next topic »

[on:]

Hello All,

I am currently working on creating a Kiln Temperature controller that utilizes the AD595 thermocouple amplifier to display temperatures in degrees C on the serial monitor.  I have my code up and running based on the mV conversion table supplied with the chip for type K thermocouples, however i am not getting accurate temperature readings from the thermocouple.  I have tried 2 or 3 brand new type K thermocouples and am getting too low of a mV signal.  I know the AD595 has built in cold junction compensation but i am not sure if i need to have a resistor in the chip somewhere to get the compensation to allow accurate readings from the thermocouple. 

To give a quick example, i had the thermocouple in water that had been boiled and was reading around 80 degrees C on a thermometer.  The output of the thermocouple was reading 1.3 mV which corresponds to a temperature around 35 degrees C for type K thermocouples.  When i plug my thermocouple wires into the chip and run my code, I get the same temperature as the thermocouple leads around 35 degrees C.

Can anyone help or point me in the right direction to get this darned thing working.  It would greatly appreciated.

- Paclassics8

[Reply #1 on:]

Do you have the connections between the thermocouple and the chip very close to the chip so the connections are the same temperature as the chip?  That's what you have to do with a thermocouple.  If you try to extend it with a cable you are creating more thermocouples!

Read the data sheet for the chip.

[Reply #2 on:]

thanks for the reply,

Yeah the connections of the thermocouple to the chip are right next to the chip itself, and since i am just testing up in the office, both should be at the same temperature ( very close to room temperature).  I was reading about this and think the data sheet refers to not putting the chip close to a motor or generator or some component that gives off a substantial amount of heat. 

[Reply #3 on:]

I am doing the same thing. I have kind of the opposite problem. At low temperatures the readings I get are fine, but they drift lower as  the temperature increases. I am calibrating against an Oton controller with both thermocouples in the kiln at once. I take it to 2180 F. What kind of kiln are you working with?

In theory I think the lead in wires should be the same metal as the thermocouple, but for the accuracy you need to fire a kiln, that seems an overkill.  The Oton uses a 5 ft twisted pair. I get sbstantially the same voltage reading on the thermocouple at the end of these leads as I do with a pair of 10 inch jumpers.

I am using a simple voltage divider to drop the output down to a 5 volt max for the Arduino A/D. I found that I could detect no voltage drop across the output of the AD595 if I used resistors of 100K ohms. How are you handling the input problem?

I wrote a VB program to process the data coming in over the USB port. You can program multiple ramps and it graphs the real time progress.  It expects a transfer function of about 0.4 X 10mv though, because of the voltage divider. If you are interested, I'll send you a copy.

Good Luck

Bob C.

[Reply #4 on:]

The lead wires have to either be the same wire or similar in terms of  how they function as thermocouples. Otherwise you get thermocouples at each union.It is also smart to make sure your connections are good. Twisting the lead wire is smart as is shielding it especially for long runs. If you are having troubles try it with short leads, a foot or two. Make sure your supply for your controller/ amplifier etc is sufficiently sheilded from your relays. I built a kiln controller with a Radio Shack Color Computer and an Analog Devices Chip in the mid 1980's.
If you get something working with an Arduino, I would love to hear about it.