Patent 4,412,299

Due to a recently issued patent I now am the sole inventor on three different patents. For various personal reasons I am not very public about the details of two of them. However, the first one, 4,412,299, does not share those concerns.

The title is Phase Jitter Detector. This was an invention to help meet speech immunity requirements for a Touch-Tone receiver. What is “speech immunity?” It turns out that because the old analog phone lines used the same channel for voice and signaling (the Touch-Tone digits) background speech, music, and other sounds would sometimes simulate a Touch-Tone digit with sufficient accuracy to insert a false digit into your dialed number. This false digit would result in a wrong number.

The problem increased in severity as the dynamic range of the receiver increased. With some reasonable precautions the problem was kept under control with normal subscriber to central office communications. But in the late 1970s and 1980s companies were creating things which they wanted to be able to remotely control. Examples from that time are answering machines to private branch exchanges (remember “dial nine for an outside line?”) where you could call the “main number” and then dial “One for Parts, two for Service, three for Sales, etc.. The receivers used for subscriber-to-subscriber communications required nearly twice the dynamic range of subscriber to central office communication. With this requirement the speech immunity problems went way up.

You could design a receiver which would would avoid the speech hits but in doing so you reduced the noise immunity of the receiver. A small amount of noise would prevent the detection of valid Touch-Tone digits. In other words, there is a trade-off between speech immunity and noise immunity. In the subscriber-to-subscriber signaling case you needed far better than normal noise immunity while retaining whatever speech immunity you could.

This speech immunity versus noise problem inspired me to get my master’s degree in electrical engineering. The graph below is from my thesis:


IIRC, the minimum A-level for subscriber to C.O. signaling was about –28 dBm. For subscriber-to-subscriber signaling the minimum A-level was about –40 dBm. Using a standard test tape from AT&T and a receiver without any guarding against speech hits the number of speech hits nearly tripled for this lower signal strength acceptance level.

My invention enabled the accurate distinction between a true Touch-Tone signal and speech/music/etc. which imitated a Touch-Tone digit. The receiver utilizing this invention, TelTone’s M-937 Touch-Tone Receiver, was able to meet any published requirements for subscriber-to-subscriber receivers for both noise and speech immunity. To the best of my knowledge no other production receiver was ever able to make that claim.


9 thoughts on “Patent 4,412,299

  1. Thank you for this. It feels like I know you better now in light of what you’ve accomplished. Engineers are a breed apart.

  2. I can’t imagine any of the gun control swine inventing anything, other than pronouns and excuses for their visits to Epstein’s Island.

  3. Neat! It’s vaguely reminiscent of the LPC adaptive noise filters that are now common in ham radio receivers (or their reverse, LPC adaptive notch filters to block interfering tones).
    Not sure why you would be quiet about the other two, since like most other patents they are public documents and readily show up in a search. Nice work there as well.

  4. The funny thing is that since I work in litigation (and often patent litigation) everything I know about SS7 (and cell phones too) comes from patent cases. I know a little more about cell phones since becoming a ham (and now understand some things I THOUGHT I knew before, like how a folded dipole works.)

    This also reminds me of the phreaker Captain Crunch, who could whistle his dialtones and didn’t need any sort of tone generator box. He started out trying to figure out why 2600 would causes his calls to hang up.

    • One of my college friends was a semi-reformed phone phreak. He was well enough known to Ma Bell that they sent some people over for a chat when he moved into a campus apartment and asked for a phone. He agreed to behave himself and asked for a nice phone number in exchange, so they gave him xxx-2000. He came to regret that since dial-a-prayer was xxx-2100 and the Tender Touch massage parlor was xxx-2900, so he got plenty of wrong number calls. 🙂

      I suppose you also know that the internal signaling tones were different from the touch tone frequencies, using round numbers rather than funny random looking ones. Those internal codes were what you’d use after sending 2600. And no, I never did this, but I heard the various stories and remember some of the details.

  5. Sweet! Looking forward to hearing about the other two some day! (Although your secrets are save with me, as I’m at the age I would forget in ten minutes.)

Comments are closed.