View Single Post
Old February 19th, 2003, 04:31 PM   #5
Peter Nachtwey
Member
United States

Peter Nachtwey is offline
 
Peter Nachtwey's Avatar
 
Join Date: Apr 2002
Location: Vancouver, WA, US
Posts: 6,618
Quote:
Originally posted by Ron Beaufort

Yes, you certainly did - but the most recent Allen-Bradley PLC-5 manual says: “... 1/5 to 1/10 times the natural period ...”. I wanted to demonstrate the results of those “published” settings - for better or for worse.
One can get by updating the PID at rates less than 10 times the system period IF one doesn't care about the response to transisients, and changes in load.


Quote:

I’ll certainly work on “a faster simulator” as soon as time permits. In the meantime, I already have a heat simulator available which has a natural period of 132 seconds. Theoretically, good control should be obtained by updating its PID every 6.60 seconds. This heat simulator also works on a 0.50 update - but I’m thinking that I should be able to alter my “adequate” Loop Update Time by a factor of 13 before I “hit the wall” of the 0.50 simulator update limitation. Calculating: 132 seconds per period divided by 0.50 seconds per update equals 262 updates per period. Now that’s “way often” to execute a loop. I won’t bet the rent - but I think that I’ll still see that going from 6.60 seconds - all the way down to 0.50 seconds - won’t give “substantially better” control. I’ll keep you posted.
In your current example the simulator updates more than 10 times faster than the PID update. For your 6.6 second PID update the .5 second simulator update should be fine.

The .5 second simulator update was not good for the previous example where the period was 15.18 seconds. Divide this by 20 at you get .76 seconds for the PID update. Divide that time by 10 and you get .076 seconds for the simulator update. This means the simulator must run about 200 times faster than the system period.

I too have learned a lot by making system simulators and tuning them.
This is good stuff.

I didn't know that that keeping the real and specified sample times the same on a PLC5 is a big problem. I agree that getting the update time right is a first or second priority along with calibration. This should have been made more user proof.
  Reply With Quote