Best Book on PID ??

If you mean the controller, here's a start:

There are two aspects of this: the most obvious is the linear loop dynamics, the classic Laplace-transform closed-loop response. Then there's the far trickier nonlinear stuff: auto/manual control, bumpless transfer, overshoot, integrator windup, process slew limits (just ask a boiler to go from 0 to 100% steam flow in 30 seconds! Or

John

Computers also make it easy to characterize the open loop and design a pretty good controller, so all is not lost.

You can often make things work well enough by fudging, though. It sometimes amazes me how quickly I can get within 20% of the optimal solution just by rule of thumb. Of course, if you need to be within 5% then you have a lot of work ahead of you...

"PID Without a PhD", and that's a "simplified" view, please. Inspired by the directions given to union millwrights by control engineers who aren't allowed to touch the equipment in many, if not most, mills. Written by some schmo named "Wescott". Available through

It certainly doesn't teach control theory, but it will let you twiddle the knobs to get a working system most of the time (predicting how well you'll like the result before you start requires control theory, however).

controllers etc.

I have Astrom's adaptive control book, and I love it. Part of my admiration is inspired by the fact that he devotes a whole chapter to alternatives to adaptive control -- anyone who's writes a book about a pretty new theory then tells you when you don't really need it has integrity, in my view.

I've got one by Astrom, also the ISA book "Tuning and Control Loop Performance", and Liptak's Process Control Handbook.

Proportional-Integral-Derivative. It was invented more than 75 years ago. It refers to feedback control with terms proportional to the error, the integrated error and the derivative of the error wrt time. (error being Process Variable (PV) - Setpoint (SP)).

Determining the three proportional terms is "tuning" the controller. Over 90% of process control loops are PI or PID.

Try

OK, I'm biased, I actually built a lot of his equipment and was impressed with the precision of the mechanics and the true realtime software in Windows.

Jay

Right. I've rarely seen derivative do much to a real-world process but make it go nuts, the main exception being fast motion control servos. I read that derivative feedback was discovered accidentally because of a manufacturing defect in a pneumatic p+i controller, circa 1930 or something. There's a story somewhere.

I did manage to fry a few hundred k$ worth of NMR probes a while back. The main system control software (a c++ horror) would occasionally ask my box to go to +3000 centigrade, so I dutifully turned the heater power to max and waited for null. We had to add a battery-backed serial-protocol blackbox recorder to our uP code to catch them in the act. Then added a max_temp variable that can *only* be manually set from the front panel of our controller. This controller includes a setpoint slew limiter that creeps even slower when we're within 5 degrees of the setpoint, to guarantee no overshoot; a few degrees over and we can poach an enzyme that a thousand rabbits died to make.

John

Damn, topped again!

I did do some interesting things on a supertanker that earned me the nickname "Lights-Out Larkin."

John

Best book? It all comes down to some nasty math.

"Servo Mechanism Analysis", by Thaler and Brown, is probably as good a book as any... It's probably even on your bookshelf, as it is a good 50 years old.

There is no easy way out of these problems, either you characterize the open loop system, and design a proper PID equation to control it, or you fudge things and hope for the best. Computer based servo loops have made fudging things much easier than it used to be... but I am pretty sure, based on your relationship with computers, that you aren't planning to do a computer based PID.

-Chuck Harris

Jim Thomps> Recommendations for Best Book on PID ??

I never could get derivative to work in this particular application. The thermocouple is in a dewar maybe a foot downstream of the heater, and the gas flow isn't fully mixed when it hits the t/c. The heater emits four little jets of hot air, sort of like Stripe toothpaste. So the temperature is literally noisy, and derivative goes sort of crazy if set to any useful-to-the-dynamics level. That's a common problem in flow systems. I designed all sorts of cute flow stirring devices (a simple spring isn't bad) but couldn't get anybody interested in poking them into the dewar.

Customers are the other big control problem.

John

Hmmm, guys, very interesting point. In most of the loops I've done, the setpoint change just shoots through the error amp and the pid stuff and sort of comes out OK, but I can see how that ought to be optimizable.

John

Aha!! The perfect pointer! Thanks!

And what do you know, old "Nichols Chart" himself... last used one about 40 years ago ;-)

...Jim Thompson

You say that like it's a bad thing! ;-)

Best regards, Spehro Pefhany

Do you want to understand it or do you just need to tune it?

If the later, search on "Ziegler-Nichols". It's a technique that will get your loop tuned and running smooth in short order. If the former, there is no end to the options but before you spend money, search on "PID Controller" and you'll find plenty of info.

x-- 100 Proof News -

Click the "Download" button... and download the 757KB PDF version... scanned, then OCRd into Word, touched up and then PDFd ;-)

...Jim Thompson

(snip)

Eliminating that overshoot is one of the uses of properly applied derivative.

Honeywell used to (and may still) have in some of their controllers a PID algorithm that handled setpoint changes via integral action only. I think that was called PID B, as distinguished from PID A which was the standard ISA PID with independent gains.

Not exactly what you describe, but a step >>

I am frustrated with industrial PID controllers that force me to choose between having the P & D terms based entirely on error or entirely on the process measurement only. What I often need is a separate gain and derivative term for the process measurement and setpoint inputs to the controller. This is almost always superior the gain and/or derivative based entirely on error (setpoint - process measurement) or base entirely just on the process measurement and lets me optimize (after I define that word for the particular loop) the process disturbance response and setpoint change response, individually. For critical tuning, I have to use the feed forward connection and external math to to get all the factors I need. Why must I choose chocolate or vanilla when I really want a swirl.