RankingsPrimerArbiterPhysicistMathematicianContrarianQEClaimsSources

Farnsworth-PT: Mathematician review (PhD mathematician role)

Source: US Patent 2,172,152 (Farnsworth, filed 21 Feb 1938, granted 5 Sept 1939), text layer re-read by me in full [read]. No source gap for this document. The folder holds nothing else; claims about Farnsworth's other work rest on [memory] or [web] only.

Key fact for all claims: the patent contains NO equations, no numbers except "about ten megacycles" and "one to four", and no data [read]. "Mathematical consistency" therefore means: are the verbal statements, the circuit description and the claims mutually consistent, and do they obey simple dynamics I can compute myself.

Corrections to Step 1

  1. Claim 3/4 wording. The patent says photo-initiated current is, "in its maximum value, a linear function of the illumination intensity up to the point of space charge saturation" [read]. Step 1 drops "in its maximum value". It then says the output current is "linearly proportional to the electron supply at the beginning of each cycle" "as stated with reference to the photosensitive cathode" [read]. So the linearity statement is made only for the photo case and is extended by reference to the gun; it is not separately derived.
  2. Claim 2 ("greatly amplified output current"): the patent does not demonstrate this. It says "as is now well known in the art, a greatly amplified output current secured" [read], i.e. it appeals to prior art (Farnsworth's earlier patents). Step 1 presents it as the patent's own result.
  3. The "one in four" figure: the patent says the fraction of secondaries passing through is "very nearly that of the apertured area to the total area, or one to four" [read]. Step 1 key_claims lists this correctly but the summary treats it as fact; it is a geometric-transparency assumption (see claim 2 below).
  4. Minor: "no focusing fields" appears only in the list of objects, not in the operating description [read]. Step 1 claim 5 is fine but it is an aim, not a result. Otherwise facts.json (claim structure 1-4 tube, 5-6 circuit, filing dates, OCR errors) checked against the text and is accurate.

Central claims

Claim 1: tube with permeable secondarily-emissive screen cathode between helical accelerating anode and collecting anode multiplies electron current under RF drive

a. Math consistent? Yes, as a description. Geometry (gun on axis, helix inside screen inside cylindrical collector) and circuit (coil 36 across screen and, via blocking condenser 39, the helix; DC +V40 on helix, +V41 on collector via tuned coil 42) agree with claims 5-6 ("energized in opposite phase" in claim 6 matches coil 36 feeding screen and helix from its two ends) [read]. Hidden assumption: secondary yield delta above 1 at the impact energy actually reached. I computed the impact energy for a parallel-plate analogue (first-order resonance, electron leaving at rest at RF phase 0.3 rad, transit = half period): 100 MHz, 1 cm gap needs about 62 V peak and gives about 32 eV impact; 10 MHz, 1 cm gives 0.6 V and 0.3 eV [my calculation, numpy]. At such energies silver oxide has delta below 1 (typical delta peaks near 300 eV; air-exposed silver max about 2.6 at 300 eV [web] https://www.sciencedirect.com/science/article/pii/S2211379722000407). So the DC bias on the helix, not the RF, must supply the impact energy; the patent states no voltages, so consistency cannot be checked quantitatively. b. Experiment? Secondary emission and electron multiplication (dynodes, photomultipliers) are established. Silver/Cs-O surfaces have high yields (Cs oxide about 10 for low-energy electrons [web] https://www.sciencedirect.com/topics/engineering/secondary-electrons). c. Tested? Patent presents no data. Farnsworth held a series of multipactor patents (e.g. 1,969,399 filed 1930; 2,071,517 "Multipactor phase control"; 2,091,439 "Multipactor oscillator and amplifier") [web] https://patents.google.com/patent/US2091439A/en. I found no source here for independent replication or for commercial use of this tube; unknown. d. If true: nothing changes. Ordinary physics. Verdict: Plausible, untested (as an amplifier tube; the underlying multiplication mechanism is Established).

Claim 2: repeated-trip (multipactor) action returns electrons through helix to strike screen again, giving greatly amplified current

a. Partly consistent, with a gap. Mechanism: electrons emitted when the helix is most positive return through it (phase-dependent) [read]. This is the standard multipactor resonance and my simulation confirms half-period transit is achievable at modest voltages. But two issues. (i) Transparency: "fraction passing through = open-area ratio = 1/4" applies to a collimated primary beam hitting a mesh. Secondaries leave the wire surface in a broad angular distribution, so transmission of secondaries is not set by open area. The "one to four" is a rough geometric guess presented as exact ("very nearly"). (ii) Loop-gain logic: if the effective yield per round trip exceeds 1, the count grows geometrically (delta^N; delta=3 gives 59,049 after 10 impacts [my calculation]) until space charge saturates, so the tube behaves as an oscillator (this is the subject of Farnsworth's separate "Oscillation generator" patent 2,071,516 [read, cited]). A linear amplifier needs the round-trip gain below 1 (regenerative, below threshold) or single pass. The patent does not say which regime it operates in, nor how stability is kept. b. Multipactor resonance is well documented in RF cavities, waveguides and satellite hardware (as a nuisance) [memory, high]. c. No test data in patent. Unknown for this device. d. Nothing. Verdict: Plausible, untested. The gain statement is unquantified; the 1/4 rule is an approximation, not a derivation.

Claim 3: output current linear in electron supply per cycle, so signal on gun control electrode modulates the carrier

a. Consistent in the small-signal regime. If each cycle's emitted charge is a fixed factor M times the injected charge, output is linear in injected charge. But the patent states linearity only "up to space charge saturation" and at saturation the output is independent of the seed. Multipactor is a nonlinear phase-focusing process; M depends on RF amplitude and bias, so amplitude-modulation distortion is expected near threshold. No derivation or measured linearity. Also the gun's DC beam, modulated by the control electrode, gives baseband current modulation; the carrier amplitude follows only if multiplication is bunched at the carrier frequency, plausible but not shown. b. Linear response of electron multipliers with beam current is established for photomultipliers (below saturation) [memory, high]. Not shown for multipactor. c. No tests in the document. d. Nothing. Verdict: Plausible, untested (linear only in the unsaturated small-signal regime, which is also the stated limit).

Claim 4: above about 10 MHz a starting electron source is needed; photo-initiated output linear in light up to space-charge saturation

a. Consistent with exponential build-up: starting from n0 electrons, N trips are needed to reach saturation (about log(n_sat/n0)/log(delta)); at 10 MHz one trip per half-cycle is 50 ns, so a build-up of tens of trips takes microseconds, which is longer than a carrier burst if n0 is very small. So the stated need for seed electrons is mathematically sensible. The "10 MHz" threshold is empirical ("it is found necessary") with no data [read]. Photocurrent proportional to light (below saturation) is standard. b. Consistent with photoemission physics (linear in intensity) [memory, high]. c. No tests recorded in the document. d. Nothing. Verdict: Plausible, untested.

Claim 5: no focusing fields needed; simple combined amplifier and modulator

a. This is an objective, not demonstrated, and the specification never explains how the beam through the helix reaches the screen without focusing beyond "small area presented by the anode, the greater number pass" [read]. Internal consistency is not violated but the claim is unsupported; "no focusing" is also in mild tension with the need for a ring control electrode and gun anode, which are electron optics. b/c. No data on gain, bandwidth or noise (secondary-emission shot noise is known to be significant [memory, medium]; not checked). d. Nothing. Verdict: Not testable as stated ("simple", "no focusing" unquantified); the specific hardware is Never properly tested (no neutral test on record that I found; not shown dismissed).

Numeric summary

  • Parallel-plate half-period resonance, my model (numpy, phase 0.3 rad): 100 MHz, 1 cm: 62.5 V peak, 32 eV impact; 100 MHz, 0.5 cm: 15.6 V, 7.9 eV; 10 MHz, 1 cm: 0.6 V, 0.3 eV. Impact energies are far below the 300 eV where silver oxide yield peaks, so the helix DC bias must be large, which the patent does not state.
  • Multiplication delta^N: delta 2, 3, 5 over 10 impacts gives 1.0e3, 5.9e4, 9.8e6.

Web checks (budget used: 2 searches)