Patents and Disputes
Arguments still continue about whether De Forest really invented the triode vacuum tube. What is apparent is that he (and everybody else at the time) greatly underestimated the potential of his original device, imagining it to be limited to mostly military applications. It is significant that he apparently never saw its potential as a telephone repeater amplifier, even though crude electromechanical note magnifiers had been the bane of the telephone industry for at least two decades. The repeating amplifiers were trouble for several years beyond the advent of De Forest's Audion tube. It was only the loophole concerning vacuum that allowed amplifying triodes to be manufactured at all, since none of the original patents by De Forest specifically mentioned the use of a vacuum as part of the tube.
De Forest was granted a patent for his early two-electrode version of the Audion on November 13, 1906 (U.S. Patent 841,386), but the "triode" (three electrode) version was patented in 1908 (U.S. Patent 879,532). De Forest continued to claim that he developed the Audion independently from John Ambrose Fleming's earlier research on the thermionic valve (for which he received Great Britain patent 24850 and the American Fleming valve patent (U.S. Patent 803,684), and became embroiled in many radio-related patent disputes. De Forest was famous for saying that he "didn't know why it worked, it just did". He always referred to the vacuum triodes developed by other researchers as "Oscillaudions", although there is no evidence that he had any significant input to their development.
In 1914 Edwin Armstrong published an explanation of the Audion, and when the two later faced each other in a dispute over the regeneration patent, Armstrong was able to demonstrate conclusively that De Forest still had no idea how it worked.
The problem was that (possibly to distance his invention from the Fleming valve) De Forest's original patents specified that low-pressure gas inside the Audion was essential to its operation (Audion being a contraction of "Audio-Ion"), and in fact early Audions had severe reliability problems due to this gas being absorbed by the metal electrodes. The Audions sometimes worked extremely well; at other times they would barely work at all.
As well as De Forest himself, numerous researchers had tried to find ways to improve the reliability of the device by stabilizing the partial vacuum. One of these, Dr Irving Langmuir of General Electric, took a somewhat unorthodox approach: instead of trying to prevent the absorption of the gas, he deliberately started out with a higher vacuum and looked for ways of making the Audion work under those conditions. He succeeded, but quickly realized that, though superficially similar to the Audion, his "vacuum" tube was really a completely different device, capable of linear amplification and at much higher frequencies.
One of the major weakness of De Forest's claims is that true vacuum triodes simply will not work if there is any trace of gas left in the envelope. In fact, before vacuum tubes could become commercially viable, quite elaborate techniques had to be developed to both initially evacuate the tubes and soak up any gas molecules that subsequently found their way in. This flies directly in the face of his original patent specification, which specifically states that gas is essential to the operation of the Audion.
Another weakness is that none of his Audion schematics denoted the provision for any sort of "grid bias", an essential feature of any true vacuum triode operation.
Unlike the Audion, the vacuum triode could not demodulate radio signals directly (although Langmuir and other researchers soon found alternative ways to do this), but it was capable of linear (i.e. undistorted) amplification, which turned out to be a vastly more useful feature. It is ironic that many "faulty" Audions, which had lost their ability to demodulate radio signals due to gas absorption, had actually turned into crude linear amplifiers (which was why they lost their demodulating ability), but nobody realized this at the time.