HI

We are measuring three analog signals ( temperature) at a single duct and want to calculate the average value of the ( i.e. [1+2+3]/3 ) so I am looking for an electronic circuit through which I can calculate the average of them and generate single analog output.

The input / output details are
Input1,2,3: 4...20mA
Output1: Avg( Input1,2,3 )

 

What are you using to measure the temperatures?

 

First you would convert the 4-20mA signals into 0-x mV signal, sum the three signals using an opamp, divide by three using voltage divider, then convert the result back into 4-20mA.

 

Dear Shagas
We are using 2-wire transmitters. They are converting mV signals from thermocouple to 4..20 mA signal.

 

Dear Kubeek
can u suggest any type of converter or instrument available in market to fullfill the required pupose.

 

Kubeek already said it. You have to convert 4-20mA to voltage.
Then sum the voltage using a summing amplifier (simple op-amp configuration).
Then convert voltage back to 4-20mA if you are not able to work with the voltage.

 

Kubeek already said it. You have to convert 4-20mA to voltage.
Then sum the voltage using a summing amplifier (simple op-amp configuration).
Then convert voltage back to 4-20mA if you are not able to work with the voltage.

I'm interested as in how one would convert a few mA into a reasonable voltage.
Using 0.01ohm current sensing resistors ? or Using a BJT?

 

Dear Shagas
We are using 2-wire transmitters. They are converting mV signals from thermocouple to 4..20 mA signal.

What 2 wire transmitters?
Why don't you skip the conversion into a current signal and just use those mV signals from the thermocouple to do what Kubeek suggested?

 

I'm interested as in how one would convert a few mA into a reasonable voltage.
Using 0.01ohm current sensing resistors ? or Using a BJT?

In 4-20mA current loops it is normal to use a 250Ω resistor when you have to convert from 4-20mA into 1-5V (or 500Ω for 2-10V).

Remember that the transmitters are constant current sources and typically have pretty high compliance voltage.

The TTH300 temperature transmitter from ABB for example, can drive loads up to (Supply Voltage-11V)/22mA according to its datasheet


I agree that he can consider leaving out the transmitters and just connecting the thermocouples in parallel to get the average temperature. If the environment allows it (EMC) and he doesn't need the individual temperature measurements. And of course he would need to have a thermocouple input available.

 

If you really have current sources, then just connect all three together and then put a resistor to scale the current to the range you desire, connect the other end of the resistor to the returns for the three current sources and call that point ground.

You are doing Vo = k (I1 + I2 + I3)

 

What 2 wire transmitters?
Why don't you skip the conversion into a current signal and just use those mV signals from the thermocouple to do what Kubeek suggested?

2 wire transmitter is a tranducer converting physical value ( temperature, level, speed , etc. ) to equivalent electrical signal and having a power source of 24V DC

 

Depending on the design of the 4-20mA transmitter you could try the following.

There are two designs of transmitter where the return loop is either on the high side (+24V) or on the low side (GND).

If the return loop is GND, connect a resistor of about 220Ω, one each for each of the 4-20mA incoming signals. Bring the three resistors to a common node and connect a 33.2Ω resistor from that node to the common return loop (GND) of all three transmitters.

Now you can read off the average current by connecting a volt meter across the 33.2Ω resistor and measuring the voltage. A reading of 400mV would be equivalent to 4mA average.

If you want to be precise, use three 100Ω 1% resistors all connected in parallel in place of the 33.2Ω resistor. Mind you there is no appreciable difference between a single 33.2Ω 1% resistor and three 100Ω 1% resistors in parallel.