Value chain · CLM-0901
In plain terms
Each antenna must hold its position along the line to the receiver within about three millimetres for a tenth of a decibel. The whole 1970s design budget, converted to the same units, was 3.4 millimetres. One metre of structural bending is nearly three thousand degrees of phase, and if that is not corrected the antennas stop adding up at all - a billion of them lose ninety decibels, which is everything.
The claim as recorded
For a transmit array at 2.45 GHz, the Ruze relation G/G0 = exp(-sigma^2) gives an rms line-of-sight element position tolerance of 2.955 mm for 0.1 dB of peak gain loss, 6.608 mm for 0.5 dB, 9.345 mm for 1 dB and 16.186 mm for 3 dB, scattering 2.3, 10.9, 20.6 and 49.9 per cent of radiated power out of the main beam respectively. The relation is confirmed against the reference design the register already holds: at the 10 degrees RMS overall phase budget CLM-0160 records from NASA RP-1076, Ruze gives 3.0002 per cent main-beam power loss against RP-1076's stated "under 3 per cent", reproducing to four significant figures with nothing fitted. Expressed as path error, the entire 1978-81 system phase budget is 3.399 mm rms at 2.45 GHz. Unmanaged, one metre of line-of-sight structural flex is 8.17 wavelengths or 2,942 degrees of phase, which decorrelates rather than degrades: a decorrelated array sums incoherently, N^2 falling to N, which at a billion elements is 90 dB.
Evidence behind this: T0-calculation: the stated values are arithmetic performed here from the Ruze tolerance relation. CALC-100 supplied as required. The cross-check against RP-1076 is the independent confirmation.
Scope and limits
Ruze treats the phase errors as random and uncorrelated. A correlated deformation - a slow structural bend - MISPOINTS the beam more than it defocuses it, and costs less peak gain than an uncorrelated error of the same rms. So these figures are the PESSIMISTIC bound for a correlated deformation and the correct model for an uncorrelated one. The helix geometry, mutual coupling and the real structure's modal shapes are not modelled. The per-cent column is power leaving the main beam, and says nothing about where it lands.
What it assumes
The Ruze relation for random aperture phase errors applies to the transmit case.
Where it comes from
This is the project's own working, shown in full:
A held script, sha256 3b96b1485af2def6, 5,335 bytes. Standard library only, no dependencies, runs on any Python 3. CALC-100 carries the arithmetic, identified 2026-09-14 by the script name and the sha256 above appearing in that calc as well - not by inference. The 10 degree cross-check value is computed by the same script.
Basis: mm (rms line-of-sight position error); dB (peak gain); per cent (power scattered out of the main beam)
Related records
- Answers question
- QST-013 not published in this release
Cite this
energyaperture.com/c/clm-0901
CLM-0901, Energy Aperture, as of 2026-09-13.
Connected records
Other published records linked to this one. A list, not a judgement.
- Also from SRC-107
- CLM-0033 (low confidence)Apart from a 1981 NASA review that found no evidence for the number and suggested a higher one, nobody was found to have revisited it, and...CLM-0060In 1981 a NASA review panel said the 230 watts per square metre figure rested on an obsolete theory with no experimental support, and made...CLM-0061The experiments were done. They found heating two or three times stronger than normal, not the tenfold effect predicted, no irregularities...CLM-0062By 1981 the theory was being revised to a steeper frequency dependence, which would raise the permitted power density.CLM-0063NASA suggested in 1981 that 550 to 600 watts per square metre might be reasonable, roughly two and a half times the figure still in use...CLM-0101The 1978 contract required the researchers to report fixes if ionospheric disturbance pushed the pilot-beam phase error past 2 to 3...CLM-0160A 1978-81 NASA design study calculated that simply pointing the satellite's structure accurately enough, without any electronic correction...CLM-0420NASA's 1981 report says the experiments found nothing supporting the 23 milliwatt limit, that the measured heating was two or three times...More on the source page.