[sdiy] 2N vs PN vs MPS, etc

[Paul Schreiber]

Back in the "good old days" (when my TRS-80 booted in 2 seconds), there was 
a government group called JDEC, that hat a transistor spec called JANTX (I 
think that meant Joint Army-Navy Transistor Spec).

There were literally 10s of thousands of transistors available. So, what the 
"industry" did is register the spec, and these guys were the "keepers of the 
spec". They assigned the '2N' numbers, so that a 2N2222 from Fairchild was 
*identical* to a 2N2222 from Motorola. Simple enough.

Then, both Motorola and Fairchild started to get 'cute'. They decided to 
market "equivalent" parts. Not the SAME part: an 'equivalent' part. What the 
hell did THAT mean? It meant "We don't want to pay the JAN commitee to keep 
the spec". So, they mape parts like this:

2N2222A: JAN part, it will cost you 8 cents
MPS2222A our 'equivalent' part, it is only 6 cents!!!!

Then TI jumped in with TIS parts (like the TIS97 used in RA Minimoogs). It 
was a nightmare for designers.

PN is what Fairchild callled their non-JAN FETs. In most cases, it is the 
*SAME DIE*. What is changed is the *SCREENING*. For example, a JAN 
transistor part would have beta specified over 3 different collector 
currents, over the full temp range. A 'MPS' (Motorola Plastic Semiconductor) 
part is only 25C speced. That sort of thing. The also screwd with the Vce 
specs. Many JAN parts were 60V, but most big users  (like Tandy) were 
turning on 5V stuff so the MPS parts were low voltage, like 30V.

Paul S.
/has *hardbound* TI transistor databook on shelf, as many pages as a Mouser 
catalog but weighs more