Temperature rise calculation for electrical devices

[srinaath]

There are 2 types of temp rise measurement for electrical devices,
1)resistance type
2) thermocouple type (direct measurement)

My query is
1) which method is superior and please explain me the reason.
2) Which gives higher value of measurement?

Thanks in advance.

 

[gleem]

To clarify your question, are you asking what devices are appropriate for measuring different temperature ranges?

 

[Borek]

Calculation or measurement, as your post doesn't agree with the thread title?

 

[russ_watters]

Easily googlable:
http://www.surecontrols.com/rtd-vs-thermocouple/

...though the question "which is better?" has two answers...

 

[srinaath]

To clarify your question, are you asking what devices are appropriate for measuring different temperature ranges?

in accordance with UL standard, there are two methods to detect temperature rise of a device,

1) resistance method: (involves detecting resistance of the coil)

2) thermocouple method: (directly measuring the temp using sensor)

I just wanted to know which method would be preferable and why is it so? or is there any advantage of one method over the other in terms of accuracy...

 

[russ_watters]

I just wanted to know which method would be preferable and why is it so? or is there any advantage of one method over the other in terms of accuracy...

Please read the link I provided, do some of your owwn research and come back with more focused questions (after you do the research, the questions you should be asking should be apparant).

 

[jim hardy]

I just wanted to know which method would be preferable and why is it so? or is there any advantage of one method over the other in terms of accuracy...

I think you'll find this a good introduction.
https://www.omega.com/temperature/Z/pdf/z021-032.pdf

"Which method is preferable" depends on preferences (and prejudices) of whoever is doing the job and of course what it is he's measuring. . You'd measure a kiln differently than the air stream in an automobile engine intake..

Beware the allure of the esoteric .
Temperature is defined in the International Practical Temperature Scale by reference to platinum resistance thermometer. See http://nvlpubs.nist.gov/nistpubs/jres/77A/jresv77An3p309_A1b.pdf
This naturally inclines one to think that a platinum RTD is endowed with superior precision.
Actually, to achieve high precision with either a thermocouple or an RTD you will have to calibrate it against a temperature standard.

Take a close look at that nist publication... first page

The thermometer resistor must be annealed pure " platinum, supported in a "strain-free" manner and have a value of W at 100°C not less than 1.39250.

It's fabricated of very delicate wire that must survive getting wound onto a mandrel and having lead wires affixed, then insertion into a sheath. .
A thermocouple on the other hand is two wires usually welded together. They need not be delicate.

Your computer will have to do a curve fit to convert the signal from either a thermocouple or a RTD into temperature. In my day we used linear interpolation in lookup tables of desired accuracy because it was faster than solving quadratics in floating point. With today's equipment i believe that's no longer a problem.
I used thermocouples to measure our feedwater temperature for reactor power calorimetric. I never regretted it. They were wired straight into the ADC so there were no 'knobs' that anybody could mis-adjust. I believe they're still using the same ones i installed in 1973.

old jim

 

[Asymptotic]

in accordance with UL standard, there are two methods to detect temperature rise of a device,

1) resistance method: (involves detecting resistance of the coil)
2) thermocouple method: (directly measuring the temp using sensor)

I just wanted to know which method would be preferable and why is it so? or is there any advantage of one method over the other in terms of accuracy...

Which UL standard are you referring to?

"Device" seems rather too broad a term, because "detecting resistance of the coil" suggests solenoids and motor windings. Provided you are referring to motors, read IEEE standard 112, "IEEE Standard Test Procedure for Polyphase Induction Motors and Generators" with a focus on section 4.4 "Temperature" and section 5.8, "Temperature Test".

Which technique is preferable? Depends on whether temperature sensors have already been installed (and where), size of the electrical machine, and what exactly it is you are trying to find out.

 

[srinaath]

Which UL standard are you referring to?

"Device" seems rather too broad a term, because "detecting resistance of the coil" suggests solenoids and motor windings. Provided you are referring to motors, read IEEE standard 112, "IEEE Standard Test Procedure for Polyphase Induction Motors and Generators" with a focus on section 4.4 "Temperature" and section 5.8, "Temperature Test".

Which technique is preferable? Depends on whether temperature sensors have already been installed (and where), size of the electrical machine, and what exactly it is you are trying to find out.

UL508
I am trying to find the temp rise of a transformer

 

[Asymptotic]

I don't know the specifics of UL 508 transformer temperature rise testing, but odds are it is similar to the procedures in IEEE 112. It may be I'm underthinking this, but temperature rise ought to be on the transformer nameplate. Let's say it isn't, and measurement is necessary.

My query is
1) which method is superior and please explain me the reason.
2) Which gives higher value of measurement?

Neither method is universally superior - it still depends on what you are trying to find out.

How do you mean "Which gives a higher value of measurement"? Which technique yields a numerically larger value? Or in the sense of which measurement technique gives results more closely aligned with actual temperature changes?

Broadly speaking, the 'resistance change' technique isn't suitable for determining the maximum temperature rise that occurs after one or more bolted fault trials. Temperature distribution isn't uniform - winding temperature is driven higher deep within the winding because the cooling medium and heat sinking can't get at it. Resistance change gives the average, under-represents this hot spot temperature, and direct measurement may be the only recourse. Problem is, the only way to get a temperature probe to the hot spot is to build it into the winding, and have some foreknowledge of where the hot spot temperature is located.

If you want to learn temperature rise at rated current over a reasonably long period of time then both methods have their pros and cons.

How large is the transformer in question? It is difficult to answer to your questions otherwise.

Thermal time constant is larger for large transformers than it is for small ones. Pulse transformers in an SCR gating application are tiny, and (unless an IR sensor is used to measure core temperature) whatever temperature probe is applied will be a significant fraction of the transformer's mass, act as a heat sink, and affect the measurement. By the same token, it doesn't take very long for such a small mass to cool off, and resistance must be measured immediately after the pulse generator is turned off. These things aren't as much of an issue, for instance, in a 50 kVA dry transformer.

 

[Kickshaw]

Interestingly the old UL motor standard 1004-1 allows for a higher max reading/calculated temperature if the rise by resistance method is used, when compared to TCs. This as I understand it through a few conversations is because UL assumes the hottest spot in the winding is in a slot. They do not believe TCs buried in end-turns capture the slot temperature and therefore require the max TC reading to be reduced. If that is the case then rise by resistance is seen as true. The problem becomes taking resistance readings while the winding is energized. That is not problem for TCs and allows for faster testing.