RankingsPrimerArbiterPhysicistMathematicianContrarianQEClaimsSources

Farnsworth-PT: Physicist review (Step 2)

Reviewer role: PhD physicist, mainstream view. Document reviewed: US Patent 2,172,152 "Radio Frequency Multipactor Amplifier" (Farnsworth; filed 21 Feb 1938, granted 5 Sep 1939). I read the full text layer myself [read: pdftotext, OCR noisy but legible]. No source gap for this document. The folder holds nothing else, so anything about Farnsworth's wider work rests on [memory] alone.

Corrections to Step 1

  1. Filing year: the drawing sheet header reads "Filed Feb. 21, 1958" in the OCR. Step 1 flagged this as OCR noise. Confirmed: the specification says "Application February 21, 1938, Serial No. 191,707", and Google Patents agrees (filed 1938-02-21, granted 1939-09-05) [web: https://patents.google.com/patent/US2172152A/en].
  2. Claim 3 in claims.md (linearity) is slightly over-stated. The patent says output is linear in "the electron supply at the beginning of each cycle, as stated with reference to the photosensitive cathode". The explicit statement is only for the photo-initiated case, "up to the point of space charge saturation". Its extension to the gun-modulated case is by analogy. [read]
  3. Claim 5 in claims.md ("needs no focusing fields") is an object of the invention in the patent, not something the description demonstrates. The description just never uses a focusing field. Step 1 half-flagged this. Treat it as an assertion.
  4. Claim 2 in claims.md bundles two things: the mechanism (electrons returned through the helix at the right phase) and the outcome ("greatly amplified"). The patent's own wording is "as is now well known in the art, a greatly amplified output current secured". So the patent itself defers to prior art for the gain. I rate the two parts separately below.
  5. Step 1 says "claims 1-4 cover the tube structure, 5-6 the external circuit". Correct, with one detail: claim 6 adds that the accelerating anode and cathode are "energized in opposite phase by a carrier current". Otherwise Step 1's facts match the text. Its notes on the garbled inventor name and the 1958 header are correct.
  6. Step 1's "[memory] multipactor ... attribution medium confidence" is now [web] checked: Wikipedia says Farnsworth identified and studied the effect in 1934 and tried to use it as an amplifier [web: https://en.wikipedia.org/wiki/Multipactor_effect].

Physics background (so the verdicts make sense)

  • Secondary emission: an electron hitting a surface with kinetic energy of tens to hundreds of eV knocks out on average delta secondary electrons. For oxidised silver or caesium-oxide surfaces delta can exceed 1 over a window of impact energies, roughly 50 to 1000 eV [memory: high confidence on the general shape; the specific peak values for Ag-O-Cs are medium and no verdict depends on them].
  • Multipactor (resonant multi-surface impact): in an RF field, an emitted electron is carried to the opposite surface; if its transit time is about an odd number of half-RF-periods, it arrives just when the field there again accelerates electrons out. If the average delta is at least 1, the population grows exponentially. [web: Wikipedia link above]
  • Growth stops when space charge (electron cloud screening the field) or detuning (electron energies drift off the delta > 1 window) limits it. So the current is not unbounded.

Claim-by-claim answers

Claim 1. A tube with a permeable secondary-emitting screen between a helical accelerating anode and a collecting anode can multiply an electron current when driven by RF.

a. Maths: The patent has no equations at all, so there is nothing to be inconsistent. The qualitative picture is self-consistent. Primaries accelerated by the helix hit the gauze, secondaries emerge. About 1/4 transmit through the gauze (the geometric open fraction), and the rest are either collected or pulled back, depending on RF phase. Energy balance: the extra output power comes from the DC supplies (sources 40 and 41), with the RF only gating the process. There is no energy-conservation problem. [read] b. Experiment: Consistent. Secondary-emission multipliers (photomultipliers, dynode chains) and multipactor discharges are textbook, measured effects. [memory: high; web: Wikipedia link above]. Gauze transparency of 1/4 is a geometric value, not an experimental claim. c. Tested: The patent reports no measurements, no gain, bandwidth or noise figures [read]. Farnsworth's earlier patents 2,071,515 and 2,071,516 are cited as prior art [read; web: Google Patents]. I found no independent replication of this specific tube geometry in what I checked (I did not do a full literature search; absence of evidence here is limited). Multipactor itself is well studied today, but as a nuisance in satellites, accelerators and waveguides [web: Wikipedia link]. d. If true, what would change: nothing. This is conventional vacuum-tube electronics. Verdict: Plausible, untested for a working amplifier in this exact geometry (no performance data on record in the folder). The underlying mechanism is Established.

Claim 2. Repeated trips (multipactor) at the right phase give a greatly amplified output current.

a. Maths: No quantitative model. Standard multipactor theory requires (i) delta greater than 1 and (ii) transit time equal to an odd multiple of half an RF period [web: Wikipedia link]. The patent says "by proper adjustment of the potentials" repeated trips occur, which is the same resonance condition stated informally, and consistent. One tension: exponential growth is inherently unstable (it saturates by space charge), so a tube operated in the multipactor regime is closer to an oscillator or a limiter than a stable linear amplifier, unless held well below saturation. The patent does not discuss stability. [memory: medium on how this was handled in practice; no verdict depends on it beyond the split below] b. Experiment: Multipactor resonance is experimentally established. "Greatly amplified" is not quantified. c. Tested: Mechanism: extensively. Farnsworth's device as an amplifier: no test results in the document. Wikipedia says only that he "attempted to take advantage of it as an amplifier" [web]. I found no evidence the tube became a commercial product; I did not establish whether it was built. Not dismissed by hostile tests that I could find; it was simply overtaken by conventional triodes and pentodes [memory: medium]. d. Nothing in mainstream physics would change. Verdict (split): mechanism Established; "greatly amplified" practical amplifier Never properly tested (no neutral test on record; not shown dismissed without testing).

Claim 3. Output current is linear in the electron supply per cycle, so a gun-control signal modulates the carrier.

a. Maths: In the growth regime, a linear multiplication process gives output proportional to the seed (N_out = M * N_in with M fixed by delta and number of trips), so linearity is natural below saturation. Above it the response compresses. The patent itself limits the linearity statement to "up to space charge saturation" for the light case. Consistent but unquantified. b. Experiment: Photomultiplier and secondary-emission multiplier linearity up to saturation is well documented [memory: high]. A subtlety: the multiplication M depends strongly on phase and amplitude of the RF, so amplitude-modulation of the carrier by seed current also couples to carrier-level noise. Not addressed. c. Tested: No data in patent; no replication found. d. No change to mainstream physics. Verdict: Plausible, untested.

Claim 4. Above ~10 MHz a seed of electrons (light or gun) is needed; photo-initiated output is linear in light up to space-charge saturation.

a. Maths: Reasonable qualitatively. Exponential buildup takes a roughly fixed number of RF cycles N, so the buildup time is N/f and falls as f rises. Relying on stray ionised particles (a stated starting mechanism) is too slow or too random at high f, so a deliberate seed helps. The specific 10 MHz figure is empirical ("it is found necessary") and stated without data. Not derivable from the patent. b. Experiment: Photoemission being linear in intensity (Einstein/Millikan photoelectric law, below saturation) is established [memory: high]. A caesium-oxide-on-silver surface is a known photocathode type [memory: medium]. c. Tested: Patent says it was "found", implying experiment, but gives no data. No independent test seen. d. No change. Verdict: Plausible, untested (the 10 MHz threshold specifically: Never properly tested in any record I have; it is an unsupported empirical statement).

Claim 5. No focusing fields are needed; a simple combined modulator/amplifier.

a. Maths: Concentric cylindrical geometry with an axial gun does not need magnetic focusing, but no beam-optics argument is given. Electron trajectories through a helix and gauze at RF are not obviously stable without transverse control, and the patent does not show they are. "Simple" is a design judgement. b. Experiment: Many tubes of the time worked without magnetic focusing, but this device's operation was not shown. c. Tested: No. d. No change. Verdict: Not testable as stated ("simple" is not measurable); the narrow technical part (works without a focusing field) is Never properly tested.

Overall assessment

This is a conventional, honest vacuum-tube patent. It makes no extraordinary claim, no energy-from-nothing claim, and no conflict with mainstream physics. The real weakness is evidential: a patent needs only a plausible disclosure, not data, so there is no performance measurement here. The multipactor effect is real and has since become mainly a failure mode to avoid in RF hardware [web]. As an amplifier it is a historically interesting dead end, in the "Historically superseded but instructive" sense, but I cannot show that from a source on this tube, so I keep the verdicts above and do not use that label.

Source tags summary