PID problem - Direcrt mode glitch and others

[TheDave2]

The following is with VL 9.3 and a V1040 with O/S 3.4 (00)

When using a PID loop in a Reverse mode ( the Action parameter bit in the configuration block = 0) everything is fine.

However, when setting the Action bit to Direct (bit =1), the only difference in function is that the CV output value is range inverted.

In other words, with Reverse action, the output goes from 0 to 100% as the PV gets further below the SP. Typical heating PID action.

Now with the action set to Direct (for Cooling control), the CV output now goes from 100% to 0% as the PV decreases further below the SP. The CV should go from 0% - 100% as the PV goes higher than the SP.

It seems that all the programmer did was reverse the CV output scaling when the Action was changed from Reverse to Direct. Ouch.

As a consiquense, I cannot use Direct action and must now interchange my SP and PV inputs to the PID to get the desired results.

Problem #2: Integral Action - Setting the Integral time to zero should get rid all integral calculations. While it does stop further integral error term injection, it does not remove the current integral error buildup in calculating the CV. I must put my own compare of the It =0 and then use the Force Error Integral to zero it.

Problem #3: Pb, It and Dt definitions. In some places, mainly in the Configuration box, there is reference to "(P gain)". This should be removed as Pb is not the same as Pg. Stick with either Gains of Bands / Time nomenclature. Having both leads to confusion as to which are being used.

Dave

[TheDave2]

[bump] - no confirmation or denial of what I see, by Unitronics ?

[TheDave2]

[bump - again - 2 weeks after original post] - no confirmation or denial of what I see, by Unitronics ?

[Emil]

Please take a log with PID server to show what you're explaining and send it along with your VLP project and some description of your system to support@unitronics.com. This will help us to better understand the problem and help to fix it.

Curently we had no any claims about inproper operation of "cooling" PID algorithm. We need more info to check the claim deeper.