OPT FOR SECURITY AND PERFORMANCE
Improve the security and performance of your plant with alarm management, performance supervision, asset supervision, web-based data visualisation, optimisation, advanced control, and tuning with the new TaiJi tool.
Rethink your working methods and obtain immediate results thanks to our services, trainings, and consulting expertise. Manage changes intelligently with our proven solutions.
March 2005
In this issue
| About us | |
| Classes and seminars | |
| Trick for this month | |
| To contact us |
See our previous issues :
http://www.topcontrol.com/techinfo/news.php
|
|
See our 2005 calendar, Classes & Seminars :
See at : http://www.topcontrol.com/training/tcalendar.php
|
|
Classes & seminars
Mexico, March 15-17
SE-100-3 - Seminario en métodos de control modernos
http://www.topcontrol.com/training/printable.php?id=90&type=cours
Montreal, March 24-25
SE-120-2 - Seminar on performance assessment techniques using PlantTriage
http://www.topcontrol.com/training/printable.php?id=29&type=cours
New-York, March 27, Easter, March 29-31
SE-130-3 - Seminar on advanced tools in process control
http://www.topcontrol.com/training/printable.php?id=28&type=cours
Montreal, April 19-21
EX-501-3 - Detailed Loop Analysis with ExperTune
http://www.topcontrol.com/training/printable.php?id=13&type=cours
|
|
Trick for this month : Kp was reduced and oscillations appeared!
It is well known that if one reduces proportional gain (Kp) in a PID controller:
- loop will be more stable,
- response time will be longer,
- oscillation will be reduced,
- overshoot will be reduced,
- robustness (process changes sensitivity) will be increased,
- etc.
Sometimes, it is not true!
If the PID controller has a parallel structure, reducing the
proportional gain will change the balance between proportional action
and Integral action. In most controllers (non parallel, Ideal/ISA,
Series) the output from the Proportional (Kp X Error) feeds other
modes (Integral and Derivative) therefore, reducing Kp will reduce Integral and Derivative at the same time. In a parallel structure, all
modes are fed by the error.
For example, both controllers below are equivalent:
| Parallel PID | ![]() |
Ideal PID | |
| Kp = 0.25 | ![]() |
Kp = 0.25 | |
| Kl = 6 r.p.m. (Tl = 10s) | ![]() |
Kl = 24 r.p.m. (Tl = 2.5s) |
If parallel controller Kp is reduced to 0.05, then Integral part remains the same and equivalent Ideal controller has now an Integral part stronger; the loop could start to oscillate. These oscillations will have a long period being induced by Integral; at first glance, they could be undetected.
| Parallel PID | ![]() |
Ideal PID | |
| Kp = 0.05 | ![]() |
Kp = 0.05 | |
| Kl = 6 r.p.m. (Tl = 10s) | ![]() |
Kl = 120 r.p.m. (Tl = 0.5s) |
Parallel controllers are popular in PLCs and in some DCSs. Flow loops cycling at long (minutes) period are frequent when parallel controllers are used; numbers seem acceptable but in fact sometimes they are not! Stay alert!
|
|
Comments, suggestions and requests are welcomed. Please e-mail us at: info@topcontrol.com
You are welcomed to visit our web site :http://www.topcontrol.com
|
49 Bel-Air Street, Lévis, QC
|
|
|
1020 James Drive, Suite A
|










