Thermocouple offest with resistor

[RTRDave]

I have a control system that only works between 70 and 100 degrees F. I want to put an inline resistor in the thremocouple leads that is variable to change the offest of the thermocouple reading so I can make my process operate in a higher range.

Any suggestions on what to use to do this? Sizing?

thanks

Dave

 

[analogkid2digitalman]

I have doubts about this. A thermocouple uses the Seeback effect to generate an low voltage based on dissimilar metals in contact with each other. Any connections made between other metals (the resistor leads, the resistor material) will also generate voltages and throw the response curve way off.

Suggest that a new control system be pursued if it is feasible. If it can be done with an off the shelf PID system, it's only a couple hundred bucks. But that is a guess since there are no details on the control system needs or specs.

Wheels within wheels / In a spiral array
A pattern so grand / And complex
Time after time / We lose sight of the way
Our causes can't see / Their effects.

 

[Skogsgurra]

I have no doubts about putting a resistor in series with a thermocouple. No doubts at all - it shall not be done. It will not work.

 

[RTRDave]

Thanks for keeping me straight....I'm off to PID it

THanks again

Dave

 

[logbook]

Well I have to disagree with the guys on this one. I don’t think it is a problem at all.

A thermocouple generates a small voltage (millivolts) and this can certainly be offset by another small voltage. Thermocouple simulators are simply resistive dividers from a voltage source. Typically the thermocouple is screwed or plugged into a connector, the temperature of which determines the "cold junction". Cold junction compensation is applied by measuring the temperature of this block. If you split one of the thermocouple leads you get an X-to-copper then copper-to-X pair of junctions, where X is the thermocouple material in one lead. Provided these two junctions are at the same temperature you do not create any thermal offset.

If you knew that one thermocouple lead was grounded at the amplifier you could put the back-off resistor in this lead and use a resistor up to the power rail to create the offset. The backoff resistor would want to be around 10 ohms, for example. The power rail resistor would be large enough to give just the right amount of backoff, perhaps a few millivolts. If both thermocouple inputs are relatively floating then you could still use the same backoff resistor but using a floating power supply such as a small Lithium battery to provide the offset. Thermocouple systems are usually not complicated and the signal bandwidth is often so low that you do not have to worry unduly about noise.

However, buying a new controller is still good advice

 

[Skogsgurra]

You are right logbook. But that is not adding a resistor, that is adding an offset and it surely needs more knowledge/experience than Dave gives the impression of having. Encouraging him to try series resistors would not help him.

 

[OperaHouse]

I was going to sell some temperature controllers on eBay. Then I scanned through the offering. You can get new digital ones for only $9.99 with the sensor. I would agree that you just can't put a resistor in series with the sensor, but I suspect you could make a voltage divider and raise it up a little.

 

[2dye4]

Hi
The reason you cannot do this is the resistor value
would depend entirely on the input impedance of the
thermocouple amplifier.
If you knew this was a linear resistive input
impedance you could calculate a resistor but it
is unlikely to be linear resistive.

more trees less bushes

 

[Warpspeed]

I agree with logbook, adding a known precisely controlled low impedance millivolt source in series is one thing. After all, cold junction correction often does exactly that.

Just sticking a series resistor in the line is a sin and a crime against nature. We all pay for our sins, one way or another in the end.

 

[OperaHouse]

I'm not saying that it shouldn't be filed under "stupid engineer tricks", but I just did the following test with my Fluke 2176A with a J thermocouple. I had a water bath that I measured at 188 degrees. With just a 220 ohm resistor across the meter input, the temperature dropped to 183 degrees. Making a 20 & 220 ohm voltage divider it dropped to 172. Of course the results are not predictable and you really have to worry about connections because they will provide wlwctro-thermal offsets. You would have to have a terminal strip with a single connector for each wire, hopefully massive enough to keep everything at the same temperature. Just holding a resistor lead will throw things way off. If you just want to bump it up 10-20 degrees it is possible with great care. I wouldn't use it in a precision situation. Any controller would have an internal span pot you could do the same thing with and then you would have to do mental math every time you read the temp. I have a home thermostat I adjuted that way in order to get around the power up setting of 68. The readout now is exactly 10 degrees higher than it actually is.

 

[buzzp]

Don't forget about the TC of the added resistor. If it is exposed to temp fluctuations close to the same fluctuations as the thermocouple this could cause problems.
Seems adding a resistor is just making a quick fix that will probably do more harm than good and not really solving the issue at hand.

 

[Warpspeed]

The thermal angel is going to get you guys.

 

[OperaHouse]

You guys have just forgot how to work without a budget!

 

[ScottI2R]

OperaHouse,
I hope your burns heal quickly from that 188 degree water bath! You can reduce your expenditures and help your buget by turning down your water heater. Hope ya find the humor in this!

Scott

In a hundred years, it isn't going to matter anyway.