PID Power Loss Problem

Hello everyone. I have a small problem with a PID block on a SLC 5/05. The PID controls fine when running normally. After starting back up after a sudden power loss however, the output CV% is at 0 and stays at 0 until the PV climbs over 85% of the full scaled value. Then it starts controlling again as usual.
I guess my question is: does anyone have experience with PID strangeness when recovering from a power loss?

Any help would be useful!
Alan

Alan,

I have seen similar situations. I think that it is due to the "P" or gain factor. For many control scemes, it seems if the PID instruction is tuned for very precise control at ranges near to the setpoint, then when there is an upset (process stop and start produces a type of upset), and the PV goes way below or above the setpoint, then the small gain setting necessary at the fine-tuning level prevents the controller from recovering rapidly.

I am sure the experts here have a tuning solution that (theoretically) prevents this problem.

Power loss to what? I shouldn't have to ask this question

alaviolette said:

Hello everyone. I have a small problem with a PID block on a SLC 5/05. The PID controls fine when running normally. After starting back up after a sudden power loss however, the output CV% is at 0 and stays at 0 until the PV climbs over 85% of the full scaled value. Then it starts controlling again as usual.

Click to expand...

if the PLC is still running during the power loss then it can't control anything. The integrator is probably winding up or in your case down and it takes time for the integrator to unwind. The simple answer is to put the PID in manual mode to the integrator doesn't wind up or down. If putting the PID in manual mode changes the integrator's state then the integrators state should be saved before putting the PID in manual mode. When starting the system back up the integrator can then be restored to the original value.

If the power is lost to the PLC then there isn't much you can do but make sure the integrator state doesn't get over written when starting back up.

Thanks! I took both of your suggestions to heart. Lancie1, the PID was very fine tuned to the setpoint. I'm always too obsessive over my loops! I loosened up the tuning a bit and incorporated your solution too Peter. Used a UPS that had a relay output to indicate when power was about to fail. This triggered an input to my SLC and ladder logic put the PID into manual. After power up, back into auto and it works great.
Big thanks to both of you guys! I read your solutions at lunch and had my program running fine about an hour later.
This forum is fantastic! Glad I ran across it! Thanks again guys, I was sweating bullets!