Value chain · CLM-1043
In plain terms
With the heater off and a tuning circuit on the cathode, one hand-picked oven magnetron was as quiet as the instrument could measure: about 197 decibels below its carrier per hertz of bandwidth, 50 megahertz away from the carrier. Detuning the cathode circuit cost 20 decibels; turning the heater on swamped the display.
The claim as recorded
Raytheon's February 1981 assessment reports one microwave-oven magnetron, Raytheon QKH2000 #13 run as a directional amplifier at 20 dB gain with its heater off and a coaxial short-circuit on the cathode support tuned for minimum noise, measured through a 66 dB carrier notch in a 300 kHz bandwidth at the edges of the 2.4 to 2.5 GHz ISM band as "Signal to Peak Noise = 133 dB/300 KHz, Signal to RMS Noise = 142 dB/300 KHz, Spectral Noise Density -196.8 dBc" (142 + 54.8 dB to 1 Hz; 136.8 dB/MHz), a reading "about 2 dB above the noise level seen on the scope without any input signal", with the cathode short at an unfavourable setting about 20 dB worse and with the normal 37 W of heater power "much higher ... the noise would almost fill the screen", no number given.
Evidence behind this: T3-tube developer's report; one selected tube at the instrument's sensitivity floor, measured.
Scope and limits
One tube, selected from fifteen, with a laboratory cathode circuit the SPS design does not include as such; the figure is at the sensitivity floor of the equipment ("approached the limit of sensitivity", 6-2) and is a floor, not a characterisation. It is at +/-50 MHz; nothing is given between 10 and 50 MHz. The source's own projection for the SPS tube is -120 dBc/MHz at 30 MHz (CLM-1046), 17 dB worse.
What it assumes
The RMS-to-peak correction of 9 dB and the 54.8 dB bandwidth conversion the source applies.
Where it comes from
PDF p.120 / 6-14, Fig. 6-7c caption, render read: "Receiver Bandwidth = 300 KHz Total Frequency Scan = 100 MHz Vertical Scale = 10 dB/div. Carrier Suppression = 66 dB Signal to Peak Noise = 133 dB/300 KHz Signal to RMS Noise = 142 dB/300 KHz Spectral Noise Density -196.8 dBc". PDF p.121 / 6-15: "When the level of noise is transferred from a 300 KHz bandwidth to a 1 Hz bandwidth an additional 54.8 dB is added, so that the spectral power density of the noise is 142 + 54.8 or 196.8 dB below the carrier ... There is a difference of about 20 dB in noise level between the two settings ... It is seen that the noise level is much higher even though the gain of the noise is reduced by 24 dB because only 42 dB of carrier suppression was used." "With 66 dB of carrier suppression the noise would almost fill the screen." PDF p.119 / 6-13: "the noise level at the edges of the frequency sweep (which corresponds to the 100 MHz width of the ISM band) is about 2 dB above the noise level seen on the scope without any input signal." PDF p.112 / 6-6: "A total of 11 Raytheon, two Hitachi, and two Toshiba magnetrons were examined as candidates".
Basis: dB below carrier per Hz; 300 kHz measurement bandwidth
Related records
- Answers question
- QST-017 not published in this release
Cite this
energyaperture.com/c/clm-1043
CLM-1043, Energy Aperture, as of 1981-02.
Connected records
Other published records linked to this one. A list, not a judgement.
- Also from SRC-560
- CLM-1045Raytheon's own numbers had the klystron just failing the interference limit at the band edge and the magnetron clearing it by 45 decibels...CLM-1046The number Raytheon wrote into the design table for the satellite tube was about fifty times noisier than its best laboratory reading...CLM-1047The oven magnetron's second harmonic came out about 70 decibels below its carrier, a tuned stub took another 40 off, and the antenna's own...CLM-1048A cheap oven magnetron run cold and hard converted 81.7 per cent of its DC input to microwaves, measured carefully by heating water...CLM-1049The fifty-year magnetron rests on a cathode that runs cool because the tube heats it only as much as it needs, on tube-life curves from...CLM-1050The interference limit Raytheon measured its tubes against is given as a CCIR requirement but never traced to a CCIR document, and one page...