Thermocouple splitter

[cntrlfrk]

Does anyone know of a device I could use that would take a thermocouple signal and split it into two seperate signals? Preferably back to a thermocouple and 4-20mA output? I currently have a skid mounted chiller with it's own microprocessor, but would like to have the ability to monitor and record the temperatures on our plant-wide control system, thus the need to maintain the thermocouple signal to the existing microprocessor, while being able to bring a corresponding 4-20mA signal back to a PLC.

Thanks for any help.

 

[SNK]

yes, there are many splitters on the market.
I have in front of me an INOR 3-wire RTD splitter, P/N MINI-PAQ-L, or Loop powered isolators (2-4-20mA outputs for 1 in) with P/N ISOPAQ-21L.
Website is www.inor.com.

They work well. The other option is to split your current 4-20mA signal and put it in series with your PLC analog input. (Just make sure that you have isolation).

 

[cntrlfrk]

Thanks for the input.

I have used several different types of splitters and signal conditioners over the years.

I guess the uniqueness of what I am trying to do here is to have a device with a thermocouple input with one thermocouple output and one 4-20mA output.

I don't recall having ever seen a splitter with a thermocouple output as an option.

I don't believe I can wire a thermocouple in 'series' through a signal conditioner then into the microprocessor to get the two signals I need.

 

[paulcs]

Since you are using thermocouples, it might be simple and very low cost to use a dual element T/C. One to the local control and the other to the PLC either directly or via a 4-20 output transmitter. Using 2 T/Cs provides some amount of system redundancy (even though they will never match exactly).

 

[rsdoran]

Since the device the thermocouple goes to has a microprocessor, is it possible there may be a way to get information from it, using an output or communicaions? Many devices, like temp indicators etc, have analog, serial, and/or other capabilities, if what you are using doesnt it may be easier to change it to a device that does.

 

[cntrlfrk]

Since the device the thermocouple goes to has a microprocessor, is it possible there may be a way to get information from it, using an output or communicaions?

That is what I am also looking into. This is a 'Carrier' chiller, with a 3200MP control microprocessor, and is an older unit. From what I've dealt with on Carrier in the past, they have their own proprietary communications. I am attempting to contact the correct person to find out if they have some type of gateway or standard communication link I can tap into.

 

[Mickey]

Wire the devices in parallel across the thermocouple. If they don't present a big load to the thermocouple it will work. I have done it,
worth a try.

 

[danw]

If I understand it correctly, you need
- a thermocuople input to a transmitter (which outputs 4-20mA)
- a separate thermocouple signal for something else.
both inputs coming from a single monitoring point that has a thermocouple installed.

If that's the case, all you need to do is to connect the T/C to one device and parallel another thermcouple wire (+ to +, - to -) of the same type from that T/C input to the other.

It is an accepted practice in the heat treat biz to parallel a T/C to both a temperature controller and a recorder.

There are two difficulties that can be encountered

a) If the thermocouple is grounded and inputs are not floating you can get ground loop(s). That can be resolved with dual thermocouple elements, with one element for one device, one element for the other.

b) Many devices have an "open thermocouple" or "thermocouple break" detection circuit. Sometimes those circuits conflict with one another, causing offsets or spikes when connect them with a common T/C. The solution, again, is dual thermocouple elements, with one element for one device, one for the other.

Paralleling thermocouples also assumes that the connection blocks or terminal blocks are isothermal, meaning the + and minus erminals on any given device are at the same temperature, otherwise error is introduced.

Dan

 

[rsdoran]

Wow I did something like this couple years ago with a Carrier Chiller, what model? They have add on modules that may allow what you need to do. Is it using thermocouples or thermistors?

 

[cntrlfrk]

Wow I did something like this couple years ago with a Carrier Chiller, what model? They have add on modules that may allow what you need to do. Is it using thermocouples or thermistors?

This is a Hermetic Centrifugal Water Chiller, model 19DHT898CQ with the aforementioned 3200MP/ ESP-II controller.

I've spoken to 6 different people at Carrier, and they are quoting some type of communications gateway. They have expressed that this is very costly, so it may end up easier to install new sensors in redundancy to the others being used by the controller.

 

[cntrlfrk]

As an update to this, I did recieve a quote from Carrier (a mere $6000) on a communications gateway for converting the Carrier comms to BACnet or Modbus RTU. I am not familiar with BACnet or if it could communicate with an A-B ControlLogix processor, or if I would be better off to just go from Modbus to ethernet and access it that way.

Also, I was wrong on my original information on the temperature sensors, they are thermistors, not thermocouples.

The most likely result in this will be to install our own RTD's to monitor the temperatures, and a relay to start and stop the unit as needed via the PLC, and possibly using an external alarm relay to provide notification of a problem.

 

[rsdoran]

I thought it might be thermistors, also appears that chiller has been updated, that model I do not think originally had a microcontroller.

Since it does have one it may offer alarm outputs that represent different types of faults etc, without the manual it is hard to say.

Just depends on your needs how you handle it, usually the chiller when it has a problem will fault and shut itself down so it may not be that necessary to monitor it that closely. You may be able to just monitor its status thru its output relays then act accordingly.

 

[MikeW]

..... I am not familiar with BACnet or if it could communicate with an A-B ControlLogix processor, ....

BACnet is the Air Conditioning industry method to provide open communications between various vendor's equipment. Based on an ASHRAE standard (135?).

Never had any direct use/ contact with it, so I'm not going to be much more help than just a little info. For more info:http://en.wikipedia.org/wiki/BACnet

 

[fish]

Hello, lots of good info on this site, I usually find a solution here.
I am also in need of a retransmitter, in my case, I need to read a thermocouple, and retransmit 2 thermocouple signals. One to a PLC, the other to a limit controller. I could use two seperate thermocouples, but I found another solution:
http://www.apcs.net.au/products/standard/tcs726.html
Sent an email requesting a quote today. Hopefully these are obtainable. Hope this helps.

 

[Lancie1]

Another way is if you have a spare analog output on the PLC, then you could run the TC to the PLC, use it internally but also retransmit it as an analog signal to the controller.

Anoter way as Mickey mentioned is because a thermocouple produces a volage (milivolts), it can be sent to 2 load devices wired in parallel, as long as it has the power capacity to drive both. If your TC runs are short (50 feet) with heavier-gage T/C wire (#18 or larger), then you should be okay to run both devices. The longer and smaller is your T/C extension wire, the less likely you will be able to parallel two devices.