If I buy an MMBT4401 from Diodes, Inc., I get a 40V, 0.6A NPN transistor in a SOT-23 package. If I buy an MMBT4401 from ON semiconductor, I get a 40V, 0.6A NPN transistor in a SOT-23 package. If I buy an MMBT4401 from Micro Commercial Co,… you get the idea. And this is hardly the only case of this; countless companies make 1N400x diodes, 2N7000 NFETs, 2N3904/MMBT3904 transistors…
I'm quite certain that these companies don't all share the exact same silicon (though some of them might be packaging and reselling silicon dice bought from the same source), so what exactly is standardized between them? Is there even a de jure standard (and if there is, who sets it), or is this all just a de facto standard?
In either case, which specific figures of merit are considered "part of the standard"? Can I trust that no 1N4001 will have a forward voltage greater than some standard value, for instance? That one I'd be pretty sure is part of the standard, but what about the parasitic series resistance? I know I wouldn't want a 1N4001 if it had a parasitic resistance of 10 ohms, but could such a diode be made and still bear the mark 1N4001?
Best Answer
A very few components are actually standardized by JEDEC (think of the 7400 logic series, for example).
Most of the components you mention are simply results of historic multi-sourcing agreements: Large customers (esp. of the military kind) wouldn't buy an obscure part that they could only get from one party.
Also, don't underestimate the history of these companies: for example, TI having all the parts that are still available in stock under the brand name of National Semi happens because the former bought the latter, and some stocks simply have a half-life of roughly eternity. Mergers and spin-offs sometimes leave multiple parties with access to the same IP.
To answer about guarantees: Um, these usually don't exist. Good luck out there and read your datasheets!