L35E PID Problems

Critt · Dec 16, 2011

Hello everyone and thanks for all the help over the years

I am currelty working on upgrading a water Plant PLC's from Micro 1500 to AB L35E. My problem is the PID instructions in the mirco are used to control flow and level to two filtration tanks. 4 PID's total. I have converted the logic and placed the PIDs in a period task. I just can't seem to get the flow or levels nailed down and working rather than they are jumping all over.

I am about to attach the logic just waiting for my membership upgrade to go through to I can attach it...

Any help would be great...

Ron Beaufort · Dec 16, 2011

just waiting for my membership upgrade to go through to I can attach it...

not sure what you mean by that??? ...

have you tried zipping the ACD file and then doing an attachment? ...

I just can't seem to get the flow or levels nailed down and working rather than they are jumping all over.

just a guess ... the PIDs in the MLX1500 use only the ISA/Dependent Gains equation ... the CompactLogix can use that format - but the default setting is for the AB/Independent Gains equation (basing this on my memory here) ...

so ...

maybe you used the original settings for the tuning - but you need the correct values for the new "flavor" equation instead ...

or ...

change to the Dependent Gains equation and see if that helps ...

DISCLAIMER: this is all GUESSWORK until we see the new program ... suggestion: post the old RSS file too ... that might come in handy ...

Critt · Dec 16, 2011

I only have 300 kb limit... I just upgraded my membership so I can post the zip of the ACD file...

Critt · Dec 16, 2011

here is the orginal .rss code

Ron Beaufort · Dec 16, 2011

if the ACD file is too big, do a File > Save As operation and save it as an L5K format ... those are usually MUCH smaller ...

then zip and post that file and we can go from there ...

Critt · Dec 16, 2011

here is the .l5k

Ron Beaufort · Dec 16, 2011

first things:

yes, your RSLogix5000 is set up for Independent ... suggest you change that to Dependent ...

you have your Loop Update Time set for 1.0 seconds ... suggest you change that to 0.1 second to match your Periodic Task rate ...

out of time ...

if the ideas above don't get things rolling, I'd look into your scaling next if I were you ...

gotta go ... wish I had more time to play ...

mellis · Dec 16, 2011

Another comment to add to Ron's...

I couldn't open the L5K file effectively (don't have rev 19 nstalled yet).

But if you entered the PID gain constants as Kc=400 and Ti=200 as they are in the 500 data table, you really want 4.00 and 2.00 instead. That will make it go nuts for sure.

dmargineau · Dec 16, 2011

Actually the Tuning Parameters Constants (gains) do no not quite match the Logix500 project ones:

Filter1 Flow MicroLogix Kc=0.20;Ti=0.75;Td=0 CompactLogix Kp=0; Ki=0; Kt=0

Filter1 Level MicroLogix Kc=4.00;Ti=2.00;Td=0 CompactLogix Kp=0.25;Ki=0; Kt=0

Filter2 Flow MicroLogix Kc=0.20;Ti=0.75;Td=0 CompactLogix Kp=0.15;Ki=0; Kt=0

Filter2 Level MicroLogix Kc=4.00;Ti=2.00;Td=0 CompactLogix Kp=0.25;Ki=0; Kt=0

Critt · Dec 16, 2011

The PID gain constants in the 1500 are as follows

Level Control
kc=4.00, Ti=2.00,Td=0.00, Loop time=0.1, E=PV-SP, Smax=1600, Smin=250

Flow Control
kc=0.20, Ti=0.75,Td=0.00, Loop time=0.1, E=SP-PV, Smax=500, Smin=0

Critt · Dec 16, 2011

The ouputs was going all over the map so I lowered them closer to zero because I was trying to find out if I needed to start from new or use the existing ones...

mellis · Dec 16, 2011

OK, based on this data... (thanks dmargineau)

Filter1 Flow MicroLogix Kc=0.20;Ti=0.75;Td=0 CompactLogix Kp=0; Ki=0; Kt=0

Filter1 Level MicroLogix Kc=4.00;Ti=2.00;Td=0 CompactLogix Kp=0.25;Ki=0; Kt=0

Filter2 Flow MicroLogix Kc=0.20;Ti=0.75;Td=0 CompactLogix Kp=0.15;Ki=0; Kt=0

Filter2 Level MicroLogix Kc=4.00;Ti=2.00;Td=0 CompactLogix Kp=0.25;Ki=0; Kt=0

Do everything Ron suggested: Loop update = Periodic Task time, Dependent gains. However, if you are still overreacting and oscillating with your Ki = 0, those probably are not the culprits. (With Ki > 0, the Loop update mismatch will have a dramatic effect.)

So, I will concur with Ron again, look to your input and output scaling. It seems like you are introducing excess gain somehow. Scaling is the next obvious place to check.

Critt · Dec 16, 2011

Can I use the same gain setpoints from the micro to the Compact logix?
I have made the changes int he program and will be onsite next week to test...

On some PID post the guys show graphs of their setpoint and such... any chance someone could tell me how to do that?

Another question, in the Micro program the is a mul and div instruction for getting the Feed forward Bias... How does this convert to the Compact? IS the FF only used in Manual mode and since my loop is auto not used?

Critt · Dec 17, 2011

Any of the PID experts have idea's??

Peter Nachtwey · Dec 18, 2011

You haven't provided enough informations to be helped as far as how to control the system. Think about what is important for the control of the tank levels.

L35E PID Problems

Critt

Member

Ron Beaufort

Lifetime Supporting Member

Critt

Member

Critt

Member

Ron Beaufort

Lifetime Supporting Member

Critt

Member

Ron Beaufort

Lifetime Supporting Member

mellis

Member

dmargineau

Lifetime Supporting Member

Critt

Member

Critt

Member

mellis

Member

Critt

Member

Critt

Member

Peter Nachtwey

Member

Similar Topics