evoo4u
Well-Known Member
Wow - this brought up some excellent resources - thanks! Among others:Adr_0 said:Have a look on Google about self regulating and non-self regulating systems and decide for yourself what we have as brewers.
http://www.controleng.com/single-article/fundamentals-of-integrating-vs-self-regulating-processes/c4511ec71b114a19097c92b5e3276089.html
I can see why we don't NEED integral action, also why trialling it in a steady-state system (constant volume, flow, and SV temperature with constant heat loss to atmosphere) worked. The integral never got to a "windup" stage, but it wasn't doing anything that couldn't be done with just P (and D as reqd).
I guess I thought that because the Auber controller goes into 'hold' mode while ramping (if time is set to one) until PV is within one degree of SV (on my setup), that integral wouldn't accumulate much. But as I understand, it's simply not needed in this process.
So, back out to the shed, and do another trial with varying P gain, and observing the results, introducing D as necessary.
Thanks again. (Sure helps to brush out the cobwebs )
EDIT: I probably had a bit of the 'Window's Font' syndrome* and thought that because I had a PID controller, I just had to use all three to get my money's worth.
( *when the computing age advanced from pin printers, and Windows gave access to a gazillion TT fonts, some folk thought they just HAD to use them all!! )