Does anyone understand the difference between the gap gain on Honeywell C300 PID controllers and the gap gain on Honeywell xPM PID? There is a choice to use "Legacy Gap" on the C300, which uses the algorithm from the xPM PID gap controller.

The C300 gap controller seems to work very poorly - causes loop instability, but the Legacy gap selection works as I would expect a gap gain controller to work (i.e. less control action when the SP - PV is within the gap).

 

The PID algorithm in Experion controllers is calculated in incremental form. That is, the output CV of the PID algorithm is calculated incrementally during each execution cycle.

The equation for CV calculation of the PID algorithm in incremental form is as follows: During each execution cycle, the incremental output value is added to the accumulated output value.

CV (t) = CV(t- l) + Del_CVt

where

CV (t) is the current CV value

CV(t- l) is the CV at the end of previous execution cycle

Del_CVt is the incremental output in the current cycle.

Del_CVt is made up of 3 components. : Proportional increment, integral increment and the derivative increment.

Del_CV =K*( DelCp + Del_Ci + Del_Cd)

Where K = gain

DelCp = proportional component of the incremental value

DelCi = integral component of the incremental value

DelCd = Derivative component of the incremental value

The gain value K may change during each execution cycle in a PID Gap equation based on the value of PV. It was felt that including a gain change component will result in a more accurate CV value for non linear gain algorithms such as gap. So in Experion the CV calculation included a gain change component in the CV calculation when configured for GAP or nonlinear gain. Even though theoretically this is accurate, users migrating from TPS systems to Experion noticed a change in behavior in PID GAP algorithms

Experion R300 provides an option to revert back to the older calculation so that there is no change in behavior after migration. The option when enabled will allow the Experion controllers to have the same behavior as the TPS xPM controller.

<b>Legacy gap gain option</b>

To provide a solution a new Boolean Configuration parameter LEGACYGAP has been added to PID type function blocks beginning with R300. The default value for this parameter is FALSE in which case the Experion PID function block will calculate CV as currently implemented.

If the user sets this parameter to TRUE, the CV will be calculated similar to the xPM controllers. In this case the gain change component will not be included in the calculation and cause the PID Gap algorithm in the Experion controller to have the same behavior as the xPM controller.

This parameter is applicable to the following PID type blocks namely PID, PIDFF, PIDER and PID-PL. The parameter will be displayed as a checkbox on the right side in the algorithm tab of the forms for PID type blocks as shown in the following configuration form. The user will have to check this box during configuration to enable the LEGACYGAP gain option.

HOPE THIS HELPS. EXTRACTED FROM MANUAL

 

The main difference is that the C300 gap controls doesn't work, where-as the legacy gap does. The C300 gap controller also has a bug in it that intermittently flips the control action from direct to reverse (or vice versa). This bug is embarrassing for the controls engineer who implements it to say the least.

The legacy gap uses the control algorithm as per the original honeywell TDC 3000 controls. The C300 uses additional maths that aren't given to you in the manual to "improve" the control response. The improvements deliver unpredictable response and make the controller work more like a positional algorithm than the velocity algorithm it is supposed to be.

> Does anyone understand the difference between the gap gain on Honeywell C300 PID
> controllers and the gap gain on Honeywell xPM PID? There is a choice to use "Legacy Gap"
> on the C300, which uses the algorithm from the xPM PID gap controller.

> The C300 gap controller seems to work very poorly - causes loop instability, but the Legacy
> gap selection works as I would expect a gap gain controller to work (i.e. less
> control action when the SP - PV is within the gap).