logic temperature control

[Koennie]

I am looking for a unit which can simulatnaeously measure 3 k-type thermocouples, average the two which are the closed together (and thus rejecting the one which is furthers off) and control a heater unit to bring the temperature to a certain setpoint.
We already have a control unit to control 1 k-type thermocouple so something which could only measure the 3 thermocouples, average the two and send this to the control unit will do to

 

[nbucska]

this is a typical job for a microcontroller

Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[itsmoked]

You are looking at some sort of lab instrument or data instrument and a computer.

If it is for a product then by all means it should be a small custom micro board that does exactly what you want and can then signal errors with indicators and can have alarm points, etc.

Do keep in mind that you can always parallel T/Cs and they automatically average their readings.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[danw2]

There are high end temperature transmitters that can take 2 T/C inputs, but I haven't run across one that can take 3.

I once did exactly the task you describe with a Honeywell hybrid controller, the HC-900, with its function block logic. The controller was ideal because the purpose was multi-zone temp control using average of closest 2 of 3 TC's in a zone. The controller had the I/O and the process logic run the PID for a complete job.

Dan

 

[GTstartup]

This is cheap and dirty but it works well. Wire them all in parallel and connect them to your existing control unit. You will get the average of all 3. If one T/C opens then you will only average the remaining good ones.

 

[GTstartup]

Sorry Itsmoked, missed your post, But good Idea!!

 

[itsmoked]

Haha. I was wondering..

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

Hi guys;
Am I going to learn something here?
This is my understanding of thermocouple theory.
With the excepion of some "Legacy" instruments, modern thermo couples generate a voltage. Modern instruments read this voltage without loading the circuit so as to avoid IR voltage losses in the extension wires.
When you connect two different voltages together you get a current flow. The resultant voltage will be higher than the lower voltage and lower than the higher voltage.
The resultant voltage will only be the average when the total resistance of each branch is equal. That will mean that all connections and the length and gage of the extension wires must be equal. Otherwise the average will be weighted in favor of the thermo couple with the lower resistance to the instrument.
Not to say that this method cannot be used, but it is well to be aware of the conditions, limitations and possible pitfalls.
Respectfully

 

[itsmoked]

While that is true it is generally not a problem. Most averaged T/Cs are relatively close together otherwise they fall into a situation where you want to know the two temperatures unaveraged.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Koennie]

Thanks for all your good advice. The solution with connecting them in parallel I already had figured out for myself. I wasn't sure whether that actually worked so i did the experiment. And it worked fine!
It is by the way for a lab experiment.
What material should I use for soldering them together?

Thanks!

 

[itsmoked]

Doesn't matter. Just make sure it has a higher melting point than your experiment.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com