You have to obey all the limits. The maximum power is a lot less than maximum voltage * maximum amperage.
Remember, dissipating power is what makes the transistor hot.
For bipolars, the absolute max values are not as useful as the safe-operating-area plot:
You'll notice one of the bounds is the maximum voltage, one is the maximum amperage, and one is the maximum wattage. Stay within the bounds to keep your transistor happy. (Also you need to avoid overheating: look at the Pc-Ta plot, or derive it yourself with the junction-to-ambient thermal resistance.)
I have a "National Semiconductor Discrete Databook" from 1978, listing the semiconductor process used to make each of their transistors (including JFETs) at the time (e.g. BC547, which was Process 04 and package 77).
Then there follows several pages for each process, showing the mask pattern for each process as well as all its electrical characteristics, including bandwidth and noise at different collector currents.
That way, all the transistors using Process 04 can be described in the same few pages of data, rather than repeating the same information for every transistor.
Fairchild may publish a similar "Discrete Databook" with the same information.
My NS databook doesn't feature a Process 61. (60,62,63 yes) I don't know if there's a correspondence between Process numbers across different manufacturers but there might be, so that they were genuinely compatible second sources for each other.
If you can find the equivalent BC547 datasheet you might check if it's process 04 or not, that might give us a clue about that.(EDIT : it doesn't name the process)
In the NS book, process numbers seem to be grouped by device type as follows:
- Process 0x to 4x : NPN
- Process 5x : N channel JFET
- Process 6x to 7x : PNP
- Process 83,84,86 : dual N channel JFET
- Process 88,89 : P channel JFET
- Process 9x : N channel JFETs, some (93-95) dual
And a separate range labelled 2C up to 5K for power transistors. Even numbers NPN, odd numbers PNP.
I wonder if there was some gentleman's agreement or de-facto standard or register of process numbers?
EDIT: a little more digging - namely google "Process 60" PNP transistor
shows the first hit is the same databook in an utterly useless form - OCR'ed text only - no tables, drawings or graphs!
Also a NS MPS8550 datasheet has the same comment "sourced from process 60" (which has apparently been uprated from 500mA collector current to 800mA since 1978!) but I can only see Nat Semi transistors when searching for Process 60.
The same search on Process 61 brings up only Fairchild devices as far as I can see.
EDIT : (last one, I promise!) First evidence of process number sharing between Fairchild and Nat Semi.
Process 58 according to NatSemi (above) is a N-channel JFET (and an impressively low noise one, under 2nV/rtHz at 10mA drain current, if anyone needs a source follower for a capacitor microphone! It's designed for switching, but the desire for relatively low on resistance gives it the large area required for low noise) Sold as the 2N5432 or J108,J109,J110.
Fairchild sell the same part under the same names (J108 to J110) and the datasheet says "Sourced from Process 58".
Best Answer
It means that the current for charging the Base is the same as the current used to discharge the Base capacitance when they do fall time testing. An exemplary circuit to do so is shown in this datasheet.
You may also have a look at:
How to determine BJT switching time