Also known as: Te, equivalent noise temperature, effective noise temperature
Noise temperature (T_e, or equivalent noise temperature) restates a component’s self-generated noise as the physical temperature a resistor would need to produce the same thermal noise power.1 Instead of saying “this amplifier has a noise figure of 0.5 dB,” you say “it adds noise equivalent to a 35 K source.” The two descriptions carry identical information — T_e = 290·(F − 1) — but noise temperature resolves the tiny noise contributions of very-low-noise systems far more clearly than fractions of a decibel, so it is the preferred metric for satellite ground stations, deep-space links, and radio astronomy.
How it works
Rewrite kTB as k·T_e·B and the meaning is direct: a device with equivalent noise temperature T_e adds the same noise a matched resistor at T_e kelvin would. The conversion to and from noise figure uses the 290 K reference:
T_e = 290·(F − 1) and F = 1 + T_e/290
So NF = 3 dB (F = 2) is T_e = 290 K; NF = 0.5 dB is about 35 K; a cryogenically cooled amplifier at 15 K corresponds to NF ≈ 0.22 dB — a difference the decibel scale barely shows but the temperature scale makes obvious.
The other advantage is linear cascading. Where noise figures combine through the Friis ratio formula, noise temperatures simply divide-and-add:
T_sys = T₁ + T₂/G₁ + T₃/(G₁G₂) + …
and at the antenna the total is just T_sys = T_A + T_e, the antenna’s own temperature (how warm the scene it looks at is) plus the receiver’s equivalent temperature. A dish aimed at cold sky sees a low T_A; aimed near the warm horizon or the ground it sees a high one.
In practice: G/T
For a receive station the headline figure of merit is G/T — antenna gain divided by system noise temperature, quoted in dB/K. It rolls the whole receive chain into one number: raise antenna gain or lower system temperature and G/T improves, directly improving the carrier-to-noise the station can achieve on a given downlink. This is why satellite operators specify earth-station performance as G/T rather than gain or noise figure alone, and why cooling the LNA (lowering T_e) is worth the cost on faint downlinks.
Relevance to SDR
Most terrestrial SDR work — scanning P25, DMR, TETRA, pagers, ADS-B — is limited by man-made noise and antenna temperature far above the receiver’s own T_e, so noise figure is the more convenient everyday metric. Noise temperature comes into its own in the weak-signal corners of the hobby: L-band and higher satellite reception, NOAA/GOES and Inmarsat downlinks, radio astronomy (hydrogen-line, pulsars), and EME/moonbounce, where every kelvin of system temperature costs link margin and cooled or very-low-T_e preamps earn their keep. Framing the front end as “35 K of added noise” rather than “0.5 dB” makes the trade-offs between LNA choice, feedline loss, and antenna pointing legible.
GopherTrunk does not compute or use noise temperature — it is a decode engine, not a station-design tool — but the concept explains why a GopherTrunk user chasing a faint signal should think about the whole receive chain’s temperature budget (cold-sky pointing, mast-mounted low-T_e LNA, minimal feedline before it) rather than expecting software to close a link the RF front end never had margin for.
Sources
-
Noise temperature — Wikipedia, equivalent noise temperature, cascade addition, and the G/T figure of merit. ↩