I would like to measure temperatures from 0 to 450 oC and send the results to PC. I drawn this circuit after months of research. As I'm a real noob in electronic, would somebody take a look at my circuit to check if I have no error in it? Thank you.

 

Hello,

You are using lineair amplification.
You know that the resistance is not lineair with the temperature.
It is written in this wiki page:
http://en.wikipedia.org/wiki/Resistance_thermometer
Take also a look at this PDF for some explanation:
Practical computational RTD linearization techniques for DAQ and embedded systems.

Bertus

 

I thought that I linearise the values in ATMEGA that I have, do you prefer using ADUC834 instead (as in your pdf link)?

as I said, I'm also not sure about whole system that I have, PT1000 interface, MAX232 to communicate with RS232, ATMEGA and it components (e.g. do I using the correct quartz there)

Do I need 1 Powerline for each of 3 PT1000s or it's enough to have one for all?

 

Hello,

Using the ATMEGA for calculating the linearization is fine.
The PDF I showed you has some formulas in it that might help.

I am only worried about the current source.
The parameters for jfets can be quite off, so the currents may be different on the three channels you want to make.
This will result in different readings from the different channels.

Bertus

 

does it help to take this transformator for current source?

http://www.reichelt.de/Printtrafos-...8AAAIAAC@QIyIb4394426f57055098e18657c5179124d

 

Hello,

Here is a page from aQD-TEC with some examples of current sources:
http://www.4qdtec.com/csm.html

I would go for circuit 4.
If you want to keep the PT1000 to ground, make the circuit "up side down" and use PNP transistors.

Bertus

 

I thought you worry about dropping the current on 2nd or 3rd instance of PT1000.
this is from Circuit 4:

If the current gets too high, the first transistor turns on and robs the second transistor of base current

EDIT: I got what you meant. I put that BF256 exactly for this purpose. see here: http://www.eleccircuit.com/fet-current-source-by-bf256/

 

Hello,

If you want three current sources, take a look on how to create them as in this post by Bill_Marsden:
http://forum.allaboutcircuits.com/showpost.php?p=117638&postcount=3

The first drawing in figure 2.4 will suit you. (just add a third transistor and limiting resistor).

The current will be about 0.7 / R .

Take care not to take a to high current for the PT1000, as self heating will disturb the measurement.

Bertus

 

Please check the datasheet for your thermistor. 450°C is beyond the range of all thermistors I have seen. You are in thermocouple territory now.