Sampling Rate Of A Control Loop, In A Process Plant

The easiest way is to contact the maker of the process control system you are working on. They will know for sure.

If you found us, you can certainly find them.

The Woodward governor that I worked with used discreet inputs (increase decrease pulses). They are on-time pulses, in other words as long as one pulse is energized the speed will move in that direction at a rate set on the card with DIP switches. There may be a slight delay, probably in the 10 to 20 millisecond range.

closer to 10 to 20mseconds is your sampling rate.

Dear Noudge,

can you please elaborate how did you come up with 10 to 20msec?

the central frequency of the micro controller is 32.6KHz, what I know is (correct me if I'm wrong) that control systems usually use 5 to 10 times the central frequency for sampling. for that I came up with an Analog to Digital converter with a 10 bit resolution and sampling rates in some what micro seconds. (f=1/T)

I'm sorry I misread the 10~20millisecond delay.

In designing a control system, it should be noted that the clearest resolution you could go or the fastest the system to respond at real time all depends on the weakest/slowest field intruments you have in your system, dont you think?

Sampling rate is the Baud rate configuration, some intruments provide a configuration adjustable to user. Read the manuals.

Connect a digital-storage oscilloscope to the output signal of the speed controller, and then analyze the waveform looking for "steps" where the signal updates to a new value (this is assuming an analog signal such as 4-20 mA). That step width on the oscillograph will be your controller's update time.

Lyn's suggestion to contact the manufacturer is a good one, but I know some industrial control systems allow the user to set their own update times, in which case the manufacturer might not be able to help you. I do not know if this is true for Woodward turbine controls, but it is definitely true for some large-scale DCS control systems (Foxboro I/A, Emerson DeltaV, etc.). One brand of control system I've heard of actually varied the loop update time based on the integral setting (the "I" in "PID") for that controller!!!

You might want to get this from the Vendor as suggested by many,

It is diffuicult to calculate accurately otherwise, Even an osiloscope will give you a fair idea only if there is a continuos change in output.