equivalent time sampling

On Sat, 18 Feb 2006 06:23:28 -0600, "maxascent"
<maxascent@yahoo.co.uk> wrote:

>The system I am designing is a pc scope. I guess I will know the trigger
>point. What I dont really understand is how you can generate precise
>offsets from the trigger if you want to sample in the GHz region as you
>are talking about picosecond values. I guess I dont fully understand the
>procedure
>

I think that last item is the one to focus on.

What you're attempting to make is a "Sampling Scope", and it was done
long before digital electronics even existed (i.e. vacuum-tube days,
1950's and earlier).  In fact, the "equivalent-time sampling
technique" itself stretches all the way back to before ANY electronics
existed....to around 1900.  It was invented in that era as a
mechanical device; used for drawing a real-time "pressure-volume-time"
diagram of an operating engine.

While the explanations given so far are good 'snippets', if you really
want to understand the basics, you should forget all about "digital"
for a minute, and get hold of the Sampling Oscilloscopes book of the
Tektronix Primer series.

It will give you an -excellent- grounding in the basic theory and
techniques of sampling; and -then- you'll be armed with the knowledge
to help you best implement as much of the system as possible
digitally.

Note that there are -two- methods of sampling....the 'linear' or
'incremental' method, as covered here by the various posters; but also
a 2nd method known as "random" sampling.  Random sampling has some
significant advantages.  With proper design, your system can be
selectable to do either.

I also highly recommend that you get a set of service-manuals
(probably $20-50) for the Tektronix 7T11 and 7S11 sampling plugins.  

These are 1980 era plugins which have pretty modern analog electronics
and also TTL-era digital stuff; so it'll be fairly close in relevance
to your goal of "all digital".  The 7T11 does random-sampling, and
these manuals contain excellent "theory of operation" sections; as
well as a full set of what I consider the nicest schematics ever
produced.  (i.e. Tek schematics of the 50's through 80's in general)

One of the biggest advantages of the "random sampling" method is that
it eliminates that need for a delay-line between the trigger ckt and
the sampling ckt.

That may sound minor, and someone mentioned that you just needed some
coax to do it; but neither is really true in your case.  You were
talking about several Ghz BW.  To get any kind of high quality (i.e.
'scope' quality) acquisition of pulses, you are looking at special
low-dispersion coax that costs 10 bucks a foot or more.

Note also that your -front end- must have flat BW all the way up to
the highest frequency to be measured; -regardless- of how slow the AD
is going to be.

For example, a vintage Tektronix 3S1 sampling-scope plugin
(1950's-60's) has excellent pulse-fidelity past a full Ghz, yet it
"digitizes" only in the Khz range.  I'm saying that this isn't RF
stuff where you can live with 1-3db of variances all through the
spectrum.  In needs to be FLAT...and issues like dispersion and
group-delay are very important in this application.

If you take a look at a commercial-quality 1+ Ghz scope front-end,
you'll quickly realize that building such a front-end with any quality
at all is a non-trivial exercise... <g>

Switchable attenuators alone will be an engineering challenge; not to
mention buffering, input protection, etc..  And in your case
(digital), the quality of your clocks, and the digitizer itself
(jitter).

Also, trigger-circuitry is a distinct field of art all by itself.
Test-equipment quality multi-Ghz triggers are not a simple matter, by
any means.

It's a fascinating field, and extremely satisfying when you succeed in
capturing events which lasted only a few picoseconds.  There's just
something awesome about that...to me anyway.  I've never had occasion
to use it in a product, but have dabbled in it solely for the love of
it for 25+ years now...

If you google for "kahrs sampling" (don't use the quotes), you might
find some other interesting material on it.  He's another "sampling
nut" who has written some journal papers on sampling and
network-analyzers (same principles) and had some useful papers and
links on his website the last time I looked.

good luck!



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! >100,000 Newsgroups
---= East/West-Coast Server Farms - Total Privacy via Encryption =---