Re: [gpsd-dev] SemPiTernal - Bounding PPS uncertainty

[Phil Salkie]

>
>> If PPS signals are always less than 500ms wide,
>
> Uh, no.  gpsd is in use with some GPS models that are exactly 500 mSec wide.
> Basically 2Hz.

Sigh, of course it couldn't be easy.

>
> My boat anchor Tek 465b still beats that. :-) 100MHz.  Only $1,200, in
> 1978, after employee discount.  They are cheap on ebay, but you'll get
> your exercise moving it.

It can be friends with the Tek 645 in my basement - I've got several
different plug-ins for it (dual trace inputs, transistor tester...)
and a wheeled cart to put it on.  It's also a storage scope - manages
that by having a long-persistence phosphor screen with an excitation
grid, lets you basically put enough voltage behind the phosphor to
keep whatever's currently displayed shining so you can still see it
long after the electron beam's done writing it. (Interestingly enough,
that's how some of the first computer memory was built - tubes which
drew dots on a phosphor, and were able to read back whether the
phosphor was in the excited or inactive state.)    I recall using it
in 1985 to figure out that a piece of industrial control equipment was
initiating serial comms by first writing two bytes at one parity
setting, getting an answer, then switching to a different parity to
write all the rest of the messages.  (I'd throw it out, but it looks
so mad scientisty, it'd be a shame to.)  Not unlike this one, but mine
has way more knobs (and lots more dust):
http://www.lydecker.org/images/image027.jpg