Review My Book

As long as you disclose that you are the author reviewing your own book is entirely respectable. I've seen a few such reviews on Amazon and think it would be good to see more.

-- Bob

--
Robert L. Knighten
RLK@knighten.org
Reply to
Robert Knighten
Loading thread data ...

i had noticed that the Chem E's in my alma mater (U North Dakota) had their *own* controls course and, once in a while, there was some ChemE student that took the EE one as a substitute. dunno exactly what differences there were, but i sorta imagine the ChemE controls course was less about servo-control mechanism and more about fluid flow and mixing controls. maybe more about non-linear plants.

all i remember is that you needed some "I" to get the error of position from set-point to eventually go to zero (ASRC has a natural "I" in the NCO). the "D" was used to anticipate some movement and react a little sooner. i s'pose "P" was something in between the two.

--
r b-j                  rbj@audioimagination.com

"Imagination is more important than knowledge."
 Click to see the full signature
Reply to
robert bristow-johnson

Integral action is about where you _were_, derivative action is about where you're _going_, and proportional action is about where you _are_.

And yes, many plants have integral action, or some slow low-pass process that looks an awful lot like it.

--
Tim Wescott
Control system and signal processing consulting
 Click to see the full signature
Reply to
Tim Wescott

...

i sorta see that, but i dunno what advantage there is to it. the

*design* imperative i remember from my control theory class 35 years ago was that having an integrator in a negative-feedback loop would (assuming it's all stable) integrate the error (between plant and set-point) and keeps on integrating it until the error gets to zero. not so with only proportional.

and that is congruent to what i remember was the design imperative around the "D" in PID.

for an Asynchronous Sample Rate Converter there is a circular buffer where the samples go in and interpolated samples come out. every sample going in increments the buffer input pointer by 1. every sample coming out increments the buffer output pointer by "step_size" which has both integer and fractional parts and, when things settle down, is the reciprocal of the sample rate ratio (which is not a given, you have to compute it with a sorta servo-control).

normally the "set point" of the output pointer is on the opposite side of the circular buffer from the input pointer (so it follows the input pointer with a delay of 1/2 the buffer length). the error signal is the difference from the actual output pointer and that set point. but the pointer position is the integral of the step_size variable (like in an NCO) and that is the reason for the existence of an integrator in the "plant" for an ASRC.

BTW, Tim, this ASRC i did on a SHArC (version 0.6 silicon) back around

1995 was the only time i have ever put anything i learned about control theory to a job that paid me. but i'm still glad i learned what i did about control theory when i was in skool.
--
r b-j                  rbj@audioimagination.com

"Imagination is more important than knowledge."
 Click to see the full signature
Reply to
robert bristow-johnson

Havinf long since retired, I didn't buy the book, but knowing Tim and what he intended it for. I'd bet on its being helpful. I revirwed Phelan's original book and designed several "couldn't-be-done" systems based on his ideas. They work even better with digital systems with enough oversampling to make te differentiators work well. Digital integrators don't drift. See

formatting link

Jerry

--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply to
Jerry Avins

mmer

t

Interesting, Thanks Jerry. (Just ordered a copy of Phelans book.) I'm stuck in the PID mind set.

George H.

=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF= =AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF=AF- Hide qu= oted text -

Reply to
George Herold

(and =3D

How could i, i am one.

?-)

Reply to
josephkk

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.