Loading ...
Sorry, an error occurred while loading the content.
 

Re: [VLF_Group] Re: Pauls VLF Stream

Expand Messages
  • Paul
    ... Yes. Here are some quick plots from this morning, http://abelian.org/vlf/live/pp100701dfc.png http://abelian.org/vlf/live/pp100701msf.png
    Message 1 of 31 , Jul 1, 2010
      Peter schrieb:

      > Promising,

      Yes.

      Here are some quick plots from this morning,

      http://abelian.org/vlf/live/pp100701dfc.png
      http://abelian.org/vlf/live/pp100701msf.png
      http://abelian.org/vlf/live/pp100701hbg.png

      'phase' is degrees relative to a positive-going zero
      crossing occurring 1970-01-01 00:00:00.000000

      There is a sawtooth of about +/- 30 degrees which I'm
      still trying to identify the cause of. It is equivalent to
      about +/- 1uS error in the timestamping. Struggling to
      cross check the timing with anything. Is this the limit
      of the GPS or have I got a software error? Anyone got an
      atomic clock handy? Perhaps I can hire one for a week...
      or is there some other way???

      > Are you going that binary way of storing?

      Probably. There are some advantages and I can make a program
      to read out the data into ASCII columns. Still some aspects of
      storage format to settle before I can commit.

      Alignment of multiple cards works very well, signal into left,
      PPS into right. Still a long way to go, much to type, another
      12000 lines or so before it is fit to release. The job would
      go quicker if the weather wasn't so nice. Too much time at the
      beach ...

      http://abelian.org/gaddings/index.php?page=img&id=56

      This really should be a winter project :)

      PS Wolf, I've had to alter the timestamp header: instead of

      int32_t secs; // Timestamp, seconds
      int32_t usec; // and microseconds

      I now have to use

      int32_t secs; // Timestamp, seconds
      int32_t nsec; // and nanoseconds

      The timestamping is running between 0.1uS and 0.01uS accuracy,
      (give or take the 1uS sawtooth error??), hence the need to increase
      timestamp resolution.

      --
      Paul Nicholson
      --
    • Peter, 54N 10E
      Hi Thomas, ... Thanks, a little bit exotic to me but may be an alternative and within my balance [^.^] Made some search today about frequency division and
      Message 31 of 31 , Jul 5, 2010
        Hi Thomas,

        > ...
        > http://www.onsemi.com/PowerSolutions/parametrics.do
        >
        > has pdip devices.
        >
        > http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=MC10H016PGOS-ND
        >
        > May have jitter less than 5 nsec????
        >
        > And, around 16 Euro??

        Thanks, a little bit exotic to me but may be an alternative and within my balance [^.^]

        Made some search today about frequency division and jitter: using asynchronous chains (ripple counters) the jitter increases by sqrt(N) where N is the number of flip-flops. That makes sense to me and confirms Tom's notices about the analogous characteristics of digital parts (noise). So synchronous dividers should be recommended?

        Peter, 54N 10E

        PS: Hi Paul, that's breakfast at "Ostfriesland".
      Your message has been successfully submitted and would be delivered to recipients shortly.