Tieback definition

Kron

Member
Join Date
Jun 2006
Location
Montreal
Posts
17
Hi to everybody...

I would like to know if somebody can explain me what the "tieback" parameter exactly does in a PLC5 PID instruction?

I am new with this instruction and I have look in the RS Help files but I can't find something clear.

Thanks in advance.
 
From the PLC-5 instruction set reference:

Tieback
a word address used to implement bumpless transfer when using a
manual control station. The tieback is an output of a BTR instruction
from the station.
 
Greetings Kron,

here’s just a little bit more detail ... suppose that your operator has a potentiometer on his control board ... suppose that the potentiometer puts out a “plain vanilla” range of analog values from 0 through 4095 ... suppose that the value from the potentiometer’s channel lands at N7:55 ... the value would be placed there by a BTR (Block Transfer Read) instruction as mentioned in Ken’s post ...


suppose that when you programmed the PID instruction, you entered in the address N7:55 for the “Tieback” setting ...



next we need to know how you’ve set up your PID instruction ... specifically, did you use an N-type control block? ... or did you use a PD-type control block? ...



for the N-type (oldie but goodie) control block, you would now go into the PID and put the “Mode” setting to “Manual” ... the PID’s output (the CV) will now track the Tieback value ... this will allow the operator to turn the potentiometer up and down and manually control the system ...



or ...



for the PD-type (new and improved) control block, you would now go into the PID and put the “A/M Station Mode” setting to “Manual” ... the PID’s output (the CV) will now track the Tieback value ... once again, this will allow the operator to turn the potentiometer up and down and manually control the system ...



in many systems the Tieback feature is not used ... in those cases a value of “0” (zero) is entered for the Tieback setting ... this basically just turns the Tieback feature off ...



going further ... another source of confusion is the OTHER manual mode ... (yes, there is another one beside the Tieback) ...



for the older N-type control block, you could go into the PID and put the “Set Output Mode” setting to “Yes” ... the PID’s output (the CV) will now track the “Set Output Value %” setting ...



or ...



for the newer PD-type control block, you could go into the PID and put the “Software A/M Mode” setting to “SW Manual” ... the PID’s output (the CV) will now track the “Output %” setting ...



and one more thing ... if BOTH manual modes are enabled at the same time, the “Tieback” (also known as the “A/M Station Mode”) will win ... specifically, it will override the “Set Output Mode” (also known as the “Software A/M Mode” ...



that should be enough information to get you on the right track ... the best way to nail this type of stuff down is to experiment with a SAFE system ...
 
Thanks for the help

Thank you Ron for the help. I've been reading all of your quotes about PID and they are really helpful.

Best Regards,

Kron.
 
I've been reading this thread; very helpful information so far, by the way, so thanks everybody for the discussion.

I did have a question about a specific implementation that I'm considering:

I'm looking to implement a bumpless transition for a PID loop in an enhanced PLC5. The loop is used for temperature control on a valve. The system has experienced a "kick" in the CV upon the transition of the PID loop from manual back to auto, due to the error between the PV and SP at the time of that transition. I'm trying to eliminate that kick.

FYI, the PID loop can be controlled manually from an HMI, where an input or adjustment can be made to the CV directly.

The way the logic works is that in a startup condition it "ramps up" the output by gradually incrementing the PDXX:XX.SO while the block is in manual mode, to open the valve to a desired position automatically. After a set period of time it switches back to AUTO so that the PID can regulate it (temperature control), and it is upon the transition that a large CV change occurs.

I'm hoping to use the "tieback" feature (currently set to "0" on the PID block) to mitigate this problem. My plan is to continuously write the PDXX:XX.SO to a free integer and then use that integer to populate the "Tieback" field of the PID block. I suspect that I'll also need to change either the A/M Station Mode or Software A/M Mode (both are "AUTO" now) to either MANUAL or SW MANUAL (depending on which one) but I'm not sure which one to use.

My hope is that upon the transition from manual back to auto, the tieback value will be retained as the control output, and an instantaneous large change in the CV can be avoided. Then, the PID block will regulate the CV in a more controlled fashion.

Does this seem like a sound approach? Ideally I'd like to have a test system to play with this on, but that may not be feasible.

Thank you all in advance for your time and consideration of this issue. Your input would be most appreciated.
 
I'm looking to implement a bumpless transition for a PID loop in an enhanced PLC5. The loop is used for temperature control on a valve. The system has experienced a "kick" in the CV upon the transition of the PID loop from manual back to auto, due to the error between the PV and SP at the time of that transition. I'm trying to eliminate that kick.

Assuming a "bumpless" manual-to-auto transition is configured correctly (discussed later), the CV action ("kick") that occurs when going to auto due to error between set point (SP) and process variable (PV) can be eliminated by setting SP equal to PV when software manual (SWM) is unlatched and the PID is scanned. If SP is coming from an HMI (and it is acceptable to override the HMI set point), care must be taken to ensure the manual-to-auto SP change is not undone by the HMI, or other external input. If is is not acceptable to manipulate the SP in this way, it may require more logic around ramping the CV until error is smaller, or tuning the PID to be less responsive. Though both of these approaches seem to be defeating the purpose of the PID loop in general, but maybe not in your specific situation.

Tieback is intended for an external input that manually controls a loop as described in an earlier post. It is unnecessary when manual control uses the SWM function through the set output (SO) register. This is what allows the CV to track manual control such that a transition to auto will not bump the output to whatever the CV happens to be when switched to auto. (There is a special case to disable this behavior, but is not typical.)

In summary, if your loop is software controlled (i.e., no hard-wired manual station), then you want to stay with SWM/SO manual operation via the HMI. Then, if it is necessary to avoid any CV action on transition to auto, you will need to make the error zero by setting SP equal to PV. At that point you could ramp the SP to the desired regulation temperature, using a rate that keeps CV action at an acceptable level.
 
Last edited:
Let me try that summary again, based on how I understand RedKingRory's situation:

1. Don't use tieback unless you have a hard-wired manual control station.
2. Do use the SWM bit to switch the loop between manual and auto.
3. When in manual (SWM latched), always use the SO register to change the CV. Avoid directly writing to the address used for the CV. This will allow for "bumpless" transfer to auto.
4. To avoid an error-induced CV change (bump) when going from manual to auto, reduce or eliminate error by either (a) setting SP equal to PV prior to unlatching SWM, or (b) manually manipulate the loop such that PV is close to SP before unlatching SWM.
 
RedKingRory

The parameter you want is
"PV Tracking"

From the manual..

PV Tracking .PVT=0
.PVT=1
Enter whether you do not (0) or do (1) want PV tracking. Displays one of the following:
NO (0) – for no tracking
YES (1) – for PV tracking
Select no tracking if the algorithm can tolerate a bump when switching from manual to
automatic control. Select PV tracking if you want the setpoint to track the PV in manual
control for bumpless transfer to automatic control.

When you set the PID to manual with the above parameter the Set Point will automatically
track the Process value (PV). So when you go back to automatic the SP and PV will be equal (bumpless transfer) . You may have to readjust the SP to it's desired value after going back to auto.
 
Thanks for the input, everybody. I'm not sure if it's acceptable to manipulate the controller SP while it's in manual, but if I'm able to restore it to its original value after the transition back to AUTO, it may be acceptable.
 
Hey everyone,

A follow up to this for anybody interested. I ended up having to reduce the error in the PID by extending the time for which the ramp was active. By prolonging the duration for which the PID was in MANUAL, and allowing the PV to come down closer to the SP, I reduced the error enough so that when the mode was switched back to AUTO, that the kick was acceptable and control would be retained.
 

Similar Topics

Hello everybody, been a longtime viewer of the forum, this is my first time posting. Nice to meet you all. I am working with a system which uses...
Replies
3
Views
2,175
We have a couple of instances in our control systems where we want to, for example, force a PID controlled motor to a minimum speed or force a...
Replies
4
Views
1,708
I've done a forum search for "tieback" but no luck with an answer; only references. I guess you all know what it is. RS5000 online help assumes...
Replies
3
Views
1,756
Can any one tell me why the max Tieback is 4095 ?and what's the harmony service because I had a problem with the RSlinx and I checked the harmony...
Replies
1
Views
1,488
Guys, Anybody here knowing what's function of "tieback" in PID instruction in RSLogix5? What value that I should fill it? and what happend if I...
Replies
3
Views
18,575
Back
Top Bottom