 |
| OwenDuffy.net |
|
An SSB receiver is thought of as an essentially linear receiver, and as such can be used to measure RF input power.
The total audio power output of an SSB receiver is approximately proportional to the total equivalent RF power input over a certain range.
There are four key provisions in that statement:
Output power should be measured with an instrument that indicates the average power of a complex wave. Even a sinusoidal external source when combined with the receiver internal noise makes for a complex wave. A True RMS voltmeter can be used, so long as it has sufficient bandwidth and sufficiently long integration time.
Like any measurement system, there is error or uncertainty. Some receivers are less linear than others, and as a result measurement error is greater.
Total equivalent RF power input includes receiver internal noise and noise due to the external signal source.
The most significant limit on range with a typical receiver is AGC which automatically adjusts the total receiver gain to maintain an approximately constant audio output power. AGC is not usually active on the weakest signals, but is delayed until the input signal has reached sufficient level to have fairly good S/N ratio. AGC threshold on typical SSB telephony receivers is around 20-30dB above the noise floor. It is the change in AGC voltage (and consequently receiver gain) with changing RF input power that is the issue, not the existence of AGC voltage.
Trying to measure a signal much below the noise level becomes impractical due to the increase influence of individual measurement error on the calculated result.
The useful range can be increased by decreasing RF gain in those receivers where doing so injects an AGC voltage. This effectively increases the threshold of further AGC action.
Note that simply turning off the AGC as is so often suggested does not necessarily increase the linear range. Many if not most receivers go into non-linear overload not much above the AGC threshold. Those offering this solution are unlikely to have measured the outcome.
Lets look at a theoretical analysis of a receiver with practical parameters.
|
|
Fig 1 shows the audio output power of a receiver with NoiseFigure=6dB, 2kHz wide IF, NoiseFloor=-135dBm, and a sine wave RF source of Sin dBm.
The blue line is the Signal, and in an ideal (ie noiseless) receiver with no AGC, this would represent the output indication.
The green line is the ouput due to the sum of generator noise, receiver internal noise, and signal for a given Sin level. Note that at the low end, it departs from the blue line as a result of the noise contribution, and at the high end it departs from the blue line due to AGC action.
The red curve shows the error in considering total output power to indicate the level of Sin.
The range for less than 0.5dB error is about 15dB from -125dBm to -110dBm.
As mentioned, the useful range can be increased by decreasing RF gain in those receivers where doing so injects an AGC voltage.
 |
Fig 2 shows the same characteristics for a theoretical receiver where the RF Gain has been reduced till the S meter reads S9.
It can be seen from the green line that reducing the receiver gain has increased the total noise at the low end (due to increase in receiver noise floor), and the new AGC threshold is not S9 or -73dBm.
The range for less than 0.5dB error is about 35dB from -108dBm to -73dBm. The range obtained will depend on error criteria, the RF gain setting, and signal level available.
Clearly, ignoring internal and source noise degrades accuracy at the low end. Taking this noise into account can extend the range of 'linear' response.
 |
Fig 3 shows that power indication using FSM which accounts for the source and receiver noise. Measurements are made of a signal generator, 1dB steps from -127dBm.
 |
Fig 4 shows the same characteristic with the AGC disabled.
Turning the AGC off has not extended the linear range as the receiver goes into overload clipping. It might be a slightly softer knee than with AGC, but it is nonlinear nevertheless and useless for measurement above onset of clipping.
Fig 4 shows the futility of turning AGC off to extend linear receiver range on this receiver.
 |
Fig 5 shows the characteristic with the RF gain reduced to read S9 on the S meter. The useful measurement range is about 30dB.
 |
Fig 6 shows a similar test of an IC7000 with Preamp ON.
 |
Fig 7 show the same characteristic of the R5000 receiver. It is similar but has significantly higher error, an apparent consistent gain compression from low to high
| Version | Date | Description |
| 1.01 | 24/10/2012 | Initial. |
| 1.02 | ||
| 1.03 |
© Copyright: Owen Duffy 1995, 2021. All rights reserved. Disclaimer.