Measuring temperature with high precission - Questions

[Reloadron]

Weird color scheme for K thermocouple. However, I have a little helpful tip. When working with Type K wire the Negative is always magnetic. Take a small maknet to a type K pair and the negative is magnetic and the positive isn't. That is for type K. Type J is the opposite.

Unless someone else does it, later today or tomorrow I can make you a circuit drawing if you would like.

Ron

Ron

 

[J_Nichols]

Thanks Ron,
But I think it isn't necessary. Maybe connecting a transformer to the output of the device will solve the problem?
I have a transformer to 15 VDC 2 Amps that I could connect to the relay of the controller. Then, that 250 VAC is transformed to 15 VDC and that voltage will be usefull to conect to the coil of the external relay.

I have other transformer of 10 VDC 2 Amps.
The external relay is 12 VDC (the coil).

I have uploaded the picture to a different server, I hope this time you can see the picture:
http://uploadpic.org/v.php?img=XWifoGM78E

 

[Reloadron]

OK. Generally speaking I would do things different. Your furnace has a 1.5 KW heating element that when on 100% will draw about 6 plus amps. Actually when cold it will draw more as the element resistance will be lower than at normal operating temperature. You want the relay / contactor that directly powers the element to be able to safely and easily handle at least 10 amps or more at 240 VAC. Obviously the controller with its relay contacts out will not do that. Controllers are not really designed to handle the load but rather drive another relay or contactor that powers the load. One big consideration is that the relay / contactor supplying the load must be able to handle current well in excess of the load current and this is for good reason. Every time that relay opens there is going to be arcing over the contacts. That arcing will, over time begin to pit the contacts. Eventually those contacts could literally weld themselves closed. Regardless of if the controller is calling for heat or not the element will be at 100% full time. This is one reason controllers have alarm outputs, it's nice to know your furnace is beginning to cook and whatever is in the furnace is beginning to fry. Personally I would be looking for a relay or contactor capable of handling 240 VAC at 15 amps for your application. I would also use a relay or contactor that runs off 240 VAC mains coil rather than low voltage but that is your call. Again, use the alarm function and set a high limit alarm set point and some sort of over temp alarm.

Remember that the relay contacts of the controller for control and alarm are rated at 250 VAC 3 Amps. Those contacts are designed for a resistive loads of AC. Therefore if you choose to use a low voltage DC coil contactor or relay to supply your element(s) you need to have the controller switching AC like the transformer primary.

Also, I assume you are aware that the originally sought control of +/- 0.1 degree C is not going to happen.

Attached are a few Type K TC wire examples. I still don't get the Blue and Red as here in the US TC wire is different. They can vary and I know Euro has different color schemes and standards. The images show a few variations of the insulating materials. Finally there is an image of a basic Type K dual thermocouple with associated connectors.

I saw the picture also. Pretty standard controller. Should work fine for you.

Ron

 

[J_Nichols]

thank you very much Ron, your explanations are helping me a lot.

I will buy a relay contact rated at 15 amps. The one I have in home I thought it was enough but I see that as you explain, arcing is a problem. And since I plan to run the furnace for long periods of time, the 15 Amps relay is needed.

The K thermocouple I have is like this
**broken link removed**

I live in Europe and it's possible that here there is different color schemes than in the USA.I don't know. Yesterday I tested the controller (only to switch on and measure temperatures) and it works fine.

Another question is... how would you connect the external relay to the controller? I said to you not to draw a schematic, but if you say me that you would do it in another way I'm interested in knowing how you would do it.

Also, I assume you are aware that the originally sought control of +/- 0.1 degree C is not going to happen.

What you mean? I will have less precision in the control of the temperature? If that is true, tell me if there is a solution.

 

[carbonzit]

What you mean? I will have less precision in the control of the temperature? If that is true, tell me if there is a solution.

[/quote]

A number of folks here have been trying to tell you that what you're asking is damn near impossible, certainly with the materials you've chosen to work with.

Again, why do you think you need such an absurdly high degree of precision? Doesn't really matter, though, because you're not going to achieve it anyhow.

 

[Mr RB]

I would not have said it quite as bluntly as Carbonzit but I agree in principle.

Are you expecting to get 0.1'C regulation using a thermocouple as a sensor and a relay as the heat control??

The thermocouple will never give you 0.1'C sensor resolution, and the relay will never give 0.1'C thermal stability even if you modulate it correctly. And the hotplate temp will vary a few 'C at different points on its surface even IF you can hold the average heater element temperature to a constant value...

It's like saying; "I'm going to whip my lame horse every now and then when he looks to be slacking, and he will run at exactly 7.000 MPH"

 

[Reloadron]

OK, one more time amd all of this has been well covered in 5 pages of post.

Starting with the sensor, the sensor is a standard type K thermocouple. The standard limits of error for a type K thermocouple are 2.2 degrees C or .75% whichever is greater. That is a given. The read and control accuracy of the controller is .1 degree C +/- 0.2% of FS (Full Scale. The full scale range for a type K thermocouple is 1300 degrees C so 1300 * .002 = 2.6 degrees plus the .1 degree = 2.7 degrees C. Thermocouple = +/- 2.2 or more plus the controller is +/- 2.7 so here we are pushing 5 degrees C.

You really need to go back and read all the post in this thread. You have also never stated exactly what this application is. You are not, not today, not tomorrow and not ever the uncertainty you seek with what you have. Way back in the beginning there were suggestions and links to the devices that deliver the uncertainty you seek. That is it. I posted T/C uncertainty way back in the thread.

Ron

 

[cachehiker]

There will also be measureable temperature gradients throughout the chamber, probably a degree or two warmer near the ceiling compared to the floor, walls will likely be a bit cooler than the center, etc. There's no way the entire volume of a typical environmental chamber could be kept within 0.1°C of a set point. Maybe a small region could be kept close to that with metrology grade RTD's, instrumentation, and well-tuned proportional control but that would get prohibitively expensive.

 

[J_Nichols]

than you very much to all.
I have enough answers to get the opinion.

 

[Reloadron]

Attached is a basic drawing.

Ron

 

[Jaguarjoe]

I will buy a relay contact rated at 15 amps. The one I have in home I thought it was enough but I see that as you explain, arcing is a problem. And since I plan to run the furnace for long periods of time, the 15 Amps relay is needed.

Only USD27 and it will last forever:

**broken link removed**