Slow down my PID Logix5000

Quick question since I am doing this online. Don't want to kill anyone! I have a PID that looks at level. SP is say 4' and I change to 5'. Gets there tooooo fast.
ki-0.25
kd-0.25
time 5.0 sec
loop update .5sec

I want it to react on the control side real slow. I tried ki-.05 but don't think so. Being online I don't like to try it too long. Should I be moving my kd down and if so what is the minimum value?

CompactLogix, V11.11 on the 5000 and unconditional rung if it matters.

Thanks,
Dave

I'm going to limit my comments to the Time portion of the question.

Are you running this PID instruction in the Continuous Task or in a Periodic Task ? Is this the ladder language PID or the function block language PIDE instruction ?

PID instructions have an "Update Time" element that you configure (it's designated Tagname.UPD). That's actually to inform the PID equation how fast you're executing it; it doesn't change how the controller executes it.

Generally you'll put the PID instruction into a periodic task whose execution interval is equal to the value you put into the *.UPD parameter.

If you're running this PID in a program in the continuous task, you'll want to condition the rung with the .DN bit of a repeating timer to get it to execute less often than every single controller scan.

Getting to SP quickly is usually a good thing.

Why not ramp the SP from 4 to 5 slowly and leave the PID tuning alone?
The ramp should not be faster than what is physically possible.

Sounds good but I really need it slow in both directions. I thought about ramp but thought cutting Gain would be more effective. There are times when too fast overloads downstream.

I appreciate the help I just want the PID to move just like it is but 10 times slower.

switch off everything but the Ki this will slow down the opening time.

NO! NO! NO!

Most people would be very happy to have a PID that has a very fast settling time. The PID is doing what it should, so don't fuss with it!

IT IS THE SP THAT IS CHANGING TOO FAST!!!!!! WELL, ISN'T IT?
WHY IS IT THAT PEOPLE THINK THEY CAN MAKE STEP JUMPS WITH THE SP ANDs AND THE PID OUTPUT WON'T SATURATE, THE INTEGRATOR WILL NOT WIND UP, THE SP WILL NOT OVERSHOOT OR OSCILLATE?

Hint for DaveMac. The SP can ramp up and ramp down as slowly as slowly as required by your application. I like to use cosine ramps. These are very smooth.

Yes, shooter, you are right but the integrator only will cause overshoot and not react to changes in load or disturbances very quickly. This is not the correct answer.

and unconditional rung if it matters

Click to expand...

*sigh*. Are you 100% certain that you have the PID instruction executing properly in a periodic task or in a conditioned subroutine ?

I'm not being picky here, it's just that you could throw tuning parameters at the damn thing all day long and it still wouldn't matter if you're executing it every ten milliseconds.

Just reassure me.

I understand that there are different ways to do this with SP. What I am actually doing is looking at:

Current:
PID1 'good and smooth working well

New
If(DownStream_Loaded, PID2, PID1) 'With that I would need PID2
'and it is just PID1 but 10 times slower therefore the simplest
'way for me would be to slow copy PID1 and paste it as 2 and slow it
'by10.

Ken,
The kd is 5 sec and the loop update is 1 sec. That is more of what I was thinking taking my good PID and copying it to my new one and being able to go from 5 sec to 50 sec kd. Would that slow the loop down by 10 fold?