Hello all,

I am new to this site but not new to the topic. However there is something I am dealing with that I don't have a lot of expertise in. I have a need to control air temperature in a room. I have installed a 3HP ventilation fan and use a VFD to control fan speed. I am trying to hook a thermocouple up to the VFD analog signal input. I have done this is the past using a 10 KOhm variable resistor. I wired it with 3 wires. 12V+, Common as well as the analog signal input to the VFD. For my thermocouple attempt, I am using a 2 wire Type K thermocouple. Does anyone have any experience with this or am I trying to do something that is impossible. If it's not impossible, how would I hook up the 2 wire sensor to the 3 terminals on the VFD?

Thanks in advance to anyone that offers any assistance.

Cheers,
Carl

 

You need an amplifier of some sort often called a signal conditioner in automation speak. It would have the cold junction reference. etc to convert TCs to a usable signal for what you want. A thermistor setup might be a lot simpler for what you are looking for but it would still be easiest to have an active device in between.

Regards
cww

 

Carl,

I have not seen a VFD that will accept a thermocouple directly, the inputs are usually Voltage or mA.

The input you describe is 0 - 12 Volts, a type "K" thermocouple measuring ambient temperature will only produce 1.2 mV @ 30 C, see link below.
http://instrumentation-central.com/TechNotes/TypeKTableC.pdf

I don't think a thermocouple is the right choice for ambient temperature other types of sensors are much more sensitive, e.g. RTD or thermistor.
You have many options depending on how accurate you need to control the room temperature. Here are just a few:
1. Use a thermostat to speed up the drive (High speed/Low speed)

2. If the drive has internal PID control it likely has an analog input that will accept 4-20 mA. Use a signal converter to boost the sensor signal, e.g 4mA @ 10 C, 20 mA @ 40 C

3. If the drive doesn't have internal PID purchase a dedicated one that will accept the sensor directly. The controller will need to have an output that the drive can accept, either 4-20 mA or suitable voltage range for the speed set-point you describe. 0-20 mA would work if you provide a suitable shunt resistor (500 Ohm for 10 V, 600 Ohm for 12 V). If you have a local vendor of commercial air conditioning systems they may have a good solution to match your drive.

4. Post some more information, e.g. VFD model, Local or remote control, etc. so we can help you further.

Hope this helps,

Roy

 

The thermocouple will not output enough voltage or current for you to drive the input of the VFD directly... certainly none of the VFDs I have come across. Depending on the input that is available on the VFD, typically either 0-10v or 4-20ma or both, you could use a converter module to convert the thermocouple signal to whatever is appropriate for the VFD. Most of the newer VFDs have some kind of PID type control loop available built in and some even have routines which are specific for fan control.

Gerald Beaudoin

 

Carl

I'm not aware of inverters with mV analog inputs.
Buy a cheap temperature controller, thermocouple input, and an analog output. It will send a "standard" signal (4-20mA, 0-10V) to the inverter.
The controller may have more resources than the standard PID found in inverters. (We're not talking about Simovert, right?) You may need it when dealing with:
. Thermal load changes
. Outside temperature changes
. Speed-flow nonlinearities or changes

It may not be easy to get a stable operation.

Regards,

Marcos

 

Thanks to all for your replies. I have done a little more research and it does coincide with all of your advice. I will look for a singal convertor/conditioner to place between my thermocouple and invertor. Yes, the invertor has a 0-10V or a 4-20ma analog input, without any PID control.

I understood that thermocouples vary the resistance during temperature changes. I was using the change in resistance to give me either a change in current or change in voltage. I actually did manage to get it working. I hade a freq. of 26Hz at 31 degrees and it went up to 33Hz at 32 degrees. so it looks like I was somewhat succesful but I am not sure if I can get more range out of it. I am going to try programming the parameters to try and calibrate my frequencies with the anaolog input values to see if I can get a range of 10Hz at 32 degrees and 60Hz after 40 degrees. (Proportional in between). I think there is gain or scale on the analog input.

If that fails then I will look for a good programmable device for in between...or an IC temperature transducer.

Thanks again to all those that replied.

Cheers,
Carl

 

Carl,
Maybe your circuit doesn't work well because you just do not understand the basics of thermocouples. They may exhibit some small resistance change with temperature, but that is incidental. A thermocouple (TC) generates a tiny emf, a voltage potential, that varies with temperature.

Actually, a TC is not designed to carry any current since at least one of the metals of which it is composed may be a poor conductor. The voltage potential exists in a circuit having a "hot" end and a "cold" end, or junction. There is no voltage potential when both ends/junctions are at the same temperature. The principal is for you to know the temperature of one end Hot or Cold.

Then the voltage potential generated at the other junction is proportional (as determined from the thermocouple curve) to the DIFFERENCE between the temperatures of the two junctions. The voltage potential is usually measured by a very high impedance circuit that does not allow much current to flow. Now you can use this voltage potential to control your fan speed, although you probably will want to use a commercial thermocouple instrument to supply a standard output signal, rather than building a custom circuit.

Google Thermocouples to find out more.

Dick Caro
===========================================
Richard H. Caro, Certified Automation Professional, CEO, CMC Associates,
2 Beth Circle, Acton, MA 01720
Tel: +1.978.635.9449
Mobile: +.978.764.4728
Fax: +1.978.246.1270
E-mail: [email protected]
Web: http://www.CMC.us

 

Carl, No, thermocouples don't change resistance they are low resistance devices but they produce a mV signal that a multimeter may miss-interpret. You may be thinking of a thermistor, a solid state device which does change resistance. One of those would be more sensitive than a thermocouple.

For example - http://www.ussensor.com/rt charts/502FG1K.htm Check out the one with 3420 curve. The problem with wiring the device straight to the drive it will not be able to correct for an offset in temperature. You should be able to purchase a small PID controller for < $300 that will accept a thermocouple and give you a 4-20 mA signal to the drive.

Roy

 

Hi Carl
Just as a note, rtds and thermistors change resistance with temperature as their effective variable. While a thermocouple will change resistance slightly because nearly all metals change resistance with temperature, that would not be what you use a TC for. I suspect you are confusing devices. The semiconductor sensor is an excellent idea it you can use them as they are linear and have a reasonably high level output. An lm35 or lm36 and a common op amp would get you the signal you need for sure and the circuit is probably right in the book.

Regards
cww

 

Carl

IF you use a digital input, just use a cheap house thermostat.

 

> Yes, the invertor has a 0-10V or a 4-20ma analog input, without any PID control.

Since you need a PID control purchase a small controller with 4-20 mA output and thermocouple or better still thermistor or RTD input.