Temperature sensor and heater and how to control it.

[driga]

Dear all,

I am new in instrumentation and control so hopefully I can get some answer or hints in solving my problem.

I am planning to use Indium Tin-Oxide (ITO) as heater. I am also would like to control the heater by checking the surface temperature of the heater using the fact that the resistance of the ITO changed when temperature changes.

I have attached a simple drawing. The big area is where the ITO layer is deposited, the two smaller area is the power pads or the bus bars, when the current is lead into the ITO layer.

What I have in mind is that, as a heater a current is applied to the layer and causing heat, after a while a much smaller current is applied to check the resistivity of the ITO layer thus gaining the temperature information. This is based on my understanding of RTD sensor, that a small current is applied so that the temperature information can be obtained. Is it possible to work like that? If it is possible, how to control the heater based on the temperature reading so that I can have the desired temperature?

If it is possible, what kind of instrument that I need?

Thank you very much.

 

[davidbeach]

Why not use a four wire connection and measure resistance continuously?

 

[DRWeig]

What are the voltage and current levels? AC or DC?

Good on ya,

Goober Dave

 

[Compositepro]

Davidbeach is correct. Resistance is easier to measure with a higher current. Also temperature will be changing when turn heater on and off. Theoretically, what you propose is possible, but it is very rarely done in practice. I think this indicates that it does not work well in practice for a number of reasons.

 

[potteryshard]

On an intuitive level, it seems like having the heater circuit electrically linked to the measurement/control circuit would be inviting fail-safe issues. Could you deposit an insulated independent island within the body of the heater for measurement/control functionality?

 

[rhatcher]

If you apply a constant voltage to the heater you can measure the difference in current resulting from the changing resistance of the heater.

If you apply a constant current to the heater, you can measure the difference in voltage resulting from the changing resistance of the heater.

If the resistance/temperature curve is known, the heater temperature can be calculated from either of the above methods.

I am pretty sure this is what davidbeach meant when he said:

"Why not use a four wire connection and measure resistance continuously?"

 

[davidbeach]

Yep.

 

[driga]

First of all thank you for the reply,

@davidbeach:
What do you mean by four-wire connections? Can you explain on this or maybe give me a link so I can read it myself?

@DRWeig
I do not know yet. How does the two parameters (voltage type and its magnitude) affect my temperature measurement and heat generation?

@Compositepro
What is the practical of making a resistive heater and resistive temperature sensor simultaneously?

@rhatcher
so I need four power pads?

 

[rhatcher]

To be specific, I would say that you need two 'power' pads and two 'instrumentation' pads. The distinction is that the power pads need to carry the heater current and will be relatively large compared to the instrumentation pads that only need to measure voltage and can be relatively small (lower current).

 

[driga]

@rhatcher
Ok

what kind of instrument?
so if I measure voltage, then I need to do calculation to find out the resistance?

 

[DRWeig]

Hi driga,

I was interested in power levels you're using -- wanted to think of a simple way to measure current, perhaps with a small known resistor in one of the power feed lines.

If your processor has both voltage and current available, resistance can be calculated simply.

It can be done in a variety of ways -- many good suggestions above.

Good on ya,

Goober Dave

 

[Compositepro]

In a four wire measurement two wires carry the current that passes through your heater/resistor and the other two wires are attached to your heater/resistor at the same or different points and are only used to measure the voltage caused by passing a current through your resistance.

The practical problems are:
Sensitivity of the resistance change to temperature versus
sensitivity of the resistance to changes caused by other factors such as humidity, oxidation, wear, corrosion.
Sensitivity to electrical noise.
How do you calibrate and maintain calibration.
How do you incorporate over-temperature safety function.

An important question is how accurate does your control need to be?

 

[rhatcher]

Also, keep in mind that this will tell you the temperature of the heater itself. It will not tell you the temperature of whatever you are heating (ie. oven, parts to be cured, air, water, etc.)

 

[davidbeach]

Google Kelvin Connection

 

[driga]

@rhatcher
Please see the attached figure, what would be the different between the two configurations? Current connection is the power pads while voltage sensing is the instrumentation pad according to your term. Exactly, I would like to measure the thin film heater temperature and not the fluid above it.

@DRWeig
I think I can give the approximation, off course assuming that all the electric current is used to increase the temperature of the heater. I believe the basic equation would be: V*I² = m.C.deltaT. What is the standard use of voltage and current related to the safety?

@Compositepro
I think that the resistance sensitivity is strongly influenced by the temperature. I guess oxidation won't be a problem (is ITO corroded?), wear (I cannot answer about this). To calibrate, probably I will calibrate it once and it won't be in-situ. What do you mean by accuracy? Do you mean the temperature accuracy or how fast the response of the control system. For temperature accuracy, I would say 0.5-1 K. To decide how fast should the control response, I really do not have it in mind yet. Or do you a good way to decide on time response of a temperature control?

@david beach
I have found it in wikipedia and I am looking for other sources at the moment.

 

[danw2]

1) See green arrows on jpg. I recall something about path of least resistance. Maybe something in this scheme alters the green arrow path, but it would appear that the current takes the short path in B.

2) No answer to DRWeig's question as to voltage and current levels. Those might be ? ?

 

[ScottyUK]

Dan,

In hybrid design we would run a bus of conductor material along the edges of the resistor to avoid current crowding and localised burn-through of the resistor film. Voltage take-off points would normally be from the conductor bus.


----------------------------------

If we learn from our mistakes I'm getting a great education!