Ever used thermistors with a PLC?

N

Thread Starter

Nathan Boeger

We've successfully interfaced RTDs to PLCs and I think thermocouples. Has anyone had any luck getting cheap thermistors to work with standard (preferably AB) PLCs? I believe AB has a unit that should work in this application, but I don't think it will register a wide enough resistance range - my guess is that you should need to be able to accurately measure at least 300K Ohms.

I'm told that there are crappy proprietary controllers and hardware for this purpose whose companies frequently get bought out.

Can anyone verify the above and point me in the right direction for what I should be looking for - A cheap "easily interfacable" standalone device or a way to read thermistors with standard PLC I/O cards. TIA,

----
Nathan Boeger
Inductive Automation
"Design Simplicity Cures Engineered Complexity"
 
C

Curt Wuollet

It depends on what range of temperatures you need to cover. The resistance/temperature curve can be calculated as a system of 3 simultaneous equations in 3 variables with one or two calibration points. I did a PC chart recorder using the game port which measures resistance via RC time constant to a modest degree of accuracy. The accuracy for the purpose was quite good as I was using it to balance heating and cooling in various rooms and the span of temperatures was quite small. The way I would intuitively use these with PLCs would utilize a precision constant current source and measure the voltage across the thermistor. I would select a thermistor whose value at the desired mean temperature was such that it produced about a 5 or 10 volt change for the span of temperatures I wanted at a current that minimizes self-heating in the thermistor, and the current value used should be much greater than the input current of the analog module. If these values were contradictory, (quite possible) then I would use an opamp to buffer the input. Your error budget would include the self-heating error, the input current error, and any linearity errors. Then using the 12 or 14 bit resolution of the analog input, you could use a lookup table or solve the equations depending on your accuracy needs.

This sounds a little complicated, but it would only have to be done once and as long as you used specified thermistors, you should be repeatable within reason. I could make a small board (signal conditioner) that would take care of the CC and buffering. Or a board with an A/D converter and uP to do the linearization and scaling. Part of the problem of doing this is that there are differences in curves between vendors, but it wouldn't be hard to do for a given product line say, Omega or YSI. The cost could still be quite low compared to thermocouples or RTDs as these are commodity analog processing items. I have tossed the idea around a couple of times, but really haven't had a good reason to do it. I have a list of needed convertors and gizmos for the automation market in case I need to do a startup to keep eating.

There is no analog electronics industry here anymore but I hear needs expressed every day that could be met easily enough with a cottage design and manufacturing business. I'll put this on the list with the others. I expect it would be more accurate than thermocouples as they are normally applied and installed for the range of temps where they can be used.

Regards

cww
 
N

Nathan Boeger

Thanks for the info. Do you know of any products that might be able to input, say, (16) 2252 ohm thermistors and have it interface elegantly to an allen-bradley controller. It could be on Ethernet/IP or something like that. Thanks,

----
Nathan Boeger
Inductive Automation
"Design Simplicity Cures Engineered Complexity"
 
DGH offers modules that will measure four 2252 thermistors from 0-100C, 0.01 deg resolution, and 0.1% accuracy. The data is output via either RS-232 or RS-485 to a serial port on a PLC.

The D5451M or the D5452M can output the data using the Modbus RTU protocol to the PLC or host computer. A datasheet can be found at the link below.

http://www.dghcorp.com/datasheets/dgh_d5km.pdf
 
Top