Key Points
- Most modern receivers are sensitive enough for HF, so sensitivity alone rarely separates a good radio from a great one.
- On the lower HF bands, atmospheric noise usually sets the floor, while on VHF and up the receiver’s own noise figure matters much more.
- Dynamic range tells you how well a receiver copes when very strong and very weak signals share the band.
- Phase noise in the local oscillator can bury weak signals close to a strong one, a problem called reciprocal mixing.
- Test numbers are only comparable when they were measured the same way, so read the test conditions as carefully as the results.
Spec sheets and test reports are full of numbers like -135 dBm, 95 dB and +25 dBm. Each one answers a simple question about how the radio behaves on a real, crowded band.
A quick refresher helps. A decibel (dB) compares two levels: 3 dB is double the power and 10 dB is ten times the power. dBm means decibels compared with one milliwatt, so 0 dBm is 1 mW and -100 dBm is a tiny 0.1 picowatt. More negative means weaker.
Sensitivity and MDS
Sensitivity asks how weak a signal the receiver can hear. The common lab measure is MDS, minimum discernible signal, also called the noise floor. It is the input level where the signal adds just as much power as the receiver’s own noise. The ARRL Lab’s test procedure measures it in CW mode with a 500 Hz filter, or the closest one the radio has.
Bandwidth matters a lot here. Noise power grows with bandwidth, so the same radio shows a noise floor about 7 dB higher through a 2.4 kHz SSB filter than through a 500 Hz CW filter. Never compare MDS numbers measured with different filters.
Older specs give sensitivity in microvolts at a stated signal-to-noise ratio. As a handy conversion, 1 microvolt across 50 ohms is about -107 dBm.
Noise Figure: HF vs VHF
Noise figure is how much noise the receiver adds compared with a perfect, noiseless receiver, stated in dB. A perfect receiver at room temperature still has a noise floor of -174 dBm in a 1 Hz bandwidth, which works out to about -147 dBm in 500 Hz. A receiver with a 10 dB noise figure would then show an MDS near -137 dBm.
On 160, 80 and 40 meters, the noise coming in from the antenna (distant storms plus your own neighborhood) is usually far stronger than the noise your receiver makes. That is why switching in an attenuator on the low bands often costs you nothing in readability.
On VHF, UHF and above, the band is much quieter, so the receiver’s own noise becomes the limit. This is where a low noise figure, good feed line and a preamp mounted at the antenna pay off, especially for weak-signal work, satellites and moonbounce.
Selectivity and Shape Factor
Selectivity is the receiver’s ability to hear one signal and ignore its neighbors. The first rule is to match the filter width to the mode: narrow for CW, wider for SSB. That keeps out interference and also keeps out extra noise.
Width alone does not tell the whole story. Shape factor compares the bandwidth far down the filter’s skirts with the bandwidth near the top, commonly the 60 dB width divided by the 6 dB width. A filter that is 2.4 kHz wide at 6 dB and 3.6 kHz wide at 60 dB has a shape factor of 1.5. Closer to 1 means steeper sides and less interference from the station next door.
You may also see ultimate rejection, the most a filter can knock down a signal far outside its passband.
Dynamic Range and Intercept Point
Dynamic range asks how strong a nearby signal can be before the receiver starts to misbehave. It is measured from the noise floor up, so a quieter receiver gets credit for its low floor too.
- Blocking Dynamic Range. The span from the noise floor to the level of a strong off-frequency signal that cuts a weak signal’s strength by 1 dB. Below that point, a big signal nearby does not make your weak one quieter.
- IMD Dynamic Range. Two strong signals can mix inside the receiver and create false signals. The test uses two equal signals and finds the level where the third-order product just reaches the noise floor.
Third-order products land close to the signals that made them, which is why they cause trouble. Strong signals on 14.020 and 14.022 MHz produce false ones at 14.018 and 14.024 MHz, right in the CW band. For every 1 dB the two signals rise, those products rise about 3 dB.
The third-order intercept point, IP3, is where those two lines would cross if you could keep raising the input. No receiver actually gets there; it is a projected number, and higher is better. A receiver with an IP3 of +20 dBm and a noise floor of -130 dBm works out to an IMD dynamic range of about 100 dB.
Direct-sampling SDRs behave differently. Instead of compressing gradually like an analog mixer, they hit a hard ceiling when the analog-to-digital converter runs out of range. For those radios, the overload point is the key number, and labs note when a figure was measured that way.
Phase Noise and Reciprocal Mixing
No oscillator is perfectly clean. Its energy spreads into a faint skirt of noise on either side, called phase noise, and it is stated in dBc/Hz at some distance from the carrier. Lower, meaning more negative, is better.
When a strong signal sits a few kHz away, it mixes with that noise skirt and spreads noise across your receive frequency. This is reciprocal mixing. It sounds like the band noise rising whenever a big station transmits, even though nothing is wrong with that station’s signal. The lab figure for it is RMDR, reciprocal mixing dynamic range.
Image Rejection
A superheterodyne receiver can hear two frequencies at once: the one you want and an image twice the IF away. For example, with a 9 MHz IF and the oscillator at 23.2 MHz, a station on 14.2 MHz and one on 32.2 MHz both come out at 9 MHz. Front-end filters must remove the image before the mixer.
Image rejection is how many dB stronger the image signal must be than the noise floor before you hear it. A high first IF puts the image farther away, which makes it easier to filter out.
How to Read a Test Report
Two well-known sources measure radios in a consistent way. The ARRL Lab tests radios for the product reviews in QST, and ARRL members can read the review archive online. Sherwood Engineering publishes a receiver test data table that ranks radios by close-spaced dynamic range, with the two test signals 2 kHz apart, or by RMDR when phase noise was the limit.
- Check the conditions first: filter width, preamp on or off, and the spacing between test signals.
- Look for notes that a figure was limited by phase noise or by the filter. Those numbers mean something different.
- Ignore small gaps. A difference of 2 or 3 dB is usually hard to notice on the air.
- Match the numbers to your operating. A contester near other strong stations needs close-spaced dynamic range; a casual operator on a quiet band needs much less.
Numbers are not the whole radio. Easy filter controls and good AGC matter every day too. For help choosing, see Which Radio, and Why.