Shop OBEX P1 Docs P2 Docs Learn Events
Propeller Chip Frequency — Parallax Forums

Propeller Chip Frequency

HumanoidoHumanoido Posts: 5,770
edited 2011-06-30 08:46 in Propeller 1
For a single chip, running various programs, what can be said about determining the overall combined frequency radiated by the chip (in terms of RF, i.e. received by a radio in close proximity)?

Is there a mathematical relationship, frequency vs running code?

I want to read "combined" RF frequency from a single chip (without touching the chip) and generalize what the chip is doing internally.

Thanks sincerely.

Comments

  • Mark_TMark_T Posts: 1,981
    edited 2011-06-29 22:55
    You'll get broadband RF noise with strong spikes at harmonics of the various clock signals. The noise will correlate with code activity.

    The wiring to the chip will do most of the radiating, not the chip itself.
  • HumanoidoHumanoido Posts: 5,770
    edited 2011-06-30 07:47
    It looks like I will need to create the specific code activity and then make the RF measurements and finally draw the correlations. Any other ideas?
  • K2K2 Posts: 693
    edited 2011-06-30 08:36
    A former boss of mine once placed an AM/FM radio on top of some old boat-anchor CPU he was using. He tuned the radio till he picked up interference from the computer and then left the radio tuned to that 'station'. Over time, he and his associates became very adept at detecting and troubleshooting CPU problems based solely on the audio output of the radio.

    It's worth noting that Morse Code developed in a similar fashion. The original equipment was designed to mark a paper tape that would subsequently be read. With time, operators found they could pick up the content simply by listening to the noise that the tape marking machine made. Eventually they dispensed with the paper.
  • mindrobotsmindrobots Posts: 6,506
    edited 2011-06-30 08:41
    I would think it the same as any type of electro-magnetic profiling. You need to create the specific situation and capture that into a library. Once you have a library of profiles built, you could take an unknown signature and compare it against the library.

    It's lke SONAR and submarines. Theoretically, you could derive the signature of a submarine from the design and massive calculations but until you have heard a specific sub and captured it's signature you can only estimate and assume. You may be able to determine if a signature is from a certain class of boat but you wouldn't know WHICH specific boat unless you had that boat in yoru library.

    Same thing for chips, you may be able to differentiate between a BS2, a Propeller and an Atmel but overal RF radiation but probably not much more than that.
  • localrogerlocalroger Posts: 3,452
    edited 2011-06-30 08:43
    My own first computer had a 4 MHz 8080A and on several occasions I used an AM radio to debug interrupt code.
  • LeonLeon Posts: 7,620
    edited 2011-06-30 08:46
    I worked on the LEO III computer 50 years ago, when I was a student. I designed the circuit for the somewhat famous audible monitor mentioned here:

    http://en.wikipedia.org/wiki/LEO_%28computer%29

    It incorporated two divide by 10 stages, using unijunction transistors created from an NPN-PNP pair of BJTs, UJTs had just been invented and weren't yet available commercially. I was told that a conventional digital divider was too expensive.

    The earlier valve-based LEO machines were so slow that an amplifier could be connected directly to the CPU.
Sign In or Register to comment.