yet another pid question

[dougpdf]

This is my first post here...so HI
I own a small injection molding operation (thermosets) and I've had good success replacing the relay logic in most of my older presses. They have fairly simple hydraulics,a few proportional pressure valves here and there and I have used linear pots on the clamp and inject. Point being, I've done ok with analog ins and outs.
Up until now I've been using little Fuji's or similar for my mold and barrel heats. In my case, molds are hot (cartridge heaters) and barrel is not so hot with heater bands over water jackets with simple sol valve allowing water flow to cool(friction).
Over the years I've wound up with a IT-6 that I've never utilized. I've been trying without success to program using pid. I've read every post here in the forum and so much of it just zings over my head. From what I read, relays aren't the norm, but I always have my Fuji units switch a small relay to switch the heater relays or mercury relays. I'd like to think I could do the same with the plc.
I've used the repeating 1 sec timer method and it's close, but still over/under shoots and I haven't even thought about the cooling part yet....
I'm waiting for an "ah ha" moment, but am now asking for some help.
Is there not some "simpler" code for this someone can point me to?

Micrologix 1500LSP AWA
IF4XOF2
1769 IT6
a couple of OW8s
Using RS 500

Thanks
Doug...in Ky

 

[Ken Roach]

Doug,

I think you'll get some better response if you can provide some more details about how you've configured your PID loop and the responses you're getting from the machine.

At a minimum, prepare a document describing how the PID equation has been configured, including the gains, and a trend of the PID in operation.

As I'm sure you're aware, injection molding temperature control isn't trivial. Allen-Bradley even made a plastic-specific "Pro-Set" controller module years ago for the SLC because the existing PID equations didn't provide the right sort of control.

Ideally, you would be able to capture a trend of the standalone loop controller in action, with the temperature and the heater duty cycle on the same trend. If you then know the tuning parameters of the standalone loop controller, you might be able to translate them into working parameters for the PLC.

 

[dougpdf]

Thanks Ken
I've been studying the parameters in my little stand alone units and have improved it a lot. The molds are close enough +/- 1 F.
If I can't get barrel temps close enough, I'll get some info together like you suggest or attach my file.

I do have one question in particular..
Just to verify, I will need to configure separate pids for the cooling portion? correct?
Keeping heat and cool from arguing with each other is going to be fun. I'm guessing dead band.

Thanks again

 

[Ken Roach]

Frankly I don't know a thing about injection molding temperature control.

Do you generally run both heat and cooling at the same time ?

I know that the ControlLogix "Split Time Range Proportional" instruction will run both cooling and heating, or just one.

You'll probably implement something like that, where the % output of the PID instruction will drive heaters and cooling valves.

In the heating applications that I do, we use exclusively solid-state relays or Watlow solid-state SCR controllers. Mercury relays are almost unheard of anymore, and our cycle rates would burn out contactors in a matter of months.

 

[dougpdf]

This would be more like what you'd see on an extruder..I think.

Most people processing thermosets control the barrel temps with temp controlled circulating fluid, like you'd typically see controlling thermoplastic molds. Frictional heat is prevalent in this type molding material.
I use cooling jackets (in conjunction with band heaters)with the fluid dead headed against a solenoid valve. The valve would open briefly for the cooling. This setup is much easier to maintain.
The process oscillates between heating and cooling to maintain set point.
Once up to temp and running, the Fuji units typically cycle heat or cool only momentarily every few seconds.
I didn't see that option of splitting the pid on my 1500, but will look closer.

 

[Ken Roach]

The PID itself doesn't have a split point; that's a separate instruction (SRTP) on the big ControlLogix family controllers.

What I meant was that when you get an output from the PID instruction in the MicroLogix, you will have to write your own logic to apply that to turn on the heater or the cooling valve over a particular time range.

This might be what you were referring to as a "repeating 1 second timer method", or you might be talking about something else.

 

[dougpdf]

Yes it is. I'm using a repeating timer to convert the pid output to on/off. I saw it in another post.
I have no idea what a SRTP instruction is. The more learn, the more I realize what I don't know. Yikes, this is complicated for me.
Thanks

 

[widelto]

time proportioning

Doug:

This is for SLC 5XX for can be used in your case.

 

[MasterBlaster]

The more learn, the more I realize what I don't know.

That, in a nutshell, is the story of my life!

 

[dougpdf]

Thanks all !!
I think I'm getting it.