Yo Larry

Robert Monsen · Mar 28, 2005

Active8 wrote:

No. Read Middlebrook's paper to understand the issue.

That the paper at your site that goes with your loop cutter macros,
right?

http://ardem.com/index.asp

--
Regards,
Robert Monsen

"Your Highness, I have no need of this hypothesis."
- Pierre Laplace (1749-1827), to Napoleon,
on why his works on celestial mechanics make no mention of God.

John Larkin · Mar 28, 2005

On Mon, 28 Mar 2005 09:07:55 -0700, Jim Thompson
<thegreatone@example.com> wrote:

On Mon, 28 Mar 2005 08:04:32 -0800, John Larkin
jjSNIPlarkin@highTHISlandPLEASEtechnology.XXX> wrote:

On Mon, 28 Mar 2005 08:14:06 -0700, Jim Thompson
thegreatone@example.com> wrote:
[snip]

Huh?

If you arbitrarily inject a signal into a circuit ALL you MAY see is
some peaking if the loop is near instability... you won't know
anything quantitative.

Transfer functions are quantitative.

John

And uninformative ;-)

...Jim Thompson

Heretic! Black helicopters are descending on Phoenix to take away your
Brass Rat Ring.

But really, if you can inject a sweepable sine wave or (better yet) a
nice slow square wave anywhere in the loop, you can then probe pairs
of points and get a good idea of the frequency-phase (or better yet,
step) response of subsections, closed loop or not. The only messy part
is when the input to a section is ugly (as in, not a clean step any
more) in which case you have a convolution problem on your hands.
Frequency domain ain't bad: amplitude and phase are easy to measure at
the two points as long as there's enough signal to resolve things.

But that's just analysis. In control loops, it's usually easier to
design the loop than to analyze it. Actually, lots of things are
easier to design than to analyze.

John

Larry Brasfield · Mar 24, 2005

"Genome" <ilike_spam@yahoo.co.uk> wrote in message
news:m9y0e.1156$kj5.480@newsfe6-gui.ntli.net...

"Larry Brasfield" <donotspam_larry_brasfield@hotmail.com> wrote in
message news:ODm0e.1492$vB1.1078@news.uswest.net...
"Genome" <ilike_spam@yahoo.co.uk> wrote in message
newsPl0e.504$xs4.117@newsfe6-win.ntli.net...
...
Hey, my name is Genome. You've been set a
challenge.

Game over. You win.
....
Oh well, if you look in ABSE again then you'll find the same circuit
with a few little modifications.

Look, I don't mean to be unkind here, but you
must realize your game can only be played and
won a certain number of times. Given your good
simulation of sincerity before, you now have a high
threshold to overcome before getting me in.

If you want to measure loop gain and phase then you use the method
shown.

Break the loop with an AC voltage source and plot V(B)/V(A). There may
be some caveats about impedances but I'm sure you'll get the general
idea.

If *I* want to measure? What do you want? (And
do remember that threshold problem. I cannot even
imagine, let alone suggest, how to solve that. ;-)

You have been measuring the closed loop response.

I don't recall that with "your" circuit. (I have no idea
whether it is yours or not.) I never looked at it with
enough care to get that far. I never saw anything to
make me think it might be or become real.

There is an example in the LTspice directory somewhere as well.

Ok. I believe you. So?

HTH

? If that's a joke, I like it.

--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.

Jim Thompson · Mar 28, 2005

On Mon, 28 Mar 2005 15:03:57 GMT, "Harry Dellamano"
<harryd@tdsystems.org> wrote:

"John Larkin" <jjSNIPlarkin@highTHISlandPLEASEtechnology.XXX> wrote in
message news:butd41ps1iqosiaisso2urf76vdihghqb9@4ax.com...
On Sun, 27 Mar 2005 09:59:41 -0700, Jim Thompson
Oh, about open-loop response: if in a simulation you inject a signal
anywhere in the "closed" loop, just to have something to work with,
can't you just observe the vector ratio of A/B, where A and B are the
voltages at any two arbitrary nodes? That's the "open loop" transfer
function from B to A. After all, in Spice there's no noise, and if you
need a gain of 1e8 to resolve things, just do it.

John
Is John not correctly in stating that special injection points are not
necessary as long as the injection point is not part of the loop under
investigation? Maybe there is a gain issue with your injection point.
Harry

Huh?

If you arbitrarily inject a signal into a circuit ALL you MAY see is
some peaking if the loop is near instability... you won't know
anything quantitative.

As someone pointed out earlier in this thread, Phase Margin and Gain
Margin tell the tale.

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice

480)460-2350 | |
| E-mail Address at Website Fax

Jim Thompson · Mar 28, 2005

On Sun, 27 Mar 2005 21:06:57 -0800, Robert Monsen
<rcsurname@comcast.net> wrote:

Active8 wrote:

No. Read Middlebrook's paper to understand the issue.

That the paper at your site that goes with your loop cutter macros,
right?

http://ardem.com/index.asp

480)460-2350 | |
| E-mail Address at Website Fax

John Larkin · Mar 28, 2005

On Mon, 28 Mar 2005 08:14:06 -0700, Jim Thompson
<thegreatone@example.com> wrote:

On Mon, 28 Mar 2005 15:03:57 GMT, "Harry Dellamano"
harryd@tdsystems.org> wrote:

"John Larkin" <jjSNIPlarkin@highTHISlandPLEASEtechnology.XXX> wrote in
message news:butd41ps1iqosiaisso2urf76vdihghqb9@4ax.com...
On Sun, 27 Mar 2005 09:59:41 -0700, Jim Thompson
Oh, about open-loop response: if in a simulation you inject a signal
anywhere in the "closed" loop, just to have something to work with,
can't you just observe the vector ratio of A/B, where A and B are the
voltages at any two arbitrary nodes? That's the "open loop" transfer
function from B to A. After all, in Spice there's no noise, and if you
need a gain of 1e8 to resolve things, just do it.

John
Is John not correctly in stating that special injection points are not
necessary as long as the injection point is not part of the loop under
investigation? Maybe there is a gain issue with your injection point.
Harry

Huh?

If you arbitrarily inject a signal into a circuit ALL you MAY see is
some peaking if the loop is near instability... you won't know
anything quantitative.

Transfer functions are quantitative.

John

Jim Thompson · Mar 28, 2005

On Mon, 28 Mar 2005 08:04:32 -0800, John Larkin
<jjSNIPlarkin@highTHISlandPLEASEtechnology.XXX> wrote:

On Mon, 28 Mar 2005 08:14:06 -0700, Jim Thompson
thegreatone@example.com> wrote:
[snip]

Huh?

If you arbitrarily inject a signal into a circuit ALL you MAY see is
some peaking if the loop is near instability... you won't know
anything quantitative.

Transfer functions are quantitative.

John

480)460-2350 | |
| E-mail Address at Website Fax

John Woodgate · Mar 28, 2005

I read in sci.electronics.design that Jim Thompson
<thegreatone@example.com> wrote (in
<mt5g415mv2nan5eqqv2e1ki39ulb9ablua@4ax.com&gt

about 'Yo Larry', on Mon,
28 Mar 2005:

R. D. Middlebrook
âMeasurement of Loopgain in Feedback Systemsâ International Journal
of Electronics
Vol. 38 #4 pp.485-512
1975

Is it accessible for a reasonable cost?
--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

Yo Larry

Robert Monsen

Guest

John Larkin

Guest

Larry Brasfield

Guest

Jim Thompson

Guest

Jim Thompson

Guest

John Larkin

Guest

Jim Thompson

Guest

John Woodgate

Guest

Welcome to EDABoard.com

Sponsor

Online statistics

Forum statistics

Yo Larry

Robert Monsen

Guest

John Larkin

Guest

Larry Brasfield

Guest

Jim Thompson

Guest

Jim Thompson

Guest

John Larkin

Guest

Jim Thompson

Guest

John Woodgate

Guest

Log in

Welcome to EDABoard.com

Sponsor