Are you certain this is not a "dog trying to catch its tail" problem?
I have not worked through the logic of this completely but let me see if I understand the problem.

The PLC controls the process temperature at a set temperature which cannot be adjusted.
Is this a PID controller or a bang-bang controller?

Now you want to change the set temperature to a different temperature.
Don't you think that the PLC will need a proper indication of the temperature, i.e. a properly functional thermistor?

What you need to change is not the resistance of the thermistor but the setpoint in the PLC. You need the thermistor to be able to correctly measure the process temperature.

 

In the currently working and unmodified system, what do the two leads of the thermister connect to?

I'm not sure what I can can contribute because I was thinking that what was being used is something along a traditional voltage divider with the thermister being one of two resistances. The voltage at the junction of the two will vary with temperature and that voltage is being monitored by a controller (as in an a2d or something) to determine whether to turn the heat on or off.

If that is the case, you want to change the voltage being monitored on your own so that you are not stuck with the set point that is fixed right now. Is that close?

 

Are you certain this is not a "dog trying to catch its tail" problem?
I have not worked through the logic of this completely but let me see if I understand the problem.

The PLC controls the process temperature at a set temperature which cannot be adjusted.
Is this a PID controller or a bang-bang controller?

Now you want to change the set temperature to a different temperature.
Don't you think that the PLC will need a proper indication of the temperature, i.e. a properly functional thermistor?

What you need to change is not the resistance of the thermistor but the setpoint in the PLC. You need the thermistor to be able to correctly measure the process temperature.

I wish I could just change the temperature setpoint in the Process PLC. That would be much easier.

The idea was that the Process PLC would still be reading a properly functioning thermistor, but with an offset.
Here is an example of what I was thinking.
Original Process:
Thermistor reads actual temperature to Process PLC.
Process PLC runs process to maintain Thermistor value to 35F, which is the temperature setpoint.

Modified Process:
Thermistor reads actual temperature to Mod PLC. (Lets say its 35F)
Mod PLC adjusts thermistor value based on user set offset. (Lets say offset = +5F)
Mod PLC sets Digital Pot to thermistor value + offset. ( 35F + 5F = 40F)
Digital Pot reads modified temperature to Process PLC. (40F)
Process PLC runs process to maintain thermistor value of 35F, which is a real temperature value of 30F because of the offset.

So the Process PLC thinks the thermistor is measuring 40F, when really its 35F, which causes the process to continue to run to maintain 30F. Because when the thermistor measures 30F air, with the +5F offset, the Process PLC thinks the measured temperature is 35F.

The setting of the digital pot would need to happen continuously to follow the change in temperature the thermistor is measuring.
All that make sense?

 

In the currently working and unmodified system, what do the two leads of the thermister connect to?

I'm not sure what I can can contribute because I was thinking that what was being used is something along a traditional voltage divider with the thermister being one of two resistances. The voltage at the junction of the two will vary with temperature and that voltage is being monitored by a controller (as in an a2d or something) to determine whether to turn the heat on or off.

If that is the case, you want to change the voltage being monitored on your own so that you are not stuck with the set point that is fixed right now. Is that close?

The thermistor is connected to an RTD input to a PLC. The PLC has an excitation voltage. It then measures the voltage across the thermistor, which gives a temperature value.

Correct. I want to alter the thermistor temperature value to the Process PLC to force it to run cooler or warmer. All of this effort because I can't simply change the temperature setpoint in the Process PLC.

 

The thermistor is connected to an RTD input to a PLC. The PLC has an excitation voltage. It then measures the voltage across the thermistor, which gives a temperature value.

Correct. I want to alter the thermistor temperature value to the Process PLC to force it to run cooler or warmer. All of this effort because I can't simply change the temperature setpoint in the Process PLC.

Ok. The PLC is designed to work with the specified thermistor and does not offer a way to change the setpoint. When the resistance of the thermistor is hit, the device does its thing.

To illustrate what I am saying....what would happen if you combined the thermistor with a 2K fixed resistor? It would hit the set point at a different temperature - right [edit I mean that the resistance that is seen by the PLC would occur at a different temperature]? So, you are, essentially, making an offset by adding a fixed resistor to the thermistor.

Now, what about if, using a 10 way switch, for example, you added one of 10 different resistors, changing the temperature of the resistance at setpoint. If you chose a different thermistor, you could likely hit your +/- 5 degree range while still maintaining temperature dependence.

 

I think we understand what you are attempting to do.
I believe the solution that you have suggested will not work. Forget about the PLC and thermistor solution for the moment. We first need an algorithm that will work for you. Can you post the make and model number of the PLC? Can you use a different PLC?
I need some time to think this through.
Maybe someone else such as @MaxHeadRoom, @strantor, or @cmartinez will come up with a solution.
(I didn't intend to miss out on other experienced members. I don't work with PLCs.)

 

Here is the transfer characteristics of a hypothetical NTC 10kΩ thermistor (truncated for simplicity).

°C - Resistance (ohms)
0 32k
5 25k
10 20k
15 15k
20 12k
25 10k
30 8k
35 6k
40 5k
45 4k
50 3k

It depends on whether the set point is below or above 25°C.
Let us assume that the set point is fixed at 35°C.
The controller will attempt to bring the temperature so that the thermistor is 6kΩ.
Now assume that we want a new setpoint of 40°C. The thermistor will approach 5kΩ while the controller's target is 6kΩ.
Insert a 1kΩ resistor in series with the thermistor.

Using this approach, you can select a different thermistor that will cover your desired range of setpoints.
Note that you need to choose between PTC and NTC thermistors. Your algorithm will depend on your choice of thermistor and setpoint range.

 

I have established that I need a variable resistor with values ranging from 800Ohm(or zero) to 100kOhm

Why 100k, if the thermistor is only 10k?

 

Why can't you modify the process PLC code to implement the offset? Is it source protected? Are you able to change the hardware config? If you can modify the code that is simplest. Next simplest solution is replace the thermistor input card with an analog input card connected to the wnalog output of your mod PLC. Can you do that? If you can't do either of those things then I think your original motorized pot idea is next best thing (and a horrible kluge).

 

Why 100k, if the thermistor is only 10k?

The thermistor is 10k @ 77F. 100k would give me a full range of temperatures to mimic.

 

Why can't you modify the process PLC code to implement the offset? Is it source protected? Are you able to change the hardware config? If you can modify the code that is simplest. Next simplest solution is replace the thermistor input card with an analog input card connected to the wnalog output of your mod PLC. Can you do that? If you can't do either of those things then I think your original motorized pot idea is next best thing (and a horrible kluge).

I agree but unfortunately I have no choice. I can't edit the code and can't modify the process PLC.

UPDATE:
I have gone forward with using the motorized pot with success. I was able to use a logarithmic equation to calculate the resistance required to input the correct temperature into the process PLC.

 

I agree but unfortunately I have no choice. I can't edit the code and can't modify the process PLC.

UPDATE:
I have gone forward with using the motorized pot with success. I was able to use a logarithmic equation to calculate the resistance required to input the correct temperature into the process PLC.

Glad you found a solution and let us know about it - don't know if I will ever need one, but I am making a note about those motorized pots...I am surprised that you can get that much precision, but obviously, you can.