Calculating resistor heat (Degrees C)

[kalel]

What is the simplest way to calculate how much a resistor will heat up?

E.g. 1w, 2w tells us the power dissipation, but whether that means 1 degrees C above ambient or 20 on the body of the resistor, depends surely on the area of the resistor itself.

If there's an online calculator that does this - all the better, as I'm not one with Math (one mistake was that I didn't understand it as something practical at school - which it is, because it allows us to make practical things). But, if there's a simple formula, that should work too (you shouldn't need to be proficient to use formulas, as long as you can read them and/or put them into a program or a calculator).

There's no specific reason (e.g. ensuring that the heat caused by resistors in circuit will not affect other elements), as likely this only needs to be calculated for high quality PCB design and such, none of which I'm doing.
I'm just curious about roughly how hot a resistor will get at certain wattage.

[halexa]

Resistor are not usually specified with junction to ambient power dissipation data.
What is needed for resistors is the de-rating of the power dissipation of the resistor at a given ambient temperature.

But i found a App note from Vishay regarding this.

https://www.vishay.com/docs/53048/pprachp.pdf

[MosherIV]

Well done on your design diligence.

I believe (because I am not a practicing electronics engineer) that you make sure the power developed is less than the power rating of the resistor rather than worry about how hot it gets.

[kalel]

Well done on your design diligence.

I believe (because I am not a practicing electronics engineer) that you make sure the power developed is less than the power rating of the resistor rather than worry about how hot it gets.

You're right, and I actually wanted to say that I'm not doing a design where such a thing is important (exact degrees), but just wondering about a rough estimation for pure curiosity reasons. I could measure the temperature of a resistor, but if there's already a way to calculate it, that might be better.

[MagicSmoker]

I'm just curious about roughly how hot a resistor will get at certain wattage.

You'll have to check the manufacturer's datasheet, then; the surface temperature of the resistor at rated dissipation and a specific ambient temperature will usually be given. It is not unusual for cement wirewound types to reach 250C at rated dissipation and a 50C ambient, btw.

[kalel]

I'm just curious about roughly how hot a resistor will get at certain wattage.

You'll have to check the manufacturer's datasheet, then; the surface temperature of the resistor at rated dissipation and a specific ambient temperature will usually be given. It is not unusual for cement wirewound types to reach 250C at rated dissipation and a 50C ambient, btw.

Is a K-Type thermocouple reasonable enough to measure the temperature?

[alm]

Datasheets for some power resistor will specify both temperature rise for the resistor body and the lead at the PCB (for some specific lead length) from 0 to 100% load (rated power). Look at page 56 of this Panasonic data sheet for example. You could use it to estimate thermal resistance. The thermal resistance will probably be similar for resistors of similar dimension and materials (e.g. other axial metal oxide resistors with the same rated power).

This data can be useful if you want to ensure that something close to the resistor (like the PCB) does not get too hot. That resistor may survive 250°C, but will your circuit board? It is not uncommon to see circuit board with darkened spots under the power resistors. To solve this they could either have used longer leads, a larger resistor (lower thermal resistance to ambient) or leads with a higher thermal resistance (like the Panasonic ERGF series with plated iron leads).

[MagicSmoker]

Is a K-Type thermocouple reasonable enough to measure the temperature?

If it's an exposed bead thermocouple - with a high temperature rated insulation like glass wool or kapton or the like - then, sure; a K thermocouple is good for up to 1300C or so, after all.

[kalel]

This data can be useful if you want to ensure that something close to the resistor (like the PCB) does not get too hot. That resistor may survive 250°C, but will your circuit board? It is not uncommon to see circuit board with darkened spots under the power resistors.

I watched a video that mentioned that you should mount the resistors which dissipate a noticable amount of power with a little distance from the board, to protect other components. Seems reasonable when possible (unless there are space considerations).

[Brumby]

I watched a video that mentioned that you should mount the resistors which dissipate a noticable amount of power with a little distance from the board, to protect other components. Seems reasonable when possible (unless there are space considerations).

The first thing an increased stand-off distance protects is the PCB.  Also the solder joints.  A longer lead provides more air between the resistor body and the PCB, which allows for greater cooling of the resistor as well as reducing the radiant heat transmitted to the PCB.  The longer leads also provide a longer path for heat to travel down them, also reducing the heat conducted to the solder joints and the PCB.  The last thing you will want is a power resistor heating the solder to over 200ºC.  Things are not going to last.

Protecting surrounding components from heat is a simple exercise of leaving enough space around the heat generating component.

[SeanB]

Just also consider both orientation, airflow and case design. A resistor operating at 200C above ambient ( a lot are specified at this at rated current for larger resistors) will need a flow of air, either forced cooling or simple convection, and will also be orientation dependant as well. For small resistors a good rule is to remember the body will run at 100C above ambient in most cases at full power, so if you are doing high dissipation relative to the resistor rating this will be a consideration, and if you are running any vitreous enamelled power resistors also remember that at more than around half power the glass or ceramic case is no longer regarded as being insulating, as it will start to become conductive. When running them with high voltages keep this in mind, you have to provide extra insulation for the mountings, or isolate them further from grounded cases if you do not want to have leakage currents to chassis. That is why you often see power resistors with separate mounting frames on a PCB, with the frame separately mounted to the board with each resistor.

[CatalinaWOW]

As the prior two post pointed out there are many, many variables involved in the actual temperature of a resistor.  Another is other sources of heat.  Two adjacent resistors will run hotter than one off by itself.

Thermal modelling software is available which can provide reasonable estimates of the temperature, but it is not the kind of thing that would work well as an on-line calculator.  They typically do a solid model of the whole board and enclosure, break it up into small elements and then solve the appropriate differential equations by iteration.  It is a lot of work and requires good judgement in how properties of each volume element are assigned and how the boundary conditions are assigned.  As with any model a certain amount of skepticism about the outputs is required.

If you really need the answer build a prototype board and use a thermal camera to observe the temperatures in operation.  Be sure to account for the worst case actual operating environment, not just the conditions in your lab.

[TimFox]

Be careful with data-sheet reading:  many resistors have different power ratings depending on the allowable temperature rise or ambient temperature.  Military ratings are generally conservative, but the same part may have a higher power "commercial" rating.