In article , "Steve Nosko"
writes:

"Avery Fineman" wrote in message
...

In article , "W3JDR"

writes:

To Len:
Not quite it. In order to derive a control voltage for the
controlled
oscillator, the PFD output MUST BE A FINITE WIDTH....

...BUT THERE IS ALWAYS A PHASE OFFSET.

Len,

This is actually _not_ the case as Joe (W3JDR), says.

Heh heh, another one of those "long days..." :-)

I'm not going into control loops per se or argue formalized "types"
here. The original thread question was from someone worried
that the "dead zone" would cause instability if built with standard
logic gates. From there on it evolved into the usual type of
non-argument wherein one individual states an "inarguable"
case about only one control loop type...and no other argument
is allowed. :-)

[I should know better after doing computer-modem communications
for two decades but any response that says "wrong on all counts"
is a sure invite to a flame-fest... :-) ]

Once upon a time three decades or so ago, I sat (as the subject
whose project was under scrutiny) in a design review where the
chief engineer of the division sat in with senior staff. I was
assured that my version of a wideband limiting amplifier "would
not work." That, despite, at the moment, it was undergoing
active temp test with monitored signals applied, etc. Wanting to
keep my job, I kept silent and agreed with "wiser heads" (two of
whom were sensed as believing me but kept silent, perhaps for
the same reason). Memos were written afterwards making note
of the comments, etc., but the design continued on as devised
and became a viable part of a system. Project team got nice
congratulations afterwards, a few patent awards, etc., plus
photos of individual smiling hand-shaking with the chief. Was not
the time to rub it in that my "unworkable" design did, indeed,
work. :-)

Those who have tangible, working hardware are often put upon by
newsgroupies who say "it won't work that way!" Okay, no problem
for me. If they say it doesn't work, well, that's it. I'll just go off for
a while and let the regulars have at it...while my stuff is working as
stated. :-)

[if anyone thinks control loops are an "easy" thing describable in a
few words...well, I don't think they've been really involved with such
hardware...but that's only my opinion]

Gotta love these newsgroups! :-)

Len Anderson
retired (from regular hours) electronic engineer person