It's getting worse - there are many "mini" or "supermini" resistors being offered by resistor manufacturers nowadays that are typically 1/2 to 1/4 the size of what I am used to.
Last night I went to buy "1/4W" through-hole resistors for breadboarding, the 6-7mm wide parts I've used for decades.
I found "Metal Film Resistors - Through Hole .4watt 1% 1/8watt body size"
That is the 1/8W body size MBA0204 rated "standard mode" (0.25W) part max. is 125°C, "power mode" (0.4W) is 155°C.
Going upsize to the
MBB0207 "standard mode" (0.4W) part max. is 125°C, "power mode" (0.6W) is 155°C. It's the size I'm expecting. At 1/4W it's temp rise is 35°C and old school would run it at 1/8W for temp rise 15°C.
Part of the new game and what has changed is:
-Higher temp materials (but who cares, nobody runs them that hot anyway) giving inflated power dissipation numbers that must compete with other spec stretchers.
-Thinner/thicker leads, and copper verses steel construction can give misleading graphs
-Specs that are for forced-air cooling use, in a fan-cooled (i.e. power supply) enclosure
So my old "1/4W" parts are rated 0.6W nowadays.
Most manufacturers are not giving thermal resistance or temperature rise data for their parts.
I commend Vishay for doing this, and curse Yageo for their game of omitting the spec and claiming a 1/4W sized part is "1W" FFS.
I did get a response from Vishay for
this thread about the test jig used for the temp rise test. They (i.e. "0.4W") are the absolute max surface temperature (say 155°C) in 70°C ambient in the jig, 56 day run?. But I did not get it all figured out yet because many tests/specs are "common knowledge" in the resistor industry but never taught, to withstand the flames from people.
Dave could do a better investigation to help with the traps in choosing resistors.
SMT resistors are other whole set of traps, I just read how 0603's have a lower thermal resistance than 0805's due to their aspect ratio. Who'd have thought?