Underground Temperature Sensors

[triden]

I have been working on a small project involving temperature measurement in a geothermal installation. I am using constant current sensors (AD592) that will be placed about 100 feet down the holes. I have the circuit designed and the PCB drawn up but I haven't printed it yet. I want to ask you guys about a few final details before I hook it all up.

I'm going to hook the sensors up to cat5 which will run to my measurement board adc to read the temps. The sensors will be encapsulated in a small copper pipe and then filled with potting compound to make them water proof. On the adc input I have a low pass rc filter and +/- rail voltage clamps. I am thinking of maybe putting a TVS on it too for extra protection. Since I don't want my circuit to blow up in the first lightning storm, I was thinking of using shielded cat5 cable for the run to the sensors. If I do this, what would be the best way to terminate the shield? Should I ground it on both ends? This is all kind of new to me and I want to make sure I am not missing anything.

Christan

 

[Bob S]

The AD 592 is a 2-terminal device with a constant current output of 1 uA per degree K. Do not ground either lead anywhere except at the receiver (op-amp or adc) end. Do use a copper shield to eliminate any induced voltages in circuit. The applied voltage V should be attached to the (+) terminal, and the (-) end is attached to a common ground to which the copper shield is attached. An appropriate resistor R should be in series with the (-) end of the AD592 at the receiver end to develop a sense voltage. With a + 20 volt power supply, the sense voltage should not exceed 2 or 3 volts (guess), so R < 10,000 ohms for best power supply voltage rejection. Read datasheet.

You do not want any current flowing in the copper shield, so do not ground it anywhere except to the common ground at the adc box.

Bob S

 

[triden]

Hi Bob, thanks for the reply.

I am most concerned about running the cable underground. With your advice, I will leave the foil shield in the cable floating at the sensor end and then tie it to Earth ground at the ADC end (inside the house). Does this sound correct? I also read about capacitively coupling one end of the shield to stop loops from forming. Is this something you would advise?

Chris

 

[Bob S]

I am most concerned about running the cable underground. With your advice, I will leave the foil shield in the cable floating at the sensor end and then tie it to Earth ground at the ADC end (inside the house). Does this sound correct? I also read about capacitively coupling one end of the shield to stop loops from forming. Is this something you would advise?

The main way of inducing a signal in your underground cable is by Faraday induction of ac signals. Faraday induction does induce voltage signals in loops of wires, which will be converted to current signals only if it is a conducting loop. To avoid this, do not connect the shield to the sensor, because then there will be a conducting loop which could generate common mode currents in the shield and signal wires. Do use twisted pair signal wires inside the shield to avoid differential mode Faraday induced signals. The shield is grounded to the case and house at the ADC end.

Bob S