6 PID Loops on a DL-260

[bfloyd806]

Hello all! I'm new to the forum and need a little help. I'm building a panel that will have 6 PID loops, each loop will control a single SCR that will heat a small tanks to 150*F. I'm having a hard time understanding how to setup the PID and need some help. I Selected the PID function in directsoft 5.3 and starting entering my addressing see below. I dont see anything in my logic and really not sure how to connect this Vmemory Locations in to my C-more and make it function correctly. Anyone have a sample ladder program and HMI screen that they know works so I could see how to make this work. Thanks for any help!! Email any files to billf@clarkpulley.com

Loop 1-
Setpoint Varible V15002
Process Varible V15003
Output V15005, Range Upper:4095 and Lower:0

Loop 2-
Setpoint Varible V15042
Process Varible V15043
Output V15045, Range Upper:4095 and Lower:0

Loop 3-
Setpoint Varible V15102
Process Varible V15103
Output V15105, Range Upper:4095 and Lower:0

Loop 4-
Setpoint Varible V15042
Process Varible V15043
Output V15145, Range Upper:4095 and Lower:0

Loop 5-
Setpoint Varible V15202
Process Varible V15203
Output V15205, Range Upper:4095 and Lower:0

Loop 6-
Setpoint Varible V15242
Process Varible V15243
Output V15245, Range Upper:4095 and Lower:0

[MikeS]

setting up PID loops in the controller is not really a DirectSOFT (or HOST Eng.) question, it's a PLC question that is better directed to the fine tech folk at Automationdirect.com.

[b_carlton]

Grab the CPU manual and go to the PID section. Read through there and use the PID setup function to set one loop. About the only thing you may have in the ladder is setting 'auto' or 'manual' mode. Possibly you may be moving the PV (measured temperature) to the PID loop location (though this can be set up to be retrieved by the PID automatically). You may also take the numeric output (CV) and use it to set up a time-proportioned firing of the outputs which control the SCRs. As suggested, re-post at the AutomationDirect forum for more extensive help.

[bfloyd806]

On a Cmore PID Faceplate Bar Meter is shows:
PV: I have this setup as the RTD input vmemory for this scr.
SP: Do I make a Vmemory address to allow the user to put the desired temp in there.
Output: This will be the analog output Vmemory used for the channel connected to this SCR.
Mode Bit: C2 (This is my question, Do I turn C2 on for Auto mode and off from Manual?)
Alarm Bit: ? Not Sure what to do with this.

[Controls Guy]

Unless you need extremely precise temperature control, you may not even need PID.  Tanks with liquid in them have adequate thermal inertia so that on-off control with a little hysteresis or anti-chatter timer should work fine.  I've done this hundreds of times even with a mechanical contactor or media solenoid and held somewhere between half a degree and maybe a degree and a half.  With an SCR (don't know if you mean a power SSR or a true SCR phase-angle fired power control), you should be able to do on/off control with even better results than I got.

[bfloyd806]

I'm using true SCR's made by Watlow. 2 of them will be controlling tank heaters and the other 4 will be controlling dryers in a hoist system. If I understanding you correctly I could write logic to say if the PV is Less then the SP then Load the value k4095 out to my Vmemory location? and then load the value K0 to my Vmemory location when the setpoint is within its range again? I guess i could write the ladder to load different vaules from 0 to 4095 depending on the offset between the SP and PV?

[Controls Guy]

Yes.  Historically the most basic form of setpoint control is to turn the output on when below the setpoint and turn it off as soon as you go above setpoint (for heating or other direct acting system).  If using a discrete output with a relay, contactor, or solenoid valve for steam or hot water, to do it this way can cause the output to chatter on and off and wear out the valve or contactor.  So to avoid that you can add some hysteresis (turn on below 150, but don't turn off till you get to 151, say) or a timer (once you've turned off the output, you can't turn it back on for 30 seconds or something).  If you already have analog, the equivalent would be as you say to set the output to 100% or 0%.  Setting it to different values depending on the error would be a basic form of proportional control.

I assume the dryers are heating air which is then blown on the parts.  If that's the case, you may still want to go with PID as the air temperature can change a lot faster than the temperature of liquid in tanks.

A couple points:

First, one of the main reasons for doing on/off control as I've detailed is to save money on the control parts if the thermal inertia of the load is such that short timebases aren't needed, but if you already have the parts needed for proportional control, you might just want to continue down that path.  Programming and tuning will require more effort, so there is still some reason to go with on/off.

When heating liquids in tanks, remember safety!  You should have low liquid level detection, so the heater isn't sitting there glowing if you cook off all the liquid, and/or an overtemperature sensor bonded to the heater.  These can certainly be wired into the PLC, so you can take appropriate control action, and so you can show the operator why the heater isn't on anymore, but you should probably also have a hardwired failsafe interlock for your safeties that physically prevents the heater from energizing even if the PLC fails to properly execute the interlock.

[bfloyd806]

I have a shunt trip installed on the breaker to trip the breaker killing the power to the SCR's. One question, for example. Say my process is working and my tank temperature is 150*F and then my RTD was to fail open showing the bath temp at 80*F. I guess this is where I would need a seperate controller as a backup?? Or would installing two rtd's on a single channel average the temp reading across the two RTDS?

[MikeS]

RTD modules typically have broken transmitter alarms to signal opens and/or shorts, I would think monitoring those alarm bits is a better option for handling error conditions.

multiple RTDs would allow you to keep running your process if one of the RTDs failed.