SM 5 BSZ - Dynamic range measurements in Handen April 2007
(April 22 2007)

Setup to measure transmitter spectral purity

These measurements were made with the same equipment as at previous meetings. See for example Measurements at the SSA meeting in Borlange 2004

Calibration.

This time Linrad was not calibrated so the raw data from the measurements have to be corrected for the frequency response of the RX2500. The table below is extracted from The spectrum of wideband pulses.

Frequency      Correction
  (kHz)          (dB) 
  7.6            2.0
  10             1.4
  15             0.3
  20             0.0
  25             0.0
  30             0.2
  35             0.5
  40             0.9
  45             1.2
  50             1.2
  55             1.4 
  60             1.1
  65             0.9
  70             0.6
  75             0.6
  80             0.8
  85             1.2
  90             2.5
  92.2           3.3


Screen dumps from the Linrad display in tx test mode were saved for the two transceiver types tested at this meeting while they were emitting an unmodulated carrier.

Orion2 (14 MHz)
IC7000 (14 MHz)
IC7000 (144 MHz)

The spectral purity of a continuous carrier is listed in table 1. In this table spurious signals are not included. The IC-7000 has a severe spur at a frequency offset of 34 kHz. It is the mirror image originating in the conversion from 17 kHz to RF. The level is a bit below -60 dB with respect to the main signal so this spur is within the specifications. On 144 MHz a spur like this will be likely to cause problems for fellow amateurs who will mistakenly take it for a DX and first after loosing some time find out that it is the strong spur from someone not so far away. In SSB it is not much of a problem because the sideband is reversed. In receive mode the IC-7000 uses a filter. The image would not be acceptable in a receiver test so it is properly suppressed......

For spurs, look at the screen dumps. Note that the mirror image that is produced because RX2500 is a direct conversion receiver is located symmetrically with respect to the spectrum midpoint.


 Model    Band      Noise floor in -dBc/Hz  
         (MHz)    5kHz  10kHz  15kHz  20kHz  50kHz 100kHz 150kHz  
Orion 2   14    128.0  127.0  125.8  124.9  119.5  128.8  136.2
IC-7000   14    115.1  119.5  119.9  124.5  125.3    -      -
IC-7000  144    109.9  116.5  118.9  122.6  125.7    -      -
Table 1. Noise floor at different frequency separations from a carrier.


Spectral purity of voice SSB transmissions.

The IC-7000 has splatter due to an inadequate ALC design, a problem very common in amateur transceivers. The Orion 2 does not show any trace of such problems but even so, the SSB signal is not particularly clean.

The raw data is here:
Orion2 (14 MHz SSB)
IC-7000 (14 MHz SSB)


                          Splatter level below PEP at      
 Model             Band   5kHz 10kHz 15kHz 20kHz 30kHz   
                   (MHz)  (dB)  (dB)  (dB)  (dB)  (dB)   
Orion 2 proc.off     14    29    47    58    65    74
Orion 2 proc.on      14    29    48    60    66    74
IC-7000 proc.off     14    38    46    52    61    78
IC-7000 proc.on      14    34    46    51    59    75
Table 2. Peak splatter level in dB below peak power at or different frequency separations from an SSB voice signal. For details, look at the spectra.


Spectral purity of keyed CW transmissions.

The Orion 2 has close to perfect CW keying. The ALC problems that were present in Orion 1 have been cured and hopefully recent software for Orion 1 has cured its problems too. Look here for details: CW keying of the Orion 2

An over-active ALC causes keying clicks with a characteristic spectral pattern. Oscillations in the ALC loop causes amplitude modulation because the variable gain amplifier is an AM modulator. The IC7000 is better than e.g. its predecessor IC-706MKIIG, but it does have this problem to some extent as you can see here: CW keying of the IC-7000

Receiver dynamic range of the Orion 2 at 2 kHz frequency separation

A good crystal oscillator at 14.158 MHz was combined with a weak signal at 14.160 MHz. The level of the strong signal was set to -19.5 dBm while the level of the weak signal was set to -115 dB (in 50 ohms) at the connector which were to be connected to the Orion 2. At these levels the noise floor was dominated by reciprocal mixing. An analysis of the loudspeaker output showed that the desired signal was 31.5 dB/Hz above the noise floor (reciprocal mixing) which means that the level of reciprocal mixing at 2 kHz is -127 dBc/Hz, well in line with the sideband noise in transmit mode.

When two signals at 2 kHz separation with a signal level of -19.3 dBm were sent into the Orion 2, the appearent level of the IM3 product was -97 dBm. This corresponds to an IP3 of +19.5 dBm but since I forgot to measure the associated sensitivity, this number is not of much interest. An observation that IM3 grows a bit, about 6 dB, as the spacing between the signals is increased. The separation where the IM3 signal has its maximum depends on what filter is in use. Presumably the amplifier (or mixer) preceeding the roofing filter is improperly terminated for signals a bit outside the passband. This is a common design weakness. The signal source will be properly loaded for a signal within the crystal filter passband. Signals outside may see a much higher impedance and that would cause saturation since there is a limit to how high voltages can grow. An additional LC resonator can be used to invert the impedance and make the signal source see a low impedance outside the passband. Depending on the nature of the signal source, this is sometimes very favourable. A small attenuator or some resistive load is another option.

Reciprocal mixing in the IC-7000 at 144 MHz

A strong and a weak signal were combined and fed into an IC7000. This was not the same unit as the one used for transmitter tests. The weak signal was set to a level of -143.8 dBm, then the level of the strong signal required to degrade the noise floor by 3 dB was measured at different frequency separations. The strong signal was produced by a crystal oscillator.

The measurement was made with preamp off and the noise floor was at -160.5 dBm/Hz (NF=13.5 dB).

Freq.sep.    Level for 3 dB      Reciprocal
              more noise           mixing
  (kHz)         (dBm)             (dBc/Hz)
    1           -64.5                96 (spurs) 
    2           -59.5               101 (spurs)  
    5           -50.2               110.3
   10           -41.6               118.9
   20           -37.7               122.8  
   50           -35.6               124.9
  100           -26.6               133.9
These valuse are very similar to the sideband noise levels observed in transmit mode on another IC-7000.

To SM 5 BSZ Main Page