Using proportional control only to retransmit thermocouple signal

Hi, I am new to pid and automation in general. I am hoping someone can verify if what I am planning will work.

I have a pid controller that I will be using a thermocouple as the input. It has a 0-10vdc linear output and a relay out for the second output. I am planning to use the relay output for a overtemperature warning buzzer, and the linear output will go to a sensor module which can interpret the 0-10vdc and send it to my mobile device so I can see the temperature of the oven at any time from anywhere. The oven is controlled by its own controls.

I'm not sure how this will work, though. If I only use proportional control and turn the gain to zero, will the output basically be an inverse of the rising oven temperature since it will never hit the set value (which I will set as max oven temp)?

I know there are likely better ways to do this, like with a thermocouple transmitter, but I am trying to do it with the parts that I already have.

Thanks for any input or suggestions.
 
I doubt your controller will allow you to have a gain of 0. The one on my desk goes down to 0.01, no lower. But you don't want gain of zero, you want zero reset action, because the device is a controller and controllers attempt to zero the error using reset action. You need some gain to get an output.

With some experimenting, I managed to get the linear control output (4-20mA but the principle's the same) on a Honeywell 2500 to reasonably track a thermocouple input with these settings:

The algorithm is not PID, it is "Proportional plus manual reset" operation (allows a fixed, manual offset value, disables the continous reset action that continously integrates the error signal)

You want a gain that will span whatever range is suitable for your magic 'sensor module'. I used a gain of 1.25 because a gain of 1.0 did not quite span the 4-20mA because the input range on this controller is -20 Deg F to 1200 Deg F

manual reset = 0
rate = 0

Setpoint = 0
action = direct

If you can't turn off the reset action the output will attempt to zero the error over time.
 
Thank you for the info. It makes sense now that I need some type of gain to get an output. Am I understanding this correct: if I cannot turn off the reset action, the output will eventually creep to match the set point? Is this the case even though there is no controlling load connected to the output of this controller?

Looking though the manual of my Loves Controls 32B-63, it does not appear to have this setting.

http://www.dwyer-inst.com/PDF_files/E_90_BPC_rev5.pdf
 
You luck out. That controller should do it.

1. PB, not gain
This controller does not use gain; it uses Proportional Band (PB).
Gain * PB = 100
Gain 1 = PB 100
Gain 2 = PB 50

adjusting the PB will adjust the 'span' of the control output, which Voltage value equals the temperature value..

2. This controller lets you configure the "in" value (integral function, same as reset) to 0, which turns off the automatic, continous reset action.

Configuring the PdoF setting will allow a trim offset to the control output, presumably bipolar, positive or negative.

See the highlights on the screen shot from the manual

linear control output tracks input PV.JPG
 
Thanks for jogging my memory from a long forgotten episode 30 years ago. I only ever encountered this somewhat out-of-the-ordinary controller configuration once before when a small food processor had bought a PID controller to automate the manual control of the injection pump that pumped an additive into a main product flowstream.

Their problem was that the recipe changes from batch to batch required varying flow rates of the main product but the operators would not always change the manual setting on the additive injection pump at the beginning of the batch. They bought a PID controller thinking it would automate the additive injection pump, which had a 4-20mA demand signal input.

They didn't understand that PID is a closed loop operation; that the variable being controlled has to be measured. But they were measuring flow rate, not the concentration of the product plus additive. The concentration needed to be held at a constant value at any flowrate, but wasn't because of the failure to adjust the manual pump control for every batch.

With no concentration measurement, the control strategy had to be open loop, rather than closed loop, with a controller output signal that changed in a linear proportion to the measured, main flow rate which was then fed to the additive pump which would then change its output as demanded by its input/demand signal.

The PID controller setup was for proportional only (with manual reset) control which provided an output signal proportional to the input where the manual reset function disabled the automatic integral/reset action. The tuning gain/PB term/factor provided some scaling for 'gain', and the manual offset term provided an offset to the output and the rate term could zeroed out. Direct action provided an increasing output signal with an increasing flow rate; decreasing output signal with a decreasing flow rate.

Configured as such, the PID controller worked OK, it eliminated the problem of the operator failing to change the additive flow rate from batch to batch where the main flow rate was changed depending on the batch recipe.

Sometimes, not often, a square peg will fit into a round hole with enough shaving around the edges.
 
I connected it yesterday with your recommended settings and spent some time messing with the proportional band setting. I was able to get a near perfect linear output from 0-10vdc, with 70 degrees F (ambient/off temperature) putting out 0.71 volts, and the output maxing out at 10.02 volts at 600 degrees F (the maximum temperature that will ever be used). The number used for pd ended up being 300. I'm glad you pointed out the direct action on cooling. I missed that when I was reading the manual, and saved me from having to do the inverse voltage to temp calculations.

Thanks again for helping out. It was a great learning experience.
 
Top