I've been experiencing issues with Woodward Governor SPC's not following demand. On two separate SPC's, the analog position being fed by the SPC back to a ControlLogix is very quickly (30 to 100 mS) deviating by over 25% for several seconds, which trips the process. The SPC's are configured to use both devicenet and analog as demand with devicenet as the primary signal and are returning their position as both devicenet and analog. The PLC looks at demand verses analog feedback and trips the process if there is a deviation greater than 25% between its demand and the analog feedback. The demand may not even be changing, but the SPC suddenly reports that the servo is off sufficiently to trip. The servos themselves are typical servos with a bias current and 2 AC lvdt's per servo. This problem only appears every few hours. Otherwise, the servos appear to behave correctly. Can anyone provide direction as to what to even look for.

 

j_dubya,

Were the SPCs working for some period of time, and they "recently" started experiencing problems? If so, how long did they run without problems and how long have they had the problems?

If the problems "recently" started, was it after a maintanance outage? Or a lightning strike? Or a problem with the PLC or the output cards? In other words, can you remember what transpired before the problem started?

Do you have the ability to monitor the mA outputs of the PLC to see if they are stable when the problem occurs?

Have you tried working with the supplier of the equipment, or Woodward? I've always had good luck with Woodward technical support?

Is the hydraulic oil pressure stable, and at normal (rated) temperature?

 

Thanks for responding.

The SPCs are new and are part of a new controls system. Everything (LVDTs, hydraulics, etc) worked well with the old system. Conditions are normal.

We tested all day yesterday and found that the values we receive from Devicenet are occasionally returning bogus values. At the time we were having the issue with valves "not following demand" we were controlling the SPC's with Devicenet as their primary control.

We also determined that when the problem happened, the analog position signal we get back from the SPC was increasing faster than what the servo is physically capable of opening, so we thought it was either the LVDTs reporting bad values or the SPC is calculating and sending bad values. The problem has occurred on multiple servos/SPCs so I don't think it is an LVDT/wiring problem. The reported position of the servo jumped up 20% in 30mS and then slowly returned back to demand. The controllers are set up for 'P' control, so there's no integration going on.

At the time the problems occurred, we were not monitoring current loop or servo drive mA. We are now, but only via devicenet.

Today we are going to test the SPC action with analog as the primary control signal. I have a feeling that the problem is related to devicenet communication problems.