>speaking of air conditioners,  I see some of the semiconductor based
>SOLID STATE cooling technology spilling out into the electric-operated
>cooler market.  These things are nice, since they use NO compressors,
>no coolant gas, no pipes.  In fact, the only mechanical parts are the
>cooler fans.

>I wonder  how hard it would be to line up a bunch of these units along the
>cool air ducts to provide environmentally friendly air conditioning.
>At the current state of technology, I would expect it to need easily above
>100 amps at 12V (maybe more), but maybe someone knows more about these
>little thingies...

This sounded great to me so I did some quick calculations.  The first 
problem is thermal efficiency.  Handbook efficiencies look like about 10% 
for Peltier coolers.  So 12V x 100 amps would produce about 120 W of 
cooling.  The heat capacity of air is about 1 kJ/kg-degC, lets assume 20 
degC of cooling:

mass of air cooled per sec. = 0.120 / 20  = 0.006 kg

density of warm air is about 1.15 kg/cub. meter

volume of air = 1.15 x 0.006 = 0.0069 cub. meter or 6.9 liters /sec

at a delivery velocity of 1 m/sec my duct size would be let's see:

Q = V * A

6900 cc/sec = 100 cm/sec * A; A = 69 cm sq or 8.3 cm by 8.3 cm

So my 100 amp circuit cools one 3.25 inch square vent inlet. hmmmm, must 
greatly increase power or efficieny or both, but 1200 W input is already is 
almost 2 hp so I can't go up too much without dragging down the engine.  
Anybody know how many hp a typical A/C unit demands?

Problem 2 is that my books suggest that coolers use "alloys of lead, zinc, 
tellurium antimony, bismuth germanium, arsenic, manganese, cobalt and 
silicon (aka semiconductors (some no problem some toxic) and heavy metals).  
If we now mass produce these items for the automotive market we could easily 
generate a great deal of hazardous waste and mining and smelting activity so 
is this solution really green?

The above is certainly not intended as a flame.  I would really like to see 
some creative thought into alternatives to the current approaches, this one 
might lead somewhere with a little engineering.

Charles Vaughan
'84 GL