Receiver Test Tables: Stop Shopping by a Single Number

I read Sherwood’s table from left to right, not merely from top to bottom. The first question is never “Which radio is number one?” It is “Which measured limitation can actually become visible at my antenna connector, in my mode, with my gain state and my nearby signals?”

The Presentation Is the Missing Legend

Sherwood’s presentation belongs beside the table because it explains why the data exist and how easily the ordering is misread. The current table says it is sorted by narrow-spaced third-order dynamic range—or by reciprocal-mixing dynamic range when phase noise limits the result. Noise floor, AGC threshold, blocking, sensitivity, local-oscillator noise, selectivity and the other columns remain measurements for that receiver; they are not separately sorted into one combined score.

The presentation’s examples are useful because they separate measured receiver behaviour from preference. Sherwood discusses dynamic range, sensitivity, attenuation, preselection and ergonomics in one decision—but he does not turn those different quantities into one physical super-number. Joeri’s point is the same: use the table to test a requirement, not to award a trophy.

Freeze the Test Definition Before Comparing Radios

A number without its state is not portable. Before comparing two rows, record:

• band, mode, filter bandwidth and signal spacing;
• preamp, attenuator, dither, randomisation, IP+ or other gain/linearity state;
• AGC state and any roofing-filter or preselector selection;
• the criterion used for sensitivity, noise floor, blocking or distortion;
• the reference plane and impedance; and
• software, firmware and any sample-specific footnote.

Microvolts, dBm, dBc/Hz and decibels of dynamic range answer different questions. Converting a voltage sensitivity into power requires the stated RMS convention and impedance, but the result is still not comparable unless bandwidth and detection criterion match. A 500 Hz CW noise-floor result and a 10 dB signal-plus-noise-to-noise SSB sensitivity result are not two spellings of the same test.

Noise Floor and Sensitivity Ask Whether the Receiver Can Hear the System

Receiver noise matters when the noise delivered by the antenna, feedline and any external preamplifier is not sufficiently above the receiver’s own input-referred noise. It matters less when atmospheric, galactic or local man-made noise already dominates at the receiver input. That boundary is installation-, frequency-, bandwidth- and time-dependent; “low bands are noisy” is not a measurement of your station.

A practical check uses a fixed receiver state and bandwidth. Compare a safe matched termination with the installed antenna, then repeat with a calibrated input attenuator or known source. Record wanted-signal level, noise power and SNR rather than S-meter impressions alone. If added front-end attenuation reduces both signal and noise together while readability remains stable, receiver noise was not yet the dominant term. When noise and SNR begin to degrade, the receiver contribution has become material.

A passive Beverage, a small loop and an active array can all be excellent receive systems, but they present different gain, pattern, feed loss, common-mode exposure and total RF power to the receiver. “Negative antenna gain needs a preamp” and “active antennas need attenuation” are hypotheses to test, not universal rules.

DR3, RMDR and Blocking Describe Different Failure Paths

A superheterodyne receiver can block through gain compression before its narrow final filter. A direct-sampling receiver can run out of ADC range. The displayed symptom may look similar, but the table’s method and the receiver architecture decide what the column means. Sherwood therefore labels the direct-sampling entry in the blocking column as an ADC-overload point referenced to receiver noise floor rather than pretending the mechanisms are identical.

If a blocking figure is defined as a difference from a particular receiver noise floor, adding those two values gives a rough input level for that exact criterion. It is not a universal safe-input limit: change the preamp, attenuator, filter, bandwidth or method and both the reference and the margin may change.

Attenuation Is a Controlled Trade, Not a Low-Band Ritual

Input attenuation lowers wanted signals, external noise and unwanted blockers before later stages. It also raises the receiver’s effective noise figure by approximately its insertion loss. That trade can improve reception when overload products fall faster than the wanted signal or when excessive band noise is driving AGC, yet it can reduce SNR when receiver noise becomes important.

Use a sweep instead of a favourite setting:

1. Fix bandwidth, AGC mode, preselection and display averaging.
2. Record wanted signal, adjacent blockers, noise power, overload indication and intelligibility.
3. Add calibrated RF attenuation ahead of the stage at risk in several steps.
4. Stop when spurs or overload cease improving, or when noise/SNR begins to worsen.
5. Repeat on the bands, antennas and time periods that matter.

The procedure is meaningfully ordered, so the numbers describe test phases rather than decorate headings. RF gain and AGC controls are not interchangeable across receivers: one may command analogue gain before a vulnerable stage, another may move an AGC threshold, and another may be largely digital. The block diagram and manual decide what a control can protect.

Filtering Helps Only Before the Stage That Is Failing

A roofing filter can protect later gain stages in some superheterodyne architectures. It cannot protect an earlier mixer or amplifier that has already overloaded. Likewise, a digital filter after an ADC cannot restore clipped samples. External band-pass, high-pass, low-pass or notch filtering is useful when it rejects the actual unwanted energy before the first vulnerable nonlinear stage and its own insertion loss, impedance and power handling are acceptable.

Measure the local RF environment rather than assuming every difficult signal is inside the amateur band. Medium-wave, short-wave broadcast, another transmitter at the site, or a line-of-sight neighbour can dominate total input power even when the wanted channel looks quiet.

Translate the Table Into an Operating Requirement

Then add what the laboratory table intentionally does not score: user interface, ergonomics, filtering options, transmit cleanliness, remote-control requirements, power consumption, serviceability, cost and whether you enjoy using the radio. Those are not excuses to ignore measurement; they are additional requirements.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.