REQ- High temperature PID control with PIC

[Parmin]

Anybody got PIC design for high temperature PID control?

I am talking about control up to 1300° Celsius or 2400° F 

[fernandodiaz]

Hi you can build digital algorithm pic16f, 18f arduino based  implementatión,  also can you implements opamps,  the low and high temp.  is the same control, thermopar

https://www.parallax.com/sites/default/files/downloads/122-28176-Process-Control-Text-v1.0.pdf

page 245

best regards

[solutions]

Would be handy to know what the heat source is and how you are planning to regulate it.

In any case, it's the thermocouple selection that determines the temperature range more than anything else.

Type B or C is what you need: http://www.asrichards.com/products/temperature/thermocouple_types.html

Other than the thermocouple front end/ADC-interface and PID tuning, I'd think you could just C&P PID code from a MicroChip app note.

[Parmin]

I guess I should reiterate, the reason I ask is because I do not want to reinvent the wheel and see if there are already a design available to suit the requirement. Generally based on my past experiences I should be able to design a system to do it, however why re design if one is already available.  Besides, since heating is not my forte, I might make a newbie error in the design which at that temperature might be quite dangerous..

The heat source would be a few 2 KW elements and the thermocouple will be K junction type since it is the easiest to obtain.

[CocaCola]

Do a Google search for 'MAX31855 PIC' breakout boards and sample code abundant...

[solutions]

K thermocouple won't work.

Only the ones I listed will.

That's why I said you may have to do the front end design, since most thermocouple circuits out there that I've seen  are K type.

[kreutz]

There are front-end ICs that include the cold junction compensation for a variety of thermocouple types (including type B), for example: http://cds.linear.com/docs/en/datasheet/2984f.pdf or similar. The good thing about this particular example is that it includes also the 24 bits Delta-Sigma ADC and offers SPI interface to the micro-controller.

[Parmin]

That is a great idea, Ok, I will use a front end thremal reader .

Thanks guys.. Consider this solved.

Posted on: 11 September 2015, 08:12:34 - Automerged
@solutions
I am reading the datasheet of MAX31855 on page 8 it listed that K thermocouple can operate up to 1370°C which is adequate for the requirement.
I am sure that the types you mentioned will do better, but since this is easy to obtain and can work to the temperature required I will try this one first.

[kreutz]

Although the temperature limit for a K type thermocouple is near 1300 degrees C, it is not convenient to work near this temperature for a long time due to premature thermocouple aging, it is convenient to work at 85% of the maximum limit or less (derating). Another thing is the thermocouple wire sheath (wire insulation) that have to resist an even higher temperature. Check also the environmental work conditions for each type of thermocouple, which also restricts the use of certain types, depending on the application (environment) rather than the temperature limits.

Please, take a look at: http://www.omega.com/temperature/Z/pdf/z017-018.pdf and http://www.omega.com/techref/colorcodes.html

[Parmin]

Yes I agree the aging of the thermocouple will be much faster at the limit temperature, it is still workable for this project.

While I want the kiln to work up to 1300 C it would not be always run at that temperature.
Most of the time the temperature used will be 800 C, occasionally it will be used at 1050 C and rarely at 1250


Posted on: 12 September 2015, 08:20:09 - Automerged
To add to the above, this kiln will be used to heat treat steel where simple carbon steel will reach austenizing temp at 800, stainless at 1050 and when needed to weld stainless it will be used at 1250.

[Gallymimu]

Also look at the Microchip MCP9600, it is a thermocouple frontend controllable over SPI AND it will handle the TYPE B Solutions suggested.

[CocaCola]

One nice thing with sticking to SPI interfaced thermocoupler chips is that you can build the CPU/display/control circuit as a stand alone modular part, and then experiment with different control chips and thermocoupler types with just software tweaks to the SPI interface...  This allows you to build and test (prototype) the device on the cheap with K type thermocouplers and upgrade later with ease if the K type doesn't work...

[solutions]

Now you have me curious, Parmin

Are you attaching the thermocouple to the steel you are heat-treating, or merely setting kiln temperature?

The latter will be an energy pig, the former will work more efficiently using a lot of initial kiln temperature overshoot - beyond even the 1370°C the clownboys at Maxim claimed you can do. As was mentioned in terms of degradation effects, the use of a shitty (technical term expressing a magnitude) part for the job and trashing ONE heat treated part as a result of that choice is a false economy.

Please let us know if you ever get a job with AirBus or Boeing 

[Parmin]

LOL, I am helping a knife maker friend to refurbish a knife heat treating kiln.
We are setting the temperature inside the kiln, and yes it is an energy waste, but that is the requirement for heat treating the steel.
Apparently to harden the steel need to be 'soaked' at the correct temperature for so long,  then quenched, then tempered at lower temperature etc.
To anneal (soften) the steel it need to be soaked at some temperature then gradually cooled over so many hour etc etc.
Every different type of steel require different heat treatment and temperature profile.

Posted on: 14 September 2015, 08:21:01 - Automerged
Oh, and the tolerance apparently can be as bad as +/- 20 C (5% or so) so it is very easy to work with.