From PY4ZBZ on Mar 31, 2005   (congratulations on the anniversary, Henri !)

 

The adjustment of transmitter power for HamDRM

 

It is necessary to carefully adjust the power output for HamDRM transmissions, because the DRM signal has a very high maximum peak power level.   The following figure shows a DRM signal at 14MHz into a load of 50 ohms.   A Bird wattmeter indicates that the RMS power is 12.5 W.

In the oscilloscope, the peak signal level is 100 V.   Therefore we can conclude that the PEP -- peak envelope power = (100 divided by the square root of 2) squared and divided by 50 = 100 W PEP.

The relationship of PEP / RMS for DRM is therefore 100/12.5 = 8 times or 9 dB !

 

Therefore, a transmitter capable of producing a maximum of 100 W PEP must be adjusted to produce no more than 12.5 W RMS in the DRM mode.   If it is adjusted for higher average power, the peaks of the signal will be clipped and cause serious distortion of the signal, degrading the decoding performance at the receiving station.

This distortion can easily be observed at the receiving station.   The transmitting station only needs to transmit the DIGTRX “Tune” signal:

This signal contains 3 tones: A = 1850 Hz, B = 1475 Hz and C = 725 Hz.  (These frequencies correspond to the reference pilots of the HamDRM signal).

If the transmitter is distorting the signal, many other signals called intermodulation products will appear in addition to the 3 original signals. The following figure shows an example of a transmitter badly adjusted (or with poor linearity), compared against the signal of a well adjusted transmitter with excellent linearity, where only the 3 original tones appear.   In practice, signals with low level distortion are acceptable.

In the case of the badly adjusted transmitter (or one with poor linearity), at least four products of intermodulation distortion, can be seen, as in the example on the left above:

1 - the lowest one in frequency is 325 Hz = 3C - A   (fourth-order product).

2 - another one is 1100 Hz = 2B - A   (third-order products).

3 - another one is 2225 Hz = 2A - B = 2B - C (third-order products).

4 - another one is 2600 Hz = 3A - 2B  (fifth-order product).

This occurs with only 3 original tones!   It is easy to imagine what happens with the 51 tones of the DRM signal: thousands of intermodulation signals will appear, creating an unusable signal !

The power amplifier of the transmitter has a transference curve "output power y ,versus input power x ".   This curve should be LINEAR, or a STRAIGHT LINE, of the type y = a x.   But in it practice, this curve is linear only for low power levels, and it flattens out when maximum power is reached, presenting a component of the type

y = a x 2 or y = a x 3.

It is these quadratic and cubical terms that cause harmonic distortion and intermodulation distortion.   Remembering the trigonometrical relations, it is easy to deduce that the quadratic term applied to the sine function generates another one with double the frequency  (the second harmonic) and the cubical term generates the third harmonic.   Also, the square of the addition of two sine functions results in two others with frequencies that are the sum and difference.   Therefore, the distortion products of intermodulation distortion correspond to a combination of the sums and differences of the original terms and its harmonic ones. An IMD product's order is the the sum of the coefficients (harmonic) of each component.

 

The basic adjustment of transmission power is made using the DIGTRX slider "TX wave out level", and it serves as a fine adjustment.

Keep in mind that this adjustment is the same as the "Wave" slider in the "Volume Control" for the sound card. The "Master volume control " slider acts as a coarse adjustment.   It is also necessary to make the proper adjustments of the modulation level of each transceiver.

The measurement of third order intermodulation distortion, IMD, confirms distortion in the transmitted signal when sending the tuning tones.

It must be more negative than -15 dB (<-15 dB).   The measurement is made by comparing signal 2 with the tone B.

 

 

Effect of the width of the SSB filter.

 

The SSB filter in some equipment is not wide enough to properly transmit and/or receive HamDRM with a bandwidth of 2.5 kHz.   The following figure shows an example of equipment with an SSB filter that suppresses the frequencies above 2.6 kHz (Fig.2).   The lack of signal can clearly be seen, as indicated by arrows 1,2,3 and 4 of fig.2 in comparison with figure 1.   A DRM signal with 2.5 kHz bandwidth and 350 offset occupies an audio bandwidth from 350 to 2850 Hz. 

In the case of this equipment, it is evident that it is better to transmit DRM using the 2.3 kHz bandwidth setting as can be seen clearly in fig.3, where the upper limit of the band is clearly visible (arrows 5 and 7).   Arrow 6 shows a flatter response of the "transfer function" when compared against the one shown by arrow 3.

 

Updated May 7, 2005    PY4ZBZ & KB4YZ