chart is of the 1001Hz 10% mod. signal at different strengths. The level of the tone is constant but the noise floor comes up as the signal drops off.
The Next Chart is of my DX380. This was measured across an 8.2 ohm resistor across the headphone terminals, with a
signal showing all bars on the strength display.
The wide filter obviously rolls off at 3Khz. The tone switch adds a bit more roll off, not that you need more.
The narrow is a bit too narrow for AM in my opinion, but hey I bought it didn't I?
Interestingly there is less detector/audio distortion than in the R2000. There is probably still some, as the high low
tone switch has quite ann effct on the signal out at >6000Hz. This is no doubt due to rolling off the harmonics due to
distortion.
The ultimate rejection still seems to be better than the R2000 which is odd for a compact radio.
More to follow .....