How to check the Wattage of a scrapped resistor?

[_q12x_]

 

[_q12x_]

Test finished !
Final Conclusion: the most important variable is the material used in making the resistor itself.
That material is dictating the temperature the resistor will rise to and its maximum wattage.
Unfortunately it is not a SINGLE TYPE of material used on all the --resistor watage types-- as I originally presumed. (I didnt think too hard)
Because the material varies, also the temperatures is varying from model to model and I can not obtain the Wattage if the temperature is not constant.
This only works if the temperature is --constant-- for every value type of resistor (10R,100R,1K,10K,100K,etc) and also for every watt type of resistor (1W, 10W, etc)
DataFiles

 

[Lucien Nunes]

I did mention this in post #6 - including a list of the common materials!

The material only affects the temperature rise slightly, because of different emissivities affecting the radiation per unit surface area, and the conductivity affecting the uniformity of temperature over the surface and amount of heat dissipated by the leads. The size and shape are much more important. The materials do however dictate the maximum permissible temperature which is critical.

 

[_q12x_]

This is a great handicap of electronics, not being able to calculate the Real Maximum Wattage of a given component ! Whatever it is. I feel sorry for this side of electronics.
I planned this experiment for a long time and I did some of it before, but now I push it to a complete stage even if it is a negative result, doesnt matter, I tried. Hopefully someone else will get ideas and inspiration from my exercise and who knows maybe in the future there will be a way to read the real maximum wattage of -any- given component. That will be cool.
Until then... I am doing what everyone else suggested so far, guessing and using a good sense in pointing what it may be - if is completly unknown and if possible, comparing to something that exists. Also asking others, always is a good idea, haha.
Thanks a lot for the support and chat !

 

[Lucien Nunes]

I build equipment that has to keep working under harsh conditions. Not space-probe harsh but there isn't space-probe budget either. If someone takes a unit to the desert and it overheats, it could cost tens of thousands of pounds of lost time while the backup is wheeled out. So thermal design is an important factor. In industrial plant you might solve the problem with lots of fans, but my equipment has to be near silent. You might think of using large heatsinks but it has to be portable. You might think of using heat-pumps but it has to run on limited battery power. There has to be intelligent balance between working temps, maximum permitted ambient temps, service life, MTBF etc.

If you have enjoyed thinking about heat dissipation in resistors, now look at the effects of heat on electrolytic capacitors. Get some data sheets and read up on the chemical and physical properties of electrolytics. Imagine you are building a unit with 1000 electrolytic capacitors of which 100 have significant ripple. They will be mixed up with 50 power semiconductors and 25 power resistors of different sizes and dissipations all packed into a unit that is going to be operating in the equatorial sun and must have an MTBF long enough that you don't expect any of 100 units to fail in their 5-year working lives. Suddenly you might find that the power limit on a resistor is not what the resistor can stand, but what it can be allowed to dissipate without overheating the capacitor next to it. What then? Move them apart? (wiring inductance problem). Heat-sink the resistor to the case? (assembly / disassembly problem) Overspec the capacitor? (size problem)... Now you have multi-dimensional thermal puzzle and the Arrhenius equation is your nemesis.

This is why I like my work.

 

[pc1966]

Even for ground-based equipment ESA has (or at least used to have) guidelines for de-rating* parts to achieve real-world reliability that is sufficient for serious applications.

[*] De-rating as in applying a fraction such as 0.8 of the rated power/voltage/current along with a lower max temperature such as 100C for Si devices instead of the manufacturer's 125C or whatever.