How to estimate power rating of unknown resistors

[TheUnnamedNewbie]

Say I have a resistor of unknown power rating. How would I go by guesstimating what rating it has?

In my specific case, I have many boxes of them, some examples in the picture below (with some reference resistors of known powerrating). Any tips, ideas, etc? Ofcourse, I could just put some power through and see what it takes to get them to certain temperature, but I was wondering if there was a better way? In addition, this technique is not possible when one is doing a repair and a powerresistor has failed, and the rating is unknown.

[schmitt trigger]

As a pure estimate,
I would say G and F are 1/2W, E and A are 1W, B and C 2W, and D at least 3W.

Part of my estimate is that these are 5 and 10% tolerance units, usually made of carbon film.
CF have a lower thermal rating than MF, which required slightly larger body sizes for a given power rating.

[bjcuizon]

How 'bout referring to some chart showing their size. Perhaps something like this one:

[b_force]

Power ratings are a bit of a weird thing.
I have seen SMD resistor rated with certain power dissipation (by respected brands) that are practically just BS.

All these unknown resistors look rather old, so you need someone with enough experience (= old enough) to recognize them.
I recognize A,B,C and E from old tube amps and keyboards etc.
Mostly around 5W, but sometimes also rated as 3W or 2W.

[Zero999]

Other than reading the data sheet, the only way to know for sure is to test them, which will be destructive. Apply increasingly more power, until either the value falls outside of the tolerance band or the component starts to physically deteriorate i.e: smoke, discolour or melt.

[b_force]

Other than reading the data sheet, the only way to know for sure is to test them, which will be destructive. Apply increasingly more power, until either the value falls outside of the tolerance band or the component starts to physically deteriorate i.e: smoke, discolour or melt.

And this is where the trouble starts.
How are you gonna test them?
DC current, AC current, pulsed, if so, what kind of pulse?
If DC current for how long?

I have seen 500mW resistors that can easily do 2-3W pulsed under de right conditions.
Yet, they were being rated as 1W resistors.
I have also seen 5W resistors that will burn down at 5.5W in a couple of seconds.
I have also seen 5W resistors easily dissipating around 8W for half an hour.

[Mechatrommer]

..or the component starts to physically deteriorate i.e: smoke, discolour or melt.

when apply the rated power to the resistor, it will warm up a little bit but not up to discolorization smoke. discolor or smoke is clear indication the resistor is badly overpowered. i would say a test of temperature raise from room temp to 60-80°C in 10 sec (or until temp stabilization for bigger resistor) will guesstimate the resistor power rating... for ac current application, we use rms value... ymmv..

[bjcuizon]

All these unknown resistors look rather old, so you need someone with enough experience (= old enough) to recognize them.
I recognize A,B,C and E from old tube amps and keyboards etc.
Mostly around 5W, but sometimes also rated as 3W or 2W.

I'm sure Mr. Carlson knows how to determine the specs of these old resistors. But I wonder if he has a video about this.

[Brumby]

You can't go on size alone with a high degree of certainty.  Advances in resistor design have allowed smaller sizes for a given power rating.

The thermodynamics has not changed - the same amount of power dissipation will occur for a given resistance, but with a smaller body, the temperature attained will be higher.  For those resistors, the materials used in their construction can withstand these higher temperatures.

To try and deduce the power rating without a data sheet will be somewhat of a challenge, especially if you have a box of random resistors collected over the years.  You would need to identify the type of resistor and have a reasonable idea of which era it came from ... and having worked with them in the past being the best chance of getting that right.

 ... or asking someone who has.  (But if you have a lot of resistors, this could really stretch the friendship.)

[IanMacdonald]

We used to joke that a well-designed resistor is one which under overload either goes o/c and turns white, or goes s/c and turns black. 

At least that way, you can still read the value correctly.

[mikeselectricstuff]

Power ratings are down to basic physics of convection losses, so primarily dependent on surface area, so comparing the size to known parts will get you pretty close.

[tszaboo]

Other than reading the data sheet, the only way to know for sure is to test them, which will be destructive. Apply increasingly more power, until either the value falls outside of the tolerance band or the component starts to physically deteriorate i.e: smoke, discolour or melt.

I was once testing fusible resistors. I wanted too break them, test how far can I overload them, when do they fuse. At around 1000%, it started glowing yellow, and burning the desk, and the damn thing was still almost within spec after that.

[Zero999]

Other than reading the data sheet, the only way to know for sure is to test them, which will be destructive. Apply increasingly more power, until either the value falls outside of the tolerance band or the component starts to physically deteriorate i.e: smoke, discolour or melt.

And this is where the trouble starts.
How are you gonna test them?
DC current, AC current, pulsed, if so, what kind of pulse?
If DC current for how long?

DC or AC with a higher frequency than the thermal time constant of course.

I have seen 500mW resistors that can easily do 2-3W pulsed under de right conditions.
Yet, they were being rated as 1W resistors.
I have also seen 5W resistors that will burn down at 5.5W in a couple of seconds.
I have also seen 5W resistors easily dissipating around 8W for half an hour.

Yes, resistors will tolerate higher power for short periods and are often rated very generously.

Power ratings are down to basic physics of convection losses, so primarily dependent on surface area, so comparing the size to known parts will get you pretty close.

The temperature rating of the material is also important. A ceramic wire wound resistor will be able to tolerate a much higher temperature, than a carbon film one, with an epoxy case.

Other than reading the data sheet, the only way to know for sure is to test them, which will be destructive. Apply increasingly more power, until either the value falls outside of the tolerance band or the component starts to physically deteriorate i.e: smoke, discolour or melt.

I was once testing fusible resistors. I wanted too break them, test how far can I overload them, when do they fuse. At around 1000%, it started glowing yellow, and burning the desk, and the damn thing was still almost within spec after that.

That's bad. Fusible resistors are supposed to fail safely, when the specified power dissipation is exceeded.

[Gyro]

Other than reading the data sheet, the only way to know for sure is to test them, which will be destructive. Apply increasingly more power, until either the value falls outside of the tolerance band or the component starts to physically deteriorate i.e: smoke, discolour or melt.

I was once testing fusible resistors. I wanted too break them, test how far can I overload them, when do they fuse. At around 1000%, it started glowing yellow, and burning the desk, and the damn thing was still almost within spec after that.

Were they Fusible or Flameproof? I'm not sure if the terms are interchangeable off-hand.

[tszaboo]

Other than reading the data sheet, the only way to know for sure is to test them, which will be destructive. Apply increasingly more power, until either the value falls outside of the tolerance band or the component starts to physically deteriorate i.e: smoke, discolour or melt.

I was once testing fusible resistors. I wanted too break them, test how far can I overload them, when do they fuse. At around 1000%, it started glowing yellow, and burning the desk, and the damn thing was still almost within spec after that.

That's bad. Fusible resistors are supposed to fail safely, when the specified power dissipation is exceeded.

I thought also. I think it was TT EMC series. 10 Ohm for sure, and 2 is or maybe 5W.
I'm not sure, it was a few years ago. Now there seems to be a much better datasheet.
http://www.farnell.com/datasheets/1851678.pdf
In any case, I was just testing it for worst case, and then pushed it a little bit. But it was still working at 10x the rated power. And if you check the "Fusing Characteristic" chart, a 2W takes 60W for quite some time, before breaking. So I just want to point out, that applying some power, and expecting to fail is a bad way to determine a resistors power rating.