I'm not too impressed with the Electronics Cooling article. The
thermoelectric effect is treated as magic--the two integrals at the
beginning of the article just integrate the magic along the length of
the wire without explaining anything. There's some useful applications
advice, but there's some pure nonsense, e.g. saying that 20 wire
diameters' worth of lead length is enough to get a good measurement of
gas temperature. The ratio of the wire's thermal conductance (in W/K)
to its surface area goes as diameter/(length**2), so for a given
accuracy, the required lead length goes as the square root of the wire
diameter.
I agree that Omega isn't the best place to look for the actual physics
of thermocouples, but they do have nice pictures. Anyway, that physics
is more or less bottomless...you can stick with classical thermodynamics
and use the grand canonical ensemble, but then you get into actual solid
state physics and have to worry about things like the density of states
differences in different crystal orientations, and then you get into the
real quantum mechanics of disordered systems stuff. As I said,
bottomless, and although I talk a good game, my actual solid state
physics expertise goes about ankle deep. (I did take graduate solid
state from Walt Harrison, who is the biggest wildman in all of
theoretical sold state physics, but didn't pay enough attention. Same
with graduate statistical mechanics.)
I'm not the worst offender, though. A lot of the explanations you hear
about physics have as much merit as the Friday afternoon stock market
guy 'explaining' what happened on Wall Street that week. The idea that
an electret mic is a variety of capacitance mic is one example, and
almost anything mentioning 'surface states' and 'traps' is another.
It's not that traps and surface states don't exist, but they're very
commonly used as a cloak for ignorance--understandably, since real solid
state measurements are hard, and tend to involve ultrahigh vacuum.
All of that said, for instrument purposes delving into the fine details
of thermocouples is putting lipstick on a pig.
Cheers
Phil Hobbs
