Robert Heinlein said, in Farmer in the Sky, that if you can't work something out to close order magnitude on the back of an envelope, you shouldn't call yourself an engineer.
I agree -- but I did find that modeling my learning after his precepts made me quite overqualified when I entered that job market compared to most. Still, not exactly a bad place to be, eh?
This particular one came up with a guy I'm mentoring off the forum, but it's instructive, and no need to name names for this.
He assumed that millions of neutrons a second was real power output, and we were going to do those kinds of things right off, when in fact I know of only a few people and places
that are even getting to that (I'm one). Too bad atoms are so tiny and the exponents involved with all this are all very large or very small.
So, for fun, and this isn't precise, just close order-magnitude, let's do the numbers here (I actually did do this on an envelope w/o a calculator). The result comes out fairly even numbers, which is fun too.
Joules per ev is 1.602...e-19. Takes a lotta eV to be a joule, in other words.
Ok, a million neutrons/second implies about twice that many fusion events with D-D, with a very gross mean output of about 3.5 MeV per. Call it 7 Mev/neutron output.
So we have 7 * 1.6 e6 * e6 * e-19 worth of power in a million neutrons/second fusor. Multiplying the rest of the way through gets us 11.2 e-7 joules/second, or about 1 microwatt.
Most of the fusors that get there in the normal modes put in about 400w (using Richard Hull's data from the last HEAS, 2009, which I duplicate here as well).
No, we're not boiling that cup of tea with our outputs! The one microwatt per million is a nice easy to remember number, handy. My best ever weird pulsed mode made maybe 500x that output/input ratio, still not that useful, as so far it doesn't scale to higher output well, just lower input, but a step on the way (I'll be documenting that fully as soon as my data aq stuff is rolling well), but we dupe the results here all the time.