Thermocouple Cable Length

[Sobakava]

I am not familiar with thermocouples much. I need to measure temperature in a room in different points. I made an thermocouple amplifier using AD595/4. I don't need so much precision measurement. (+/- 2C degree in 20..60 range is enough I think)

I need to know how much can I extend the cables of thermocouples? (J or K type) Because I suppose I can not place the amplifier board in to the room because of temperature (>80C Degree) and high moisture.

I've seen such system and they used regular cables longer than 10meter.I don't know how good is it...

What do you think?

 

[FANT]

I have used cables until 50 meters.
Take a lot uf care about the gradient between the two cables !


Mandi

 

[river78]

It depend WHat is kind of your Thermocouple? Current or potential. If it's kind is 4-20 mA u can use long cable.

 

[rauol]

8) Be carefull about shielding. try to use a double shielded wire.
if your temperature is less than 100 Deg. C. you can use the amplifier near to the source which will reduce noise problems.

 

[Sobakava]

do I have to connect shield of thermocoupe cable to my board's GND?

 

[goshka]

Why You down't use digital ds1820 sensor?
300*ds1820 sensor on IEEE1394 cable 300 meters lengs with active pull-up - works ok.

 

[goshka]

See also AD693 - Loop-Powered 4–20 mA Sensor Transmitter

 

[Sobakava]

goshka, are u sure about ieee1394 cable? isn't it for video?

 

[Ale]

To extend the thermocouple cable, you need use a Weasthone bridge, to eliminate the cable selft resistance.

cooming soon i post the diagram.

Sorry, i'm not good with english.

best regard

 

[golfbumb]

Actually, you should not have any trouble measuring anywhere in a "room" from the outside of that room if you are using an AD594 which has sufficiently high impedance to negate IR losses in fairly long wire itself.

The best setup would be to use have a TC cable with a long enough pigtail to reach your 594 directly. You then take out the variable of in-between connections.

But if you need to extend an existing thermocouple, you must use the same kind of wire for each of the wires to be extended and use thermocouple connectors classified for the type of thermocouple you are trying to extend. Dissimilar materials in any connection, other than the junction itself, will introduce undesired juction voltages and large measurement errors. Manufacturers also make extension wire which is materially similar to the TC wire, adds small amounts of error (plus or minus a degree or so), but is cheaper than the TC wire if costs are an issue.

With regard to your second post, since you have a shield, connect to the board ground ONLY if your board ground is referenced to earth ground. But if you are powering with an isolated supply, there will be no place for the noise to go (e.g. stray RF) but back into the power supply. Otherwise connect the shield directly to known earth ground. Also unless you have twisted TC wire and extension wire be careful to stay away from electromagnetic interference sources (motors, other electronics, fluorescent light fixtures, etc.).

 

[Mr.Cool]

you might enjoy reading:

PRACTICAL DESIGN TECHNIQUES
FOR SENSOR SIGNAL CONDITIONING

published by Analog Devices

it talks about compensating for long wire length and offset errors etc.

Mr.Cool

 

[sztibi82]

You have to use same type of cable for both terminals, and so you can connect the termocouple with 10m. As i know it is better to use Cu cable.

Only if you use 2 types of cable the length has to be mittle.