Using a Signal Generator to Test a Receiver
A signal generator transforms receiver testing from an art based on listening and experience to a precise science based on measured numbers. With a calibrated RF signal generator, you can determine a receiver's minimum discernible signal (MDS), noise figure, selectivity at various offsets, image rejection, and spurious response rejection — all expressed as dB or dBm values that can be compared to published specifications or to measurements taken before and after a repair or modification.
This is not an abstract laboratory exercise. When a receiver is performing poorly — missing weak DX signals, being desensitized by nearby signals, or hearing ghost signals on frequencies that should be quiet — the signal generator measurements described in this lesson will tell you exactly what the receiver is doing and by how much it deviates from expected performance. This is the difference between troubleshooting by guesswork and troubleshooting by measurement.
Test Setup
The test setup for receiver testing requires care. RF leakage between the signal generator and the receiver — through the air rather than through the cable — can cause the receiver to respond even with the generator output disconnected, completely invalidating the measurement. Requirements:
- Shielded connection: Connect the signal generator to the receiver's antenna input through a 50 Ω coaxial cable with good-quality connectors. Use as short a cable as practical.
- 50 Ω termination: The receiver's antenna input should see 50 Ω from the signal generator. Most RF signal generators have a 50 Ω output impedance. Do not use an open-circuit connection or adapter that changes the impedance.
- External attenuator: For measurements below −90 dBm, insert a precision 10 dB or 20 dB attenuator between the generator and the receiver. This pads down the residual generator output leakage and improves the accuracy of very low level measurements. Add the attenuator value to the generator's displayed output level setting.
- Antenna disconnected: Disconnect the receiver's antenna and connect the signal generator in its place. Any ambient signals on the antenna will corrupt the measurement.
- AGC consideration: Many receivers have an automatic gain control (AGC) that compresses the audio output level as input level increases. For sensitivity measurements, either disable the AGC or account for its effect. The most reliable method is to use a signal level just at the threshold of discernibility and not rely on audio level readings.
The MDS test setup. The signal generator feeds the receiver input through a precision step attenuator; the receiver's antenna port must be disconnected to prevent ambient signals from corrupting the measurement. The audio voltmeter measures the noise floor with the generator off, then the generator level is raised until the audio output is exactly 3 dB (1.414×) above the noise floor — that generator level is the MDS.
View LargerMinimum Discernible Signal (MDS)
The minimum discernible signal is the smallest signal that the receiver can detect with any reliability. It is defined as the input signal level that raises the receiver's audio output by 3 dB above the audio noise floor (the noise output with no input signal). This "3 dB above noise" criterion is equivalent to a 0 dB SNR — the signal power equals the noise power in the measurement bandwidth.
MDS measurement procedure:
- Set the receiver to the test frequency, USB mode, 2.4 kHz bandwidth (or as close to standard conditions as possible), full RF gain, no preamp, no attenuator
- Tune the BFO or passband to center the signal generator's tone in the audio passband
- Connect an audio voltmeter or the receiver's S-meter to quantify the audio output
- With the signal generator set to zero output or disconnected, note the audio noise floor level (N)
- Increase the generator output level until the audio reads 3 dB above N (voltage increases by a factor of 1.414, or power doubles)
- The generator output level at this point is the MDS
Test conditions: 14.2 MHz, USB mode, 2.4 kHz roofing filter, RF gain maximum, no preamp, no ATT.
Audio noise floor with generator disconnected: 2.8 mV RMS at audio output (call this N).
3 dB above noise requires audio output = N × √2 = 2.8 × 1.414 = 3.96 mV RMS.
Generator level adjusted until audio output reads 3.96 mV. Generator level at this point: −135 dBm.
MDS = −135 dBm. This is an excellent result — the TS-590SG achieves −135 to −136 dBm MDS in typical independent reviews, confirming this is a properly functioning receiver.
Noise Figure from MDS
The noise figure (NF) is a fundamental measure of receiver sensitivity, independent of bandwidth. It tells you how much the receiver degrades the SNR relative to a theoretically perfect noiseless amplifier. Once you know the MDS, noise figure is calculated as:
NF (dB) = MDS (dBm) − kTB (dBm)
Where kTB is the thermal noise power in the measurement bandwidth at room temperature (290 K):
kTB (dBm) = −174 dBm/Hz + 10·log₁₀(BW in Hz)
MDS = −135 dBm, measured with 2.4 kHz (2400 Hz) SSB bandwidth.
kTB = −174 + 10·log₁₀(2400) = −174 + 33.8 = −140.2 dBm
NF = MDS − kTB = −135 − (−140.2) = +5.2 dB
A 5.2 dB noise figure is excellent for an HF receiver. The theoretical minimum (for a 290 K source impedance) is 0 dB. Most high-quality HF transceivers achieve 8–12 dB noise figure, so this example is particularly good. A receiver with 10 dB NF would show MDS ≈ −130 dBm in the same bandwidth.
Sensitivity in µV
The ARRL and most receiver specifications quote sensitivity not as MDS (0 dB SNR) but as the signal required for 10 dB SNR. This is a more practical criterion because a 0 dB SNR signal is barely detectable, while a 10 dB SNR signal is actually copy-able. The test procedure is the same as MDS, except you raise the generator level until the audio output is 10 dB above the noise floor (3.16× in voltage).
The result is then converted to µV EMF (open-circuit voltage from a 50 Ω source). This conversion: V_emf (µV) = 10^((P_dBm + 107) / 20).
Adjacent-Channel Selectivity
Selectivity describes how well the receiver rejects signals on frequencies other than the desired frequency. The test procedure:
- Apply a signal at the desired frequency and set the generator to a level that produces a comfortable audio output (e.g., 20 dB above the noise floor)
- Without changing the receiver, move the generator to an offset frequency (e.g., 3 kHz, 6 kHz, 10 kHz offset for HF SSB)
- Increase the generator level until the audio output is back to the same level as in step 1
- The difference between the level in step 3 and the on-channel level in step 1 is the selectivity at that offset, in dB
On-channel generator level for reference audio: −100 dBm at 14.200 MHz.
Generator moved to 14.210 MHz (10 kHz offset). Level increased to −55 dBm to restore same audio.
Selectivity at 10 kHz offset = −55 − (−100) = 45 dB.
This receiver rejects signals 10 kHz away by 45 dB. A good performance figure for 10 kHz offset with a 2.4 kHz bandwidth filter.
Image Rejection
A superheterodyne receiver converts the received RF signal to an intermediate frequency (IF) by mixing with the local oscillator. An unwanted signal at the image frequency — located at RF ± 2×IF from the desired frequency — also converts to the same IF and appears in the receiver's passband. Image rejection is how well the receiver suppresses these image-frequency signals.
For a receiver with a 455 kHz IF operating on 14.200 MHz with a high-side LO: LO = 14.200 + 0.455 = 14.655 MHz. Image frequency = LO + 0.455 = 14.655 + 0.455 = 15.110 MHz.
Image rejection test: set the generator to the image frequency (15.110 MHz in this example), increase the level until the receiver responds at the same audio level as an on-channel signal. The difference in level is the image rejection in dB. A well-designed HF receiver achieves 60–80 dB image rejection from pre-mixer filtering and double- or triple-conversion architectures.
Receiver Sensitivity Calculator
This calculator computes the theoretical minimum discernible signal (MDS) for an ideal noiseless receiver (0 dB NF), the MDS for a receiver with a specified noise figure, and the corresponding sensitivity in µV (for 10 dB SNR). Comparing the calculated ideal MDS to your measured MDS tells you how far the receiver is from the theoretical limit.
Receiver MDS and Noise Figure Calculator
Enter the receiver bandwidth and noise figure to calculate MDS (0 dB SNR) and sensitivity (10 dB SNR). Or enter a measured MDS to back-calculate noise figure.
Frequently Asked Questions
How do I perform a sensitivity test without a calibrated RF signal generator?
You can make a relative sensitivity test using an uncalibrated signal source plus a precision step attenuator. Set the signal source to any convenient level that produces readable output in the receiver. Then reduce the attenuator in 10 dB steps while observing the audio output. The number of 10 dB steps required to reduce the audio output to 3 dB above the noise floor, multiplied by 10, gives the receiver's sensitivity relative to your starting level. This relative measurement is useful for comparing before/after receiver modifications or verifying that the RF gain, preamp, and attenuator controls are working correctly.
Why does the receiver seem more sensitive with the preamp on, but the noise figure is worse?
The preamp increases gain (making weak signals louder and therefore more audible), but it also adds noise of its own. In most HF environments, the external noise from the antenna (from atmospheric noise, man-made interference, and the cosmic background) exceeds the receiver's internal noise. In this case, adding preamp gain does help you hear weak signals because the external noise-to-signal ratio of the antenna is what limits reception, not the receiver's internal noise. But if you measure noise figure (which accounts only for receiver-generated noise), the preamp's own noise contribution increases the total NF. Use the preamp on HF when external noise is high (most real-world conditions); disable it on VHF/UHF where external noise is lower and receiver NF dominates.
Test Your Knowledge
Answer the questions below to check your understanding. Every answer can be found in the lesson above.